 After decades of silence, the CDC is speaking up about America's gun violence epidemic. Gun violence in America is a public health emergency. Gun violence is a public health issue. And we need to start thinking of this as a public health crisis. Republican congressman Jay Dickey is infamous for a 1996 law prohibiting the CDC from using federal funds to advocate for gun control. The Dickey Amendment. The so-called Dickey Amendment. Remember the Dickey Amendment? Which is effectively kept the CDC from studying gun violence? What inspired the Dickey Amendment was a blockbuster 1993 CDC-funded study, which claimed to show the owning guns made Americans overwhelmingly less safe. Having a gun in your home not only doesn't protect you, but it puts your family at risk. According to a recent film published by The New York Times, the NRA lobbied for the Dickey Amendment because the 1993 study made an overwhelming case against gun ownership. And it wanted to stop the CDC from ever backing studies of this sort again. The NRA didn't think that that would be good for business. And this really spurred them on to attack us with much greater fury. The problem with this story is that the 1993 study doesn't actually show what its proponents claim. It had serious statistical problems that stem from its approach to the topic of gun violence, underscoring why the Dickey Amendment was good policy. A public health agency isn't well suited to analyzing social science issues of this sort. This is not serious research designed to improve individual decisions or social policy. It's scientific trappings used to camouflage arguments no one would take seriously if they were presented clearly. Statistician Aaron Brown has taught at NYU and the University of California at San Diego. He's a columnist for Bloomberg and an expert on risk management. The initial conception of this study is that a gun is like a pathogen in your home that causes homicide in unspecified ways. A gun is like a homicide virus. Brown points out that unlike with death from disease, when a gun kills someone in a home, the surrounding circumstances are highly relevant. The researchers count all deaths equal, whether the homeowner or her family or a homicidal home invader. Someone not blinded by disease control thinking would focus on how these guns were used for self-defense, suicide, accidental injury, domestic violence, or seized by an attack ring used against the owner. This could be useful information for someone weighing the benefits and risks of keeping a gun in the house. A major flaw in the study is that it attributes homicides to the presence of guns in the home even if a gun wasn't involved. In fact, the majority of the homicides don't have a gun at all. And if a gun was involved, the researchers don't bother to ask if it was the homeowner's gun. The risk was 200% greater that someone in your family would be shot and killed with a gun. And the risk that someone in your home would commit suicide with a gun was five times greater, a 400% increase in risk. That may be what Rosenberg would have liked the study to find, but it did not. It did find that people living in homes with guns were murdered in their homes more frequently than people living in similar houses without guns, but only 25% more frequently, not 200% or five times. And far more dangerous than guns were living in a gated community, living alone or renting. Having been in trouble for drinking at work made you 20 times more likely to be murdered in your home. The politicized nature of this study is shown by ignoring the big risks it found and highlighting one of the smallest. But what this really shows is the methodology is absurd. No one would take it seriously if we applied it to rent the gated communities and no one should take it seriously applied to guns. Brown says that the authors also fail to establish convincingly that the 25% finding isn't the result of random chance. It is notoriously easy to find 25% or 50% effects of the kind shown in this paper for any effect you want. The study considered 388 people murdered at home and found that 174 of them had owned guns. They matched them with 388 similar people who were not murdered, 138 of whom owned guns. The first question you might ask is whether this difference could be the result of random chance if gun ownership has no effect on your chance of getting murdered at home. Statisticians call this the significance of the study. By arbitrary convention, most journals will accept publications if the significance is less than 5%. That is if there is less than one chance in 20 years on searches random variation. The authors compute a 4% significance for the 174 murder victims with guns, but this calculation assumes all data is measured without error, that their sampling is perfectly random, and that all important control variables were included. Those assumptions are never met exactly in any real study, and this study is very far from ideal. A rough rule of thumb used by many practicing statisticians is to be taken seriously and observational after the fact study like this needs to find a 3 to 1 effect. That is 3 times as many gun owners murdered at their homes as similar non-gun owners, regardless of idealized statistical calculation. Brown says that when studying topics like gun violence that are fully determined by human behavior, it's necessary to take a multidisciplinary approach, a requirement that may be lost on public health scientists trained to study disease. You fight an epidemic by trying to eradicate a pathogen, so the only solution to a gun violence epidemic is to get rid of guns. But violence is a more complex social problem than disease. Violence can do good as well as harm, and guns can protect as well as kill. In an epidemic, all deaths are equal and all are bad. But some gun deaths are good. The result of legitimate self-defense or law enforcement, and some are worse than others. Innocent bystanders, for example, versus suicide or armed gang members trying to kill each other. Gun control policies should be informed by criminologists, sociologists, psychologists, political scientists, legal experts, historians, and others. All of these fields have built up extensive knowledge about guns and violence. Yet this study, and most public health gun control research, cites only other public health researchers. That NRA strategy of hyperpolarizing the people worked out well for them, but it had devastating consequences in terms of continuing gun deaths. The Times video asserts that not enough research has been done on the causes of gun violence, and that Dickey is partly to blame. The real issue, Brown says, is quality, not quantity. The Dickey Amendment did not stop gun control research. The Rand Corporation tabulated 28,000 recent papers on gun control proposals. The problem is quality. Only 123 of those papers meant minimal statistical standards. He said, isn't it true that you really want to get rid of all guns? And I felt, how dare him? This guy is really bad. Worse than not understanding, he doesn't care. Jay Dickey and I, at that point, were really kind of mortal enemies. The film is structured around the story of an improbable friendship. In the 1990s, Congressman Dickey and Mark Rosenberg, then director of the CDC's National Center for Injury Prevention and Control, went from adversaries to best friends. We started talking, and we actually started to trust each other, and we started to like each other. And over the years, that developed into an incredible friendship. Rosenberg convinces him to change his mind and repent. He is the former self-described NRA point person and now is a strong supporter of research. I think the CDC should do the gun research that relates to gun violence. You're the guy. It's your bill. I feel different now than I did before. Dickey, who died of Parkinson's in 2017, may have come to regret pushing for the amendment that bears his name. But he remained concerned about the 1993 study, which would have complicated the New York Times narrative in which the filmmakers left out. The New York Times film uses this clip. What changed? How did you have the change of heart? Well, I just think it's just the weight of all of the incidents that have occurred, and that is that the kids and the innocent people who are being killed deserve our attention. What the filmmakers cut out was Dickey's ongoing concerns about the 1993 study and how it was carried out. We wanted research done for gun violence, and that's what the money was paid for. But we found out that as we went along that not only was the research being done just to support gun control, but we weren't even given access to what the collected data was. So it was clear that we needed to do something and to stop what was being done. That's exactly correct. Research funded by the CDC on gun violence is unscientific from the start. Guns are not a pathogen, and violence is not a disease to be controlled. What is needed is not propagandistic films or sloppy research. It's cross-disciplinary research on what drives human beings to violence, whether with a gun or not. Everything we have seen on this issue from the CDC indicates to me that they are the wrong organization to manage this research.