How to Think About Guns (Ep. 114)

Listen now:

There are an estimated 300 million guns in America. This photo, from Kyle Cassidy’s Armed America, shows (from right to left) Donno, Judy, and their son Uzi. Cassidy’s latest book is War Paint: Tattoo Culture & the Armed Forces

Our latest Freakonomics Radio podcast is called “How to Think About Guns.” (You can download/subscribe at iTunes, get the RSS feed, or listen via the media player above. You can also read the transcript here; it includes credits for the music you’ll hear in the episode.)

This episode is a straightforward conversation between Stephen Dubner and Steve Levitt, keeping in mind recent events like the Newtown, Ct., school massacre and long-standing traditions like the American embrace of guns.

Levitt has focused much of his academic career on crime research, including all sorts of gun policies that do and do not prevent violence.  He has also analyzed the relationship between the economy and the crime rate, whether increased police presence affects crime, and whether deterrents like capital punishment and sentence enhancements actually work.

We begin this episode with some basic data. In the U.S., there are roughly 11,000 gun homicides and 20,000 gun suicides a year. (Our podcast “The Suicide Paradox” looked into why we hear so much less about the suicides than the homicides.) What we hear about more than anything are the relatively rare but extremely disturbing mass shootings. From the podcast:

Mother Jones magazine recently built a database of mass shootings – four or more fatalities — over the past 30 years. Not everyone likes this database – it excludes, for instance, all gang shootings and armed robberies. But here are those numbers: since 1982, there have been 62 mass shootings with 513 fatalities, or an average of 2 mass shootings and 16.5 fatalities a year. (Now, remember, keep in mind there are 11,000 gun murders each year in total.) Over just the past 10 years, those numbers are a bit higher – about 3 shootings a year, with 26 fatalities. But 2012 was a very bad year: 7 shootings with 72 fatalities, more than 4 times the average number of victims in a year from mass shootings.

Levitt helps put all these numbers in context, and make sense of overall crime trends. We also hear what he thinks about current proposed gun policy. He’s not optimistic:

LEVITT: I would just say that anyone with any sense looks at the current political climate, thinks about the kinds of proposals that are being made and accepts the fact that none of these proposals are going to have any real impact at all.

So what could diminish gun violence? We’ve asked that question before; good answers are hard to come by.  Levitt says mandatory sentence enhancements work. You’ll also hear about Geoffrey Canada‘s book Fist Stick Knife Gun, which might change the way you think about violence in general.

caleb b

Guns are not constitutionally protected for self-defense, hunting, or recreation. They represent the means in which citizens can fight tyranny. Any law governing such should begin with this premise.


I find it insincere to attempt to have an "honest" discussion about gun violence and use a fraudulent database that intentionally provides bogus data in an attempt to push gun control efforts.

I am sure we could create a database of shootings that have been stopped by armed civilian (and off duty police) carrying weapons, and claim there is no problem - but it wouldn't change the fact that mass shooting exist.

I have no problem discussing mass shootings and the peoples rights to bear arms, and I can understand some people no liking guns or the second amendment. The fact that the government doesn't "give" you rights is often forgotten, the governments responsibility is to PROTECT our rights.


Why can't we start with a tax on guns at the manufacturer or distributor level?

It's obvious that guns have tremendous externalities associated with them. The injuries from gun violence not counting the deaths costs billions of dollars per year for treatment alone. Most of that is paid by taxpayers. Aside from that, you also have associated loss of work time, and potential partial or full disability (which costs the taxpayers even more).

Guns also increase the costs of police forces in various ways. Gun deaths are likely to account for an astounding amount of lost productivity which can be quantified as well.

We are unlikely to be able to completely eliminate gun violence, and it may be cold but how about we start by putting a number on the cost of gun violence, and forcing the gun industry (and those who buy guns) to shoulder that burden.

Why not start with a pigovian tax? It may reduce gun violence. It may not reduce gun violence. But at least it is unlikely to increase gun violence, and at least non-gun owners wont end up subsidizing the 'hobby' and 'self defense right' which generates significant negative externalities nationwide.


tung bo

The Canada book points to another historical development and possible avenue for mitigating urban violence. As Canada described, the urban culture of 20-30 years ago were of LOWER lethality because few people had access to guns. That they are common today is most likely due to cheap mass produced hand guns. (If any one has data to contradict this, please correct me.) This suggest an economic solution: tax the sales of cheap hand guns at the factory level and also maintain a correspondingly high buy back price using this tax. The point is to elevate the street price of hand guns above the cost/demand equilibrium and hopefully achieve a fewer number of guns in circulation on the street. If we can approach the lethality of 20-30 years ago, that might be a big achievement already.


What about rasing the price of bullets? Through taxing?


Mat, your passionate diatribe besides being impolite and venomous lacks any semblance of an attempt at a solution. It's ok to disagree, but if you want people to consider your discourse worthwhile please temper your distaste by offering what you perceive as a solution to the problem at hand. To just hammer someone for their opinion and run away comes across at best as childish, at worst unstable. If you want to be taken seriously show some restraint and add something besides insulting language to the discussion.


When thinking about homicide we rarely make allowance for the fact that not all homicides are created equally. With regard to crime we tend only to look at the method and result. To have any indication of how to solve the problem I think we need to look a bit deeper, at both motive and incentive.

Much of the street crime in the US is drug-related and this is where much of the homicides occur as well. When you think about it, it makes sense. There’s huge consumer demand for illegal drugs. This demand means huge potential for profit and therefore a significant, completely rational incentive to get involved in illegal trade. Crime pays (or at least makes a convincing promise to pay).

A criminal involved in the drug trade for example has plenty of incentive to arm up and use violence. Drug gangs need to defend themselves against what’s usually the biggest armed gang in the city, the city police, AKA the drug warriors.

They need to defend their market-share against encroaching business competition. In their line of business, there’s no such thing as suing an infringing competitor or an economic sanction or a leveraged buy-out. Violence is the one and only recourse in illegal business for solving what are common and mundane facts of business in the legit world.

Drug entrepreneurs also don't have banking services to safely store their assets or fixed alarmed and secured warehousing services to safely store their product, or insurance to help mitigate the cost of risk. So massive piles of cash and product represent huge carrots for anyone with a stick to try to score, and you can't insure the carrot against theft. In this arrangement violence is the method for stealing these carrots, as well as protecting them from costly loss.

Then when you compare the US to other somewhat similar rich countries like the UK, we have many more metropolitan areas of 250,000 people or more. This density seems to be a bifurcation point where violence experiences a huge trend upward. I wonder about this, is this the nature of humanity in dense groups showing through, or is it simply that once population in an area reaches around this number it forms a reasonably robust market for illegal services and making it a compelling landscape for setting up business? Not just illegal industries, but all industries seem to flourish in large population centers, and likely for shared reasons. If the violence is a manifestation of the only available business techniques in an illegal market then it is of no surprise that it is found in the same places. On top of the robust consumer base, the fact that these densely populated areas are the nodes in the transportation network across the whole country and provide anonymity in the crowd, make it very logical to conduct illegal business in these areas.

To help solve this problem customers of illegal markets could willingly boycott the products and services they presently enjoy, thereby starving these violence machines of their life-blood. If that is not a sacrifice we want to make as a society we should consider letting these markets exist in a legal framework so violence is no longer such an advantageous and necessary business strategy.



I was surprised that there was no discussion about the situation here in Australia. It seems there is some great data here – a clear ‘before and after’ situation, where certain bans were introduced together with an amnesty period and buyback for older weapons (paid for by a temporary tax or levy) before all guns that fit certain criteria, whether existing and out there, or yet to be manufactured, were banned in certain contexts. It was also made much harder to get a gun too. It takes months to buy a gun here, and most people don't even know anyone who owns a gun, let alone how to go about getting one. Ostensibly, it all seems to have had a dramatic impact on gun violence, statistically at least, but I’m aware of course that there is always a hidden side to everything. The existence of such an overt ‘experiment’ climate though, with a measureable data-set, seems like a missed opportunity if not covered when discussing what can work and what won’t.