Campus Gun Control Works- Why Guns and Schools Do Not Mix


Had a long article published over at the Boston Review.  The column is reproduced in full below:

Why Did Chris Die?

After his son Christopher was gunned down near the campus of the University of California, Santa Barbara on May 23, Richard Martinez sounded what has become a famous plea.

“Why did Chris die?” he asked, choking back tears. “Chris died because of craven, irresponsible politicians and the [National Rifle Association]. They talk about gun rights. What about Chris’s right to live?” He went on, “When will this insanity stop? . . . We don’t have to live like this.”

In response to Martinez’s impassioned appeal for gun control, the cavalcade of bumper-sticker slogans rolled in—“guns don’t kill people, people kill people,” “control criminals, not guns,” “don’t punish law abiding citizens,” and so on.

The NRA has been silent on the shooting, as is its usual media strategy following high-profile gun violence. But we know its position: the solution to gun violence is always more guns.

Thus the express goal of the NRA and other pro-gun groups is to promote the concealed carrying of firearms on college campuses. As the NRA puts it, “Colleges rely on colorful ‘no gun’ signs, foolishly expecting compliance from psychopaths.”

To this end, the NRA and state legislators are pushing guns at every level of schooling. The lobby backed a new Indiana law that allows guns on school property, so long as they are contained within parked cars.“Teachers have to leave their 2nd Amendment rights at the front door when they go to work,” said Indiana Senator Brent Steele, explaining why he supported the measure, in spite of the fact that the courts have never wavered on the constitutionality of gun bans on school property. A bill in Nebraska, if passed, would allow teachers and school employees to carry concealed handguns in schools. In Idaho Governor Butch Otter recently signed a law that allows residents with “enhanced concealed-carry permits” to keep firearms on college campuses. A similar bill passed a Florida Senate panel but ultimately was voted down.

The consistent refrain from conservative lawmakers and the gun lobby has been that such legislation will enhance security in schools. The logic is that if students and teachers are armed, or at least protected by armed guards, shootings such as those at Columbine High School in 1999, Virginia Tech in 2007, Sandy Hook Elementary School in 2012, UCSB, and, yesterday, Seattle Pacific University, either will not occur or will be less deadly.

Yet the evidence points in the opposite direction. Schools, including college campuses, exemplify the success of gun control. Though our schools are far deadlier than those of other countries with stricter gun control policies, they are safer than other corners of America that lack stringent constraints on guns.


How Safe Is School?

Despite the fact that the United States compares favorably to other high-income nations in terms of school bullying rates, we are the exception in terms of lethal school violence. Themost comprehensive study of school shootings to date—encompassing thirty-eight countries between 1764 and 2009—found that the United States had one less mass shooting than all the other countries combined.

The disparity in lethal school violence between the United States and other countries is almost entirely a function of firearm prevalence. It is not a coincidence that, in the United States, the vast majority of mass killings are carried out with a firearm, while in China, which had the second highest rate of mass killings in the dataset, not a single one was carried out with a gun.

But while American schools may be less safe than their international counterparts, they are still among the safest places in the United States.

Among school-age children, less than 1 percent of homicides occur either on school grounds or on the way to school, even though far more than 1 percent of students’ time is spent in school and en route. A Justice Department study showed that, between 1995 and 2002, college students between the ages of eighteen and twenty-four experienced 24 percent less violence than non-college students in the same age group. When college students experienced violence, it occurred off-campus 93 percent of the time.

These sanguine statistics are a reflection of the near universal prohibition of firearms by academic institutions. At least thirty-eight states ban firearms on school grounds, and sixteen explicitly prohibit concealed carry on campus. Such policies enjoy massive public support: according to one survey carried out by researchers at the Harvard School of Public Health, 94 percent of Americans feel less safe when fellow citizens “bring their guns into restaurants, college campuses, sports stadiums, bars, hospitals, or government buildings” and “overwhelmingly, the public believes that in many venues gun carrying should be prohibited.”

So just what sort of effect would guns on school grounds have? For starters, we can be confident they would not decrease school violence.


Public Carrying Doesn’t Reduce Crime

One of the intellectual touchstones behind the pro-gun movement’s support for extending concealed carry permits to schools is John R. Lott’s book More Guns, Less Crime, first released in 1998 and since updated twice. In response to the book’s claims, a sixteen-member panel of the National Research Council convened in 2004  to address the relationship between right-to-carry laws and crime rates and concluded that the existing evidence did not support the more guns, less crime hypothesis. A reexamination of the NRC’s findings in 2010 found that, at best, concealed carry laws have a negligible effect on crime rates and, at worst, concealed carrying increases rates of aggravated assault. Two legal scholars, Ian Ayres and John Donohue, further reviewed Lott’s findings and discovered that his data contain numerous coding and econometric errors that, when corrected, yield the opposite conclusion: right-to-carry laws increase crime. This was the second time Lott presented findings with “convenient” coding errors, and, when confronted by Ayres and Donohue’s research, he removed his name from a paper that claimed to confirm his results.

One of the largest and most recent studies on gun violence in America concludes that widespread gun ownership is the driving force behind violence. The study compiles data from all fifty states between 1981 and 2010 to examine the relationship between gun ownership and homicide. After accounting for national trends in violent crime as well as eighteen control variables, the study concludes, “For each percentage point increase in gun ownership the firearm homicide rate increased by 0.9%.” This research is consistent with evidence showing that even in “gun utopias” such as Israel and Switzerland, more guns means more violence.

Another large study compared 91 case workplaces with 205 control workplaces and found that workers whose job sites allow guns are about five times more likely to be killed on the job than are those whose workplaces prohibit all firearms.

Given the weight of evidence demonstrating the danger of carrying guns in public settings, it is extremely unlikely that more guns would make schools safer.


Why Allowing Guns on Campus is an Especially Bad Idea

In a recent editorial in the Chronicle of Higher Education, former Idaho State University Provost Gary Olson spoke to the realities of firearms on campus, their limited potential to improve safety, and the near certainty that they would have the opposite effect. “There is no recorded incident in which a victim—or spectator—of a violent crime on a campus has prevented that crime by brandishing a weapon,” Olson wrote. “In fact, campus police officers report that increasing the number of guns on a campus would increase police problems exponentially, especially in ‘active shooter’ situations.” Ninety-five percent of university presidents share his opposition to concealed carrying on campus.

If we take a sober assessment—one that will be sorely lacking at college keggers—it is not difficult to imagine the ramifications of widespread gun ownership at colleges. Alcohol abuse, bullying and hazing, high population density, and academic stressors are all predictive of violence—and all are ubiquitous on college campuses.

Guns and Alcohol Don’t Mix

Thirty-one percent of college students meet the DSM-IV criteria for alcohol abuse, andalcohol is used in 95 percent of violent crimes, 90 percent of rapes, and 66 percent of suicides among college students. Alcohol consumption renders police officers, people trained to use firearms, unfit for duty, so what should we expect from students who lack the preparation and discipline of police officers?

The most recent survey of firearm ownership on college campuses found that gun-owning students are more likely than non–gun owning students to engage in dangerous behavior such as binge drinking and, when inebriated, participate in activities that increase the risk of life-threatening injury to themselves and others. These include drunk driving, vandalism, and physical violence.

Given excessive consumption of drugs and alcohol on campus, the best a college can do is take precautionary measures to minimize the chance that lapses in judgment and drug- or alcohol-induced impulsivity will become lethal in the presence of a firearm. The only way to do this is to prohibit or at least strictly control guns on campus. It is simply not possible for campus police to monitor every party to ensure that those possessing guns are sober enough to do so. In any case, gun control is practically required in light of court rulingsthat force universities to provide safe premises to residents and visitors. Universities can be held liable for criminal assault on school grounds and for negligence in connection with social life on campus.

It should be obvious that the combination of alcohol abuse and firearms increases the potential for serious violence. After all, the archetypical “rational actor” is painfully sober. On a typical weekend, the average college student hardly fits the profile of a “good guy with a gun” advanced by gun advocates.

Accidents Happen

Even without the presence of alcohol, accidents happen much more often than gun advocates would like to admit. And when accidents happen with guns, they are often deadly. Individuals in households with firearms, for example, are four times more likely to die of accidental death than those in households without firearms.

The NRA supports bills that permit guns to be carried in vehicles on school grounds, arguing that firearm owners should not be punished for accidentally leaving a gun in their car. Curiously, there seems to be little concern for what happens if the same careless owner accidentally forgets to lock his car, accidentally fails to put the safety on, or accidently pulls the trigger, ad infinitum. It seems clear that there are many more ways to accidentally go wrong with a gun than there are ways to go right, and this is especially true in a densely populated, anxiety-ridden, alcohol-saturated, hormone-fueled school environment.

Guns and Suicide

While suicide is the second leading cause of death among college students, the rate ofabout 6.5 to 7.5 per 100,000 is roughly half that of a matched non-student population. The difference in suicide ratesbetween student and non-student populations is explainedalmost completely by the reduced access to firearms on college campuses. Consider that suicides committed with firearms represent only five percent of suicide attempts but more than half of suicide fatalities. About 1,100 college students commit suicide each year, and another 24,000 attempt to do so. Given that suicide attempts with a firearm are successful90 percent of the time, each one of these more than 25,000 attempts would almost certainly result in death if carried out with a firearm.

The best studies to date show that the majority of suicides are impulsive,with little deliberation prior to the act. We also know that youths between the ages of eighteen and twenty-five experience the highest rates of mental illness in the general population. These factors, combined with high rates of alcohol and drug abuse, provide a compelling reason to believe that the nation’s suicide rate will increase if firearms are allowed on college campuses.

Gun Theft

According to a Department of Justice report, between 2005 and 2010, an average of 232,000 firearms were stolen each year, primarily in residential burglaries. In a survey of incarcerated felons, about one-third of respondents report having stolen their most recently acquired handgun.

Adorm room is one of the least secure places to store a firearm. School dormitories are small, cramped, shared spaces, and they receive a large number of visitors. It would be difficult to conceal the fact that a dorm resident owns a firearm; more likely, the student would flaunt this fact. This means it is a lot easier for a thief to identify potential targets and successfully steal a firearm. And once a gun is stolen, it is much more likely to be used in a crime than if it were in possession of its rightful owner.

Armed Students Are Unlikely To Stop Shooters

Even if a student or professor were to confront a shooter, their chances of stopping a bad guy with a gun would be slim. This should be self-evident given that New York City Police, for instance, only hit their target in 18 percent of cases. The average student or professor would likely have a substantially lower hit rate, thereby increasing the threat to innocent bystanders.

20/20 segment, “If I Only Had a Gun,” showed just how hopeless the average person is in reacting effectively to high-stress situations. In the segment, students with varying levels of firearm experience were given hands-on police training exceeding the level required by half the states in order to obtain a concealed carry permit. Each of these students was subsequently exposed to a manufactured but realistic scenario in which, unbeknownst to them, a man entered their classroom and begin firing fake bullets at the lecturer and students.

In each one of the cases, the reaction by the good guy with a gun was abysmal. The first participant, who had significant firing experience, couldn’t even get the gun out of his holster. The second participant exposed her body to the assailant and was shot in the head. The third, paralyzed with fear, couldn’t draw his weapon and was shot by the assailant almost immediately. The final participant, who had hundreds of hours of experience with firearms, was unable to draw his weapon and was shot at point blank range.


Stand Your Ground

A recent New York Times article, in brilliant tongue-and-cheek, exposes some harrowing prospects that could result from arming college campuses. The author satirically asks if students using laser-pointers in class or arguing over coffee is sufficient cause to fire away. While this may sound absurd, lax gun laws have created shooting scenarios just like this. In recent years, people have been shot over skittles, popcorn, and their choice of music. It is easy to think up a whole laundry list of relatively common occurrences that could provide legal justification to shoot at a student.

Heightening the risk of needless bloodshed, the states most likely to push for guns on campuses often have stand-your-ground laws as well. As Judge Debra Nelson told jurors in the trial of George Zimmerman for the killing of Trayvon Martin, Zimmerman “had the right to stand his ground and meet force with force, including deadly force if he reasonably believed that it was necessary to do so to prevent death or great bodily harm to himself or another or to prevent the commission of a forcible felony.”In other words, in a stand-your-ground state, authority to end another person’s life rests with one’s own perceptions and convictions, with all their attendant biases. In a high-stress environment such as college, where rationality can be sorely lacking in dangerous moments, the presence of a gun can only make the situation worse, and stand-your-ground laws provide ample room to shoot first and justify later.


Back to School

You are in college. You show up at a fraternity party late one weekend. You don’t know much about those attending, except that some may be carrying a firearm due to a new policy permitting concealed carry on campus. Do you feel more or less safe knowing that some of the party attendees may be armed and intoxicated?

If you are like 94 percent of Americans, you feel less safe knowing that people in your community carry guns into public spaces such as colleges. But we need not rely only on the public’s expressed preferences when it comes to gun control in schools. The evidence is clear. While gun advocates complain that control measures don’t work, the case of our schools—and workplaces—stands as a sharp rebuke: where guns are carefully controlled, there is less gun violence. And where young people are most vulnerable to heavy drug and alcohol use, accidents, theft, poor judgment, and impulsive behavior, more guns won’t mean less crime but more mayhem.


*correction: Previously the article mentioned that the NRC panel met in 2004 and 2010. They only met in 2004, and it was a separate study reexamining the NRC’s findings that was issued in 2010.

Related posts

  • Kyle

    I agree on issues regarding alcohol and drugs with guns on campus, but regarding carrying guns into workplaces, I am not sure I agree. Criminals can just as much carry guns into workplaces to shoot the place up if they really want to. That has happened. Not allowing guns legally into a workplace would only stop possible crimes of passion, where someone decides while at work that they want to shoot. But what can happen is someone gets too stressed and takes a gun into work and shoots people.

    It is also true that people can occasionally be shot over minor arguments where legal concealed carry exists, but plenty of people also are not shot over such things as well. Areas with concealed carry do not turn into the Wild West as the gun control proponents always argue will happen.

    Regarding Richard Martinez, the guy’s son who clearly had mental problems, goes and stabs three people to death, then shoots some others, and hits one with his car, and his choice of blame? Guns and the National Rifle Association. Maybe he ought to blame the police for not doing their job appropriately. It is clear that people were going to die from this man anyway as he was stabbing them and rammed one with his car. He talks about his son’s right to live. One’s right to live does not mean infringing on the natural rights of others. If someone insults Islam and an extremist of Islam blows someone up in response, we wouldn’t call to end free speech in response.

    I also do not agree with the 20/20 segment. Regular police officers carry guns and they aren’t much better trained than regular citizens in how to gun fight. Most of their gun shots miss in such scenarios. The logic seems to be that because a person may not be able to draw out their weapon in time, that it is thus okay to ban their carrying of a weapon. There are, according to the gun control proponents, hundreds of thousands of defensive uses with a gun each year, which far exceeds the 15,000 or so gun homicides a year.

  • Spree shootings are almost exclusively in areas where guns aren’t allowed, with the attempted Giffords assassination the only counterexample I am aware of. Probably the worst response to spree shooters is to add more places where decent people can’t carry unless it is coupled with a gun ban so strict as to require a repeal of the second amendment. Even this would be unlikely to end spree killings–someone will figure out an effective alternative, the media will publicise the details and the next killer will have a starting point to refine the method.

    Schools are ‘less deadly’ because of demographics rather than gun laws–Elementary age school children aren’t likely to be murderers, gang members of college age aren’t generally college students. Even if guns are allowed to be carried in toy stores and banned in bars (as was the case here in Ohio until recently) the bars will have more violence, and more gun violence (as was and is the case in Ohio)

    No evidence that someone brandishing a weapon where guns are not allowed…proves what, other than the people whose carry may be beneficial generally follow the law so don’t have a gun available to brandish? Where has a legally carried civilian gun made any active shooter situation worse? Active shooters end their spree as soon as they come across a defender with a gun. (I’ve yet to find even one counterexample, feel free to point out any I missed) The most critical factor in how many victims is the time the shooter is unopposed–legal carry won’t increase that time, and in some cases shortens it.

    Eliminating restrictions on carry on school grounds would not eliminate existing restrictions on drunken carry, nor would it give 18 year olds the ability to get a carry license.

    “individuals in households with firearms are 4 times more likely…”. Is that limited to legally owned firearms, or like most anti-gun statistics does it mix criminals to increase the violence statistics? Is there any gun control study that compares the violence rate of license holders to the general population?

    “Accidentally pulls the trigger”–maybe negligently, but it’s pretty hard to do accidentally. If anything that’s an argument to minimize handling–allow people to leave their guns holstered where there is no chance of accidentally pulling the trigger. I’ve also not seen anything where the penalties for negligent discharge would be reduced.

    There is a correlation between the seriousness of the attempt and the method chosen–someone with “valid” reasons is more likely to use an effective method, There’s a difference between a jilted lover taking a handful of Tylenol and someone who is terminally and painfully ill who can’t take it anymore, to use two incidents I’m personally familiar with. There is also a considerable difference between guns kept in a dorm room by 18 year olds, and allowing a 50 year old license holder to carry.

    “If I only had a gun” was obviously designed from the outset to make guns fail. Even an ordinary holster would be difficult to draw from with the grossly oversized thick sweatshirts that the students were required to wear. The retention holsters used are designed for police officers who may have to wrestle with criminals rather than what a typical civilian would use. They are designed to require specific and abnormal motions to draw, even without the oversized sweatshirts. In at least one of those cases, the belt was absurdly loose, something anyone with more than a few days experience with carry would not do. The “gunman” was a highly skilled police firearms trainer who appeared to know exactly where the “defender” would be sitting. There is also editing to consider–were successful defenses left on the cutting room floor? Was the gear modified until it failed?

    How about this for a more realistic scenario-Instead of unfamiliar gear, absurdly bulky clothing and paintball, let a few people with carry licenses use their own gear and a laser trainer gun matching their own. History is that gunman attack multiple classrooms, so the gunman won’t know which room to expect the defender–sometimes the defender will hear shots and screams before the gunman appears.

    The record of police marksmanship is indeed abysmal. The record of civilian license holders in self defense situations is phenomenally better. To be fair, we aren’t required to detain people and we are allowed to run away.

    Whether or not Zimmerman was right or wrong, Stand Your Ground was not a factor, Zimmerman’s defense didn’t use it. . If Zimmerman attacked Martin, neither Stand Your Ground nor self defense would apply, if Martin attacked as Zimmerman claimed, Stand Your Ground was not necessary. The only thing Stand Your Ground did was keep Zimmerman out of jail until political pressure. Can you find examples where people “got away with murder” due to Stand Your Ground where simple Self Defense wouldn’t be enough for an acquittal?

  • Arch Stanton

    The 20/20 video seemed stacked against the students. Despite varying levels of experience, not one student had ever carried a concealed weapon ever. The students were laden down with protective gear and extra long shirts that hung up when needed. The students were also seated in the exact same place and they were the second one targeted by a highly trained police instructor. How about running the test again, after allowing the testers time to train with their weapons, wear their own clothing and allow them to sit anywhere in the room? Or, throw a few local CCW holders with as much experience as the police instructor to participate? It seems like 20/20 went out of their way to make self defense difficult.

  • I was never that convinced with the 20/20 video.. however, we know for a fact that the body reaction to stress makes the likelyhood of a “good guy” stopping a “bad guy” to be rare event… and is backed by several studies.. Some people think the situation is unrealistic but in an open carry situation where the shooter scouted the location he/she may know which targets to “take down” first.. Situations where t-shirts catch firearms, or holster malfunctions also happen..

  • When has a spree shooting been made worse by an armed good guy? I’ve never heard of an incident where anyone but the shooter or in rare cases the armed good guy got hurt after the armed good guy shows up. “an open carry situation” is a red herring, even where legal it remains rare–how many openly carried guns have you seen, not counting cops or gun events?

  • Max

    “Another large study compared 91 case workplaces with 205 control workplaces and found that workers whose job sites allow guns are about five times more likely to be killed on the job than are those whose workplaces prohibit all firearms.”

    This does not imply causality in the slightest. Jobs that allow the usage of guns tend to be high-risk such as security and whatnot. This reminds me from that study from Branas’ which Wintemute called out.

    Where any of the homicides committed with guns carried by employers? The study does not address that.

    “A 20/20 segment, “If I Only Had a Gun,” showed just how hopeless the average person is in reacting effectively to high-stress situations.”

    In this garbage experiment, the shooter knew who was carrying and who wasn’t. It proves absolutely nothing.

    Nonetheless you cannot cite any empirical evidence that violent crime has increased amongst schools that allow carry, and that’s because there is no evidence.

    This blog has some really well-written posts, sadly this is not one of them.

    • Thanks for your well-written reply, Max. A couple quick comments

      –Correlation does not equal causation has become the default response by people who are simply uninterested in reading studies. These guys are PhD’s who are fully aware of what correlation entails. That being said, the study adjusted for workplace characteristics and preventative measures– “further analysis suggested that the increased risk associated with employer policies allowing guns was not completely explained by either characteristics of the workplace that may be indicative of its inherent ‘riskiness'”

      –There seems to be this bizarre preoccupation with cherry-picking among readers of this blog. I have said this before: not all studies were created equal. The point of this essay was to catalog all the evidence available. There may be individual flaws in many of the studies, but when study after study points to the same conclusion, over and over again, across disciplines, across institutions, across states, across countries, across time, then you can begin to accept an emerging consensus.

      The probability that all these studies just are getting it wrong over and over again, with one small errors, time and time again, across institutions, and states, and countries is astonishingly small. That being said, the preponderance of evidence related to suicide, accidents, theft, and homicide suggest that status quo prohibitions and regulations of guns on campus are a good idea.

      –The 20/20 segment is obviously problematic. It’s not a scientific publication. Unfortunately, there are simply no studies on the question of the efficacy of concealed carriers to respond to high-stress situations because such interaction is inherently low probability. As such, I found the best investigation of the issue that I could. If you could find a competing journalistic investigation, I would think differently about how useful shooters are in high-stress situations.

      –The reason I can’t cite a study that shows “violent crime has increased amongst schools that allow carry” is because there simply isn’t enough sample size of schools that permit concealed carry on campus, nor enough years in the dataset to examine variation in school violence (which is already small — largely thanks to gun control). If I were to show increases in violence among schools that permit guns, you could just as easily make the same accusation as you did in the workplace study: “the schools that permit guns did so because those schools are risky.” You are not availing yourself to opportunities to be disproven. I’m not sure you want to be persuaded.

      • Max

        “Correlation does not equal causation has become the default response by people who are simply uninterested in reading studies. These guys are PhD’s who are fully aware of what correlation entails. That being said, the study adjusted for workplace characteristics and preventative measures– “further analysis suggested that the increased risk associated with employer policies allowing guns was not completely explained by either characteristics of the workplace that may be indicative of its inherent ‘riskiness’”

        It’s absolutely vital to know if any of the homicides were committed by any employers who were carrying a weapon. Were socioeconomic factors adjusted for? Perhaps workplaces that allowed guns were located in rough areas? I am reminded of the study by Arthur Kellermann where only a minority of homicides involved household guns.

        “The reason I can’t cite a study that shows “violent crime has increased amongst schools that allow carry” is because there simply isn’t enough sample size of schools that permit concealed carry on campus, nor enough years in the dataset to examine variation in school violence (which is already small — largely thanks to gun control). If I were to show increases in violence among schools that permit guns, you could just as easily make the same accusation as you did in the workplace study: “the schools that permit guns did so because those schools are risky.” You are not availing yourself to opportunities to be disproven. I’m not sure you want to be persuaded.”

        No, it’s because there is a lack of empirical data. Schools that allow the carry of weapons have yet to see an uptick of violence or homicides as you seem to imply greatly in your article.

      • “There seems to be this bizarre preoccupation with cherry-picking among readers of this blog. I have said this before: not all studies were created equal. The point of this essay was to catalog all the evidence available. There may be individual flaws in many of the studies, but when study after study points to the same conclusion, over and over again, across disciplines, across institutions, across states, across countries, across time, then you can begin to accept an emerging consensus. ”

        Except there isn’t an emerging consensus. There are plenty of studies that have been presented that you have stated are “not valid”, while you trot out obviously unethical studies where the numbers or methods were cooked to produce a set result.

        This is a horrible combination of confirmation bias and appeal to authority while ignoring that yes there are lots of scientists who are perfectly willing to adjust their methods or pick their numbers to make sure that the grants keep coming.

        It shows that you are 100% agenda driven rather than actually interested in your stated goal of public safety.

        • I’ll make this simple: Find half the number of peer reviewed academic studies on gun violence we have in our database that agree with you. I’ll even round the number you have to find down to 60 for you. If you find this many (even including the chronically dishonest work of Lott and the shoddy research of Kleck), you have an excuse to claim we are cherry picking. Otherwise, you can stop pretending that there is anything close to a similar amount of research supporting our respective sides. Best of luck.

          • Is that including the shoddy and dishonest work by Kellermann, Hemenway, and Branas?

            Garbage in, Garbage out!

          • If you wish to exclude Hemenway, Kellerman, and Branas, then I request you exclude any paper that was written by Lott or Kleck (no matter how many coauthors that paper may have). In this case find 40 peer reviewed academic studies on gun violence that agree with you. Otherwise you need to find 60.

            As I am but a mere “cherry picker” and you clearly an expert on the literature, this should be no problem for you. Otherwise, it would be more than a tad hypocritical to label us as cherry pickers when you can’t find even half the peer reviewed studies we provide (and this isn’t even including the dozen or more studies I have found since I last updated the database).

        • People always lump Hemenway in with Lott or Kleck as if it just two sides of the same coin. This is just wrong.

          You find me a Lott or a Kleck study and I will explain the specifics of what is wrong with the models, what is wrong the data, what is wrong with the analysis, and so on. I have done this many times throughout the blog.

          Whenever I bring forth a Hemenway study, the only response I get is, “Oh! Joyce Institute! Everything is biased!!!” I have never, not even once, received a comment on this blog explaining what it is that’s wrong about the content of the studies I cite. Why? Because nearly everyone (with the exception of Kleck and Lott) believe that Hemenway’s work is the GOLD STANDARD in gun violence research. They are consistently the most cited work, published in the highest quality academic journals, and Hemenway doesn’t have any history of academic dishonesty. If readers of this blog cared to open their eyes, they could read about the laundry list of egregious broaches in integrity that colors Lott’s history (making up data in the 1998 research, secretly changing models in-between publications without telling anybody, removing his name from papers once he’s been caught with lying, pretending he is a girl named Mary Rosh on the internet in order to defend himself, and so on.)

          • I’m glad you posted first; I wouldn’t have been so kind. Prior to 1990s, CDC research into “gun violence” was “bias”, now all research not done by pheudo-scientific shils like lott, kleck/gert, … (and so on) is “bias”; its a “no win” situation…. One way or another research will continue, and everything pro “gun control” will be viewed as “bias” (in their own minds)…. which is the majority of said research…

  • Max

    Jus wanted to add one more thing regarding the study:

    “Our data suggest that, much as residents of households with guns are more likely to become victims of homicide,4,5 workers in places where the employer’s policy allows guns may have a higher chance of being killed at work. These findings bear directly on policy for workplace safety. In light of the evidence, it is reasonable to question the costs and benefits of polices permitting firearms in the workplace.”

    It explictly cites the same study by Arthur Kellermann that represented a completely spurious correlation. It is thus highly unlikely that most of the homicides involved guns belonging to the employees.

  • Arch Stanton

    “As such I found the best investigation that I could.” Seriously? That test was so stacked against the volunteers that it is painfully obvious that 20/20 didn’t want them to succeed.

  • John

    Regarding your latest Slate article: “Children from states where firearms are prevalent suffer from significantly higher rates of homicide, even after accounting for poverty, education, and urbanization. A study focusing on youth in North Carolina found that most of these deaths were caused by legally purchased handguns. A recent meta-analysis revealed that easy access to firearms doubled the risk of homicide and tripled the risk for suicide among all household members. Family violence is also much more likely to be lethal in homes where a firearm is present, placing children especially in danger. Murder-suicides are another major risk to children and are most likely to be committed with a gun.”

    Why do you refuse to mention that homicides are rarely committed with household guns?

    Why do you continue to perpetuate this myth? Is it because if this simple fact came to light, nobody would take these studies critically anymore?

    You have yet to reply on the new study that found no association between homicide rates and noncriminal gun prevalence, the authors address a fundamental problem that previous studies have failed to address, thus the spurious correlations:

    Did you publish the article under the assumption that most people are not well-versed regarding the literature?

    • John

      I await a reply from the authors from this blog regarding my concerns. Thank you.

      • Good luck with that John. While the authors always seem to be available to post more junk science to support gun control, they seem surprisingly absent when solid evidence of bad science for the sake of pushing an agenda is presented.

        That’s one issue with Appeal To Authority presented by the authors. Yes many of the paper cited are written by PhDs who are experts in their field….but they are in fields where all their bread and butter comes from grant money.

        It doesn’t matter if it’s public health, public safety, sociology, or government biology, when the author knows it the grants or public funding will slow or stop if the results section reads one way, and more checks will arrive if the results read the other way, many choose to continue their career leaving their ethics in the ditch, than finding another career.

        It isn’t right when they do it, but I can’t necessarily blame them.

        Still when the authors here, who are educated enough to know better re-publish their junk findings for no economic gain, I really start to frown.

    • For the deeply flawed Kleck rebuttal, I’ll merely copy and paste an earlier reply:

      “Yes, my response came from Kellerman’s response, just as your argument is basically copy and pasted from Kleck. As I mentioned, Kleck’s argument is far more flawed than even the alleged flaws of Kellerman’s study. Let’s take a closer look.

      “Here we can see that the risk factor Kellermann identified was solely attributable to handguns, rifles and shotguns were not correlated. Are we to believe that shotguns and rifles are not owned at home for the purpose of self-defense.”

      In the original study, 42.9% of the homicides were committed with a handgun, 2.4% with a rifle, and 3.6% with a shotgun. There weren’t enough rifle or shotgun homicides in the sample to even determine statistical significance. Further, the study was conducted in the most populous counties of Tennessee and Ohio (translation: urban). Handguns are a far more popular means of self-defense highly populated areas than rifles or shotguns. In rural areas, this may be different given the necessity of owning a rifle or shotgun for hunting or protecting livestock from animals. But for exclusively self-defense, the handgun is the weapon of choice. This is unsurprising.

      “Now here’s the problem, why does the risk factor decrease of the gun was unloaded? This would only make sense if the majority of decedents were murdered with loaded and unlocked firearms kept in their own homes. The issue is that they weren’t.”

      Incorrect, they only have to be a significant enough minority to sway the statistics, which they are.

      I will stop quoting from what you said from now on as it would take up a large amount of room and merely present my rebuttal.

      First, Kleck arrives at the 14.2% by taking data from Kellerman’s 1998 study and then extrapolates the 9.7% number from a study published in 1958. The problem with using the 1958 study are twofold. First, it’s ancient. Second, Kleck uses it to imply that cases where the owner in the house is not the victim are irrelevant. This is absurd. If there is an argument, and the owner of the house pulls out the gun and shoots the person he invited in, that certainly counts and does nothing to undermine Kellerman’s study (keep in mind that only 14% of the cases in the original study had signs of forced entry, meaning the other person in the house was overwhelmingly allowed in the house.

      Also, using the 14.2% number is deeply suspect. The 1998 study looked at both assaults and homicides, whereas the original study looked at just homicides. Guns are more deadly than other methods of assault, that is simple fact. Therefore, we would expect that this percentage should be higher when only homicides are included. But, for the sake of argument, let’s just assume that the 14.2% stat is accurate.

      Here is where Kleck makes another major logical error. He assumes that this is the only way a gun in the house could have a causal role in the homicide. This is just silly. There are 30 cases where the victim is shot with a gun kept in that home. There are also 21 cases where the victim unsuccessfully tried to use the gun and was killed. So there are at a minimum 51 cases where the household gun directly contributed to the death. At a bare minimum. This is ignoring the other ways the gun could have contributed (e.g. the victim went for the gun but never got to it).

      Now Kleck makes a massive category error. He compares these 30 cases to ALL of the homicides in the study area, and uses this as his damning evidence against Kellerman. He completely overlooks the fact that the study only focused on homicides in or around the home, and makes no claim about overall homicides. This is a stunning oversight. Even had he just limited his comparison to all of the homicides, it would have still been flawed. The correct way to look at the data would have been to focus on how these 51 cases (the bare minimum proper number) relate to the rest of the homicides in homes with guns. So, there are 191 cases in the study of homicides in homes with a gun. of these, at least 51 are a result of the gun in the house. The presence of the gun in the home then at the very least increases the odds of homicide by 1.36 (or 36%). Astoundingly, these rough calculations provide the exact same number Kleck arrives at in his own study, though he desperately tries to downplay his findings as insignificant. Keep in mind that this is the absolute bare minimum (and also best case scenario for Kleck) effect a gun in the house would have.

      Before continuing, it is worth noting that Kellerman’s warning against owning a gun is justified in every way. First, this analysis excludes suicides, and there is overwhelming evidence of a causal link between gun ownership and the rate of suicide. Second, look at the likelihood of effectively using the gun versus being killed by it. In 3.6% of the cases the homicide was justified. That represents about 16 cases (leniently assuming that all of these justified homicides were with guns). The odds in this case are 2 to 1 that your own gun is turned against you. Add in the other 21 cases (died trying) and this becomes even worse. Another Kellerman study found twice as many people lost their gun to the intruder than used it successfully. These odds are really bad for gun owners.

      And I would not assert causality based on the Kellerman study alone. However, not admitting that there is a very high likelihood of causality when there are 5 other studies that have examined this study and using different techniques and samples have come to a similar conclusion would be silly. Here is a very recent meta analysis that looked at each of these studies and arrived at the same conclusion:

      Oh, and the “bought the gun in response to a threat” logic would only make sense if the presence of a gun in the house also significantly increased the risk of non-firearm homicide. It doesn’t. Only one study (Cummings if you are curious) came close to a statistically significant positive effect, but not even it reached significance. Rather unsurprisingly, Kleck attempts to hype this effect as much as he attempts to downplay his own findings (and not very persuasively given his other massive errors in the response to Kellerman). Further, only in 14% of the cases was there forced entry, meaning in the vast majority of cases the acquaintance was certainly and acquaintance (or at least somebody you knew/trusted enough to let into your house, a label that would not apply to loan sharks wanting their money back).

      Given the strong evidence supporting Kellerman and the rather lackluster case brought by his critics, it would be foolish not to think that guns in the house have at least some significant harmful causal role on homicides in the home.”

      Now, for the other Kleck study. Reading beyond the abstract, it quickly becomes apparent that there are severe flaws that render the study nearly useless in drawing any substantive conclusions.

      The first (and most damaging as the rest of the paper rests on this assumption) flaw is the assumption that there is reverse causation between firearm ownership and gun homicides: i.e. people see the homicide rate go up and buy guns, or they see the homicide rate go down and sell their guns. This sounds reasonable until you realize that practically nobody knows what the homicide rate is where they live (and it what direction is may be going). A recent survey showed that most people think the homicide rate has been going up recently (it has actually been flat since the mid 90s). People may base some of their gun purchases on what they PERCEIVE the homicide and crime rate in the surrounding area to be, but this PERCEPTION (based far more on media reporting than actual statistics) is a far cry from reality.

      So already, any conclusions the study comes to (even if they would’ve happened to agree with me) are deeply suspect. Further, Kleck finds that is there is no reverse causation there is a strong, significant association between gun ownership and gun homicide (but not non-gun homicide).

      However, let’s just accept for the sake of argument that this assumption (which has no logical or empirical supporting evidence) is 100% accurate, and that the criminal vs. noncriminal dichotomy is a sophisticated and accurate assumption (it’s not). The study is still deeply suspect. Why? The instrumental variables are weak. So weak that the authors even admit they are weak and warn readers to treat the tenuous results with caution. The tables show that using barely different functional forms leads to wildly different results, which is not supposed to happen if your instrumental variables are solid. Another example of highly suspicious results (suspicious as in they don’t make logical sense) is that the authors find good guy gun ownership reduces gun homicide, but has no effect on non-gun homicide. This makes no sense as a bad guy with a knife should be just as deterred as a bad guy with a gun (even more so in fact).

      There are other flaws (less central to the core claims made by the study), but these alone would make any intelligent person well-versed in the literature to not take this study very seriously (at least not until there are 3+ more academics using superior methodology that come to the same conclusion).

      • Adam

        What John was taking issue with is because you refuse to mention the mechanism of which the guns-homicide association at the household operates, you mislead readers to believe being murdered with a gun kept in the household is a common thing when it isn’t.

        If people were aware of the mechanism that this study functions on and others by extension, would it be cited as critically as it is? No it wouldn’t. And if it had Kellermann would have told us long ago how many guns were the ones kept in the household. The fact that it was published in the NEJM without this datum being released shows how politicized it is. But then again, the NEJM also published the severely flawed Loftin paper in 1991, a study which you conviently forget to cite in your “database.” I digress.

        In the end it turned out that the claim of spouses or children grabbing the household gun and firing at someone else turned ou to be laragable tenable, did it not?

        1.36 is garbage and very weak. If that’s the risk involved, I’m going to own guns, and I’m going to do it big. Actually as somebody else mentioned on here, there’s good reason to assume it’s probably less, that is if you assume all unknown guns were the same ones stored in the house.

        Hell I would own guns and again do it big even if the odds ratio was 1.5 because even if it’s a 50% added relative risk, the absolute risk is very small. Safety queen’s really do grind my gears.

        • Adam

          largely untenable* apologies as I was posting from a phone.

      • John

        Here’s another study saying the same thing:

        The guns homicide association hinges mainly on culture.

      • tofiy

        Devin, here’s a really bad hemenway study I like:

        hemenway finds an association between guns and homicides, however virtually all this association can be attributed to the united states.

        If the US is excluded and fs/s is used as the proxy, the relationship ceases to be statistically significant. hemenway downplays this finding and instead mentions that the results remain significant when the cooks index is used (an outdated and inferior proxy, even the researchers stopped using it in future studies)

  • Adam

    So wait, you’re claiming that the reason why schools are safe is because they do not allow guns, Talk about the worst case of post hoc ergo propter hoc. If this was even remotely the case, the campuses that alliwed conceal carry should be a bloodbath by now. Except they aren’t, just like every antigun prediction in the recent years.

  • Paul

    I don’t see how you can claim schools are safe because they ban guns. It was quite common for high schools to have rifle teams and a range during much of the 20th century. School shootings were extremely rare back then.

  • ididthemath

    In the 1998 study by Kellermann et al he reported that 438 gun assaults and homicides took place in the home.

    He reported that one in five assaults resulted in a death (20%) which would be 87 deaths.

    Amongst those 438 assaults, 49 were committed with a gun kept in the household, 295 brought from elsewhere, and 94 which were not noted by police.

    20% of 49 equals 10 deaths due to household guns.

    20% of 295 is 59 deaths due to external guns.

    20% of 94 is 18 deaths due to unknown guns.

    10 + 59 + 18 = 87

    Conclusion: most household gun homicides do not involve the household gun.

  • Pingback: Texas Just Made College Less Safe | Gun Free UT()

  • Pingback: Should Guns Be Allowed On Campus? | Millennial Speech & Debate()

  • Pingback: Need help to write an Argumentative Essay, Should guns be permitted on college campuses? – Writer Jojopapers()

  • William Ashbless

    The 20/20 coverage of “If I only had a gun” is so thoroughly stacked against the participants that it’s use by Hughes and Defilipis show their commitment to their ‘academic and scholarly’ debunking of myths.

    The shooter(who is a professional range officer) enters the room and targets the instructor. He then selects the armed volunteer despite the fact the volunteer looks like everybody else and has a concealed weapon. So, the active ‘shooter’ knows exactly where the armed person is sitting(every time) and targets that person immediately after shooting the teacher. The thing about concealed weapons is the shooter wouldn’t know who would have a concealed weapon and the exact seat they would be sitting. FAIL!
    Wearing a new shirt with a new holster in the first minutes of those items being assigned to you leads to problems? Wow! Who could have foreseen that? Like anything new, your concealed carrier would have to try different holsters, positions of carry and clothing to see what works best for him or her.
    I’m curious as to why 20/20 didn’t recruit any off duty cops or people that had carried concealed even once before in their lives. It’s almost as if 20/20 wanted their participants to fail.