Statistics Don’t Matter (But Ours Are Right)

Some time back I had a productive exchange with a commenter (Scott) on my blog post about the public health approach to gun violence. Scott and I got into the question of the role of “data” or “statistics” in debates over guns and gun violence. He usefully summarized both the pro-gun and anti-gun approaches to “facts”:

“For my side, for me a least, the facts don’t matter because whatever they are they wouldn’t trump my freedom and rights. . . . But the other side, for a great many anyway, the facts don’t matter because at all costs they must get my gun away from me (or at the very least burden me with all sorts of “common sense regulations” that they foolishly believe will actually affect the criminal or accomplish the goals of safety etc).”

I was reminded of this exchange while reading Alan Korwin’s “Gun Rights” column in the July/August 2014 edition of American Handgunner magazine, unhelpfully titled, “Can Logic Sway Gun-Grabbers? Can Pigs Fly?”

Photo by Gage Skidmore
Photo of Alan Korwin by Gage Skidmore


Korwin is a full-time freelance writer who is best-known for his books on gun laws. Of statistics he writes in American Handgunner:

“It seems like the numbers should be the be-all and end-all, and should settle matters but they don’t. . . . Human rights are not subject to cost-benefit analysis. Statistics aren’t a proper gauge in this arena . . . [Gun rights are] based upon civil rights and your unalienable right to survive, not on fungible figures manipulated to make points and support agendas. In this sense, it doesn’t matter what the numbers are.”

So, statistics don’t matter because the issue of rights, as my commenter Scott said, cannot be trumped by any statistics — no matter how accurate those statistics are. I think this is an important point. And it reminds me of an interview conducted by a team of sociologists for the book, The Good Society. An economist at the Environmental Protection Agency was asked, “What about the theory that human life is priceless?” So caught up in cost-benefit analysis that he could not see the bigger picture, the economist answered, “We have no data to support that” (p. 117).

Korwin continues offering another reason statistics don’t matter, one that I find more troubling: “For some fun, put two statisticians in a room and let them have at it. There are numbers to support every position, making statistical arguments — especially on wedge issues — almost worthless. Just find out who supplied the numbers, and you’ll already know what the numbers will prove. . . . People aren’t looking for truth; they want to win. Anti-rights people aren’t interested in learning the truth about guns — they don’t care about your numbers, or whether they’re right or wrong. The evidence could be overwhelmingly on your side (and it is). They just hate and fear guns and want them to go away. . . Statistical approaches are wasted on them. If they cling to any numbers it’s only those to bolster their fears and help them rationalize their desire to rid the world of these horrible things . . . Surveys and stats are a subterfuge — it’s all about accumulating power against freedom. When you read or hear about news surveys and stats, especially from so-called “news” media and politicians, realize this: You’re not getting facts, you’re getting pitched. Those numbers have been developed by someone with an agenda, paid for by someone seeking a result and promoted by someone who decided the numbers work for them.”

Like many, I have seen evidence for this recently in the case of the Everytown “data” on “school shootings.” But Korwin’s logic suggests that this mis-use of statistics cuts both ways. If we put a Brady Campaign statistician and an NRA statistician in a room, what would happen? As Korwin says, “The Brady’s do not promote numbers to support the NRA, and vice versa” (emphasis added).

In a recent email exchange, someone else summarized the situation even more succinctly:

“The pro-gun folks don’t want any laws and the gun control folks don’t want any guns.”

For the truly committed on both sides, statistics are just a tool toward these ends.


Here I begin to part company with Korwin’s analysis. Yes, statistics don’t matter in terms of fundamental rights. And, yes, statistics don’t matter when it comes to political advocates debating contentious issues. But Korwin wants to have his cake and eat it too when he argues that statistics don’t matter, but ours are right. Having argued that if you find out who is supplying the numbers then you will already know what the numbers will “prove,” Korwin also argues:

“Prof. John Lott did amazing statistical work, and settled some of these issues once and for all. His studies turned out to be accurate and above reproach.”

Now, I just heard John Lott speak at the 20th anniversary dinner of Grass Roots North Carolina a couple of days ago, and he did a masterful job of debunking statements about guns made by Barak Obama in the wake of the Reynolds High School shooting in Oregon. But to debunk your opponent’s data does not make your data definitive.


It is simply not the case that we can say, “John Lott published More Guns, Less Crime, so the empirical debate is settled.” Lott’s work is significant, but it is also part of an ongoing and unsettled debate about the relationship between guns and crime (about which I have commented here and here). One that, honestly, given the complexity of the world and the fact that human beings have free will, will probably never be settled once and for all.

Despite this overreaching assertion, Korwin does offer a ray of hope for someone like me: an empirically minded social scientist interested in the role of guns in American society. Korwin writes:

“The right to keep and bear arms without infringement isn’t due to, and shouldn’t be influenced by, statistics on crime, accidents, armed households, ammo sales, caliber performance, incarcerations, self-defense incidents, numbers of permitees . . . anything. There are no valid statistics on the core issue of whether you can morally and legitimately protect yourself or others from crime or death. If more or fewer gun owners, carrying this or that, have some measureable effect on something — your rights are immune to it.”

Fair enough. But then Korwin concludes, “Your behavior might change, but not your rights.” Here statistics may yet matter. Crime is not good. Victimization is not good. Injury is not good. Understanding more about what makes crime more or less likely, what makes victimization more or less likely, what makes injury more or less likely, these are important things to know. This knowledge could then influence people’s behavior for the good, while leaving their rights alone. The NRA’s Eddie Eagle GunSafe program and the National Shooting Sports Foundation’s more recent “Project Child Safe” comes to mind as examples of a pro-gun rights behavior-modification effort based on the empirical reality that guns that are left unsecured pose a danger to children.





  1. Are there some particular statistics that you believe are accurate and should have sufficient logical weight to sway public opinion towards much more restrictive laws in regards to the possession of firearms? To have “sufficient logical weight” I would think it would have to overwhelm the amount of real good done with firearms to protect lives and property. The “if it will only save one life” is not a logical argument if it ignores lives saved.


    • Thanks for the comment. I hope I did not imply that there currently exist statistics that are sufficiently accurate, in either direction. Thoughtful and helpful work about behavior (not fundamental rights) should absolutely include consideration of the benefits of gun ownership and use. It should also distinguish between ownership and use of guns by criminals and by the law-abiding (I wrote about this earlier:

      These things I hope will enhance our understanding more about what makes crime more or less likely, what makes victimization more or less likely, what makes injury more or less likely. If we can come to a solid understand of that, then we can begin to think about ways of encouraging good behaviors and discouraging bad ones, while leaving the issue of rights aside.

      Beyond this, an overarching concern of mine is people invoking statistics on both sides in a way that only serves to undermine the credibility of empirical research as a whole. That is one of the most depressing parts of getting involved in this line of work, for me.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.