What is the social cost of gun ownership?

by on December 27, 2012 at 2:41 am in Economics, Law | Permalink

Via @JustinWolfers, finally we can see some numbers, provided by Philip J. Cook and Jens Ludwig:

This paper provides new estimates of the effect of household gun prevalence on homicide rates, and infers the marginal external cost of handgun ownership. The estimates utilize a superior proxy for gun prevalence, the percentage of suicides committed with a gun, which we validate. Using county- and state-level panels for 20 years, we estimate the elasticity of homicide with respect to gun prevalence as between +0.1 and + 0.3. All of the effect of gun prevalence is on gun homicide rates. Under certain reasonable assumptions, the average annual marginal social cost of household gun ownership is in the range $100 to $1800.

By no means should you take this as the last word, but it is one place to start.

1 KPres December 27, 2012 at 4:09 am

And what about the social benefits? What about the homicides prevented, the criminals who hold back because of the possibility of an armed citizen? Did they include any of that in their estimates?

2 Jan December 27, 2012 at 4:28 am

Nice try, Wayne. Why don’t you just look at the paper and see?

Obviously 310 million murders per day, every day are prevented due to the possibility of an armed citizen.

3 Andrew' December 27, 2012 at 5:25 am

Read it. They don’t consider anything.

4 Rahul December 27, 2012 at 7:17 am

I admit I only skimmed through the paper; but why is the range so wide? “$100 to $1800” sounds like a huge amount of uncertainty.

5 Andrew' December 27, 2012 at 9:07 am

“There is a literature on everything” and this is how it happens. That’s not a criticism.

6 Mr Economist December 27, 2012 at 6:58 am

The marginal social benefit will be a helluva lot lower than in most other western countries where they are not idiotic enough to have such crass gun laws perpetuated by a society of morons.

7 careless January 3, 2013 at 9:07 pm

This has to be my favorite: the guy who wants to ban guns simultaneously arguing that more guns in places that banned them would be better.

8 GW74 December 27, 2012 at 7:05 am

230,000 legal guns are stolen every year in the USA, mostly from houses with unsecured non-sporting guns used for home defence. Stick that in your “estimate” of homicides “prevented”.

9 T. Shaw December 27, 2012 at 11:04 am

Are all academic economists innumerate?

10 Willitts December 27, 2012 at 12:08 pm

With about 15,000 murders each year, 215,000 gun thieves are lousy shots.

Stick that up your “estimate” of homicides committed with stolen guns.

11 T. Shaw December 27, 2012 at 12:44 pm

In 2010, there were 11,015 American gun homicides. That would be one out of 218, or 0.46%, of 2,400,000 American deaths. Approximately, 25% of US deaths were the result of heart disease.

Hostess recently went bankrupt.

Thank God! Twinkies kill!!

12 Tom B December 27, 2012 at 2:14 pm

Your statistics are the average. It’s actually been declining since 2005. The BJS reports it was 233,000 in 2005 and 135,000 in 2010. Maybe people are becoming more aware of the theft issue and taking better care. Certainly I always have.

Another interesting statistic that you should be aware of is that the ATF says that between 10% to 15% of stolen guns are used on crimes (that’s all crimes). Most firearms used in crimes were purchased illegally from people, either through straw purchases or from unscupulous licensed dealers (representing 8% of the total licensed dealers in the US). The ATF believes that these are the two categories that need to be targeted to solve the problem.

The ATF didn’t indicate the number of those stolen firearms that were used in murders however, just in total crime statistics. I’ve seen some numbers out there that say it was as high as 90%, but in that particular blog (not exactly a reliable source) the author failed to indicate or provide a link to the source of that information, so I suspect it was anecdotal at best. In fact, all of the sources I found when researching this showed numbers that were all over the place, and all of them based on speculation and hearsay rather than on an legitimate study. If you can find some numbers to back up a case that stolen firearms are the biggest source of murder weapons I would be interested, so long as you don’t quote sources from any organization, pro or con, that has a axe to grind, like the Brady bunch or the NRA or any other agenda based group, all of which manipulate the data to justify their own agendas.

13 Paul December 27, 2012 at 7:35 am

It’s covered. If guns were a deterrent to crime, they would find negative estimates, and would be talking about the marginal social benefit of gun ownership. Alas, it would appear there is none.

14 Robert December 27, 2012 at 10:03 am

Paul, I’m new and learning, so please correct me if I’m wrong. If the MSC were positive, but lower than the Marginal Private Cost, would that imply an external benefit since the Marginal External Cost would have to be negative? It would still be a social cost, but one carried by the producer, correct?

15 willitts December 27, 2012 at 12:13 pm

Marginal social costs are added to the supply curve. Marginal social benefits are added to the demand curve. An estimate of one tells you nothing about the other.

16 dead serious December 27, 2012 at 7:49 am

I stock torture instruments for when I catch me an armed burglar. Is THAT included in their estimates? Is it? IS IT?

17 T. Shaw December 27, 2012 at 12:47 pm

I’m kind’a with you.

I have my wife sleep on the side of the bed closest to the door.

18 Jeff Abbott December 27, 2012 at 10:11 am

The study does account for that. Homicide rates increase with gun ownership. Nothing is being filtered, nothing left unaccounted for. However, the method used to calculate total gun ownership is not entirely accurate. They cite one case where voting exit polls were used as a proxy (estimate) for ownership rates. Another uses subscriptions to Guns and Ammo magazine. The rates of homicides and crime are the actual rates. What the study does not do is determine the rate deterrence as compared to rate of homicide. Meaning are those killed more likely to be gun owners or non-owners in the same region (count/state)? If non-owners, then the gun ownership would tend to increase homicide rates for non-owners relative to gun owners.

19 Joshua Lyle December 27, 2012 at 12:43 pm

“Homicide rates increase with gun ownership.”

Yes, in the sense that is identical with “gun ownership increases with homicide rates”.

20 dead serious December 27, 2012 at 5:48 pm

Yes, the causation arrow is mighty confusing.

If you don’t believe that an increase in gun ownership increases homicide rates, you’d have to believe that a society of knives and baseball bats would suffer the same kill rates.

On the other hand, if you believe the causation arrow points the other direction, you’re basically saying that we’re caught in an endless cycle of escalation. Something no other nation on the planet seems to be facing.

My gut says it’s the former.

21 Correlation December 30, 2012 at 6:45 pm

It doesn’t imply escalation at all. Joshua’s point is that, assuming for the sake of argument that the study is accurate, all it actually shows is that homicide rate and gun ownership have a slight positive correlation. It’s quite possible that living in a high-crime area make people more likely to keep a gun in their household. (Inner cities are the dominant source of homicide in America, so specifically it means that living in an inner city makes you more likely to own a gun.) There’s no “endless cycle” at all though. Crime rates are decreasing in America as well as worldwide.

22 Alexei Sadeski December 27, 2012 at 4:12 am

Average annual marginal social cost of legal handgun owned by a responsible owner?

Probably a lot lower.

Average annual marginal social cost of each ILLEGAL handgun?

Probably a lot higher.

Calculating the social cost of a marginal firearm isn’t the same as calculating cost of, say, CO2 emissions. Not every marginal firearm’s social cost is equal, but every pound of CO2 emissions is equal (well, more or less – supposedly CO2 emitted at altitude is worse than that emitted at sea level, but these differences are easily discernable to the observer).

This study does not differentiate between firearms legally obtained and firearms illegally obtained (study’s use of FSS data as proxy for household firearm ownership rates). It is possible – nay, near metaphysical certitude – that legally obtained firearms have far lower average marginal social cost than do illegally obtained firearms. Given this, should not the social costs of these two category of firearms be measured separately?

After all, the policy question concerns legal firearms, not illegal firearms.

23 prior_approval December 27, 2012 at 4:54 am

‘Average annual marginal social cost of legal handgun owned by a responsible owner?

Probably a lot lower.’

Well, apart from accidents, suicides, thefts, and situations where a legal owner simply decides to kill people – Joe Horn, in 2007 comes to mind, or better, this case –

‘Trevor Dooley stood his ground, brandished his gun and killed a man after an argument over local skateboarding rules in a Florida town.

He argued in court last month that he had a right to do so under the state’s Stand Your Ground law.

Outrage over the death of 17-year-old Trayvon Martin, killed by a crime watch volunteer, has focused new attention on the law, which permits those in Florida “to meet force with force, including deadly force” when attacked. As my colleague Lizette Alvarez reports, the Justice Department is pursuing an investigation into Trayvon’s case.

As that investigation goes forward, the law is currently being invoked as a key defense by Mr. Dooley.

The man he killed, David James, had been playing basketball with his 8-year-old daughter in September 2010 when he and Mr. Dooley began arguing over whether a boy on a skateboard had a right to ride on the court, according to an account in The St. Petersburg Times. There was a “physical confrontation,” the police said, during which Mr. Dooley fired the weapon he was carrying, killing Mr. James in front of his daughter.

“You agree you do not want to go to prison for killing David James?” he was asked at the trial, according to televised footage from the courtroom.

“I don’t think I should,” responded Mr. Dooley, who has been charged with manslaughter but says he feared for his life during the altercation with Mr. James.’


Legal gun owners are perfectly capable of killing a father in front of his child.

24 Andrew' December 27, 2012 at 6:30 am

Killing a father that is capable of getting himself shot in front his child, that is.

25 prior_approval December 27, 2012 at 6:52 am

‘Killing a father that is capable of getting himself shot in front his child, that is.’

I’m curious – is there any confusion about who the gun owner was, or who the gun’s pulled the trigger?

Because that quoted formulation would seem to equally apply to the murdered children in a Connecticut school, who were also capable of getting themselves shot in their classroom.

26 JVM December 28, 2012 at 8:00 pm

> capable of getting himself shot

I guess everybody besides superman?

27 David K December 27, 2012 at 5:33 am

My issue with arguments like this is “owned by a responsible owner.”

We don’t have firearms laws that only permit legal gun ownership by RESPONSIBLE people. We permit legal gun ownership by virtually anyone who has not (yet) committed a felony. And at gun shows or online we don’t even check that.

There’s a two-step involved in many pro-gun arguments. First, if guns are owned by law-abiding responsible people who would only use them for defense, then that’s ok (and even a positive good). But, second, we cannot have waiting periods or require background checks for private sales or have intrusive checks into people’s fitness to carry weapons or even have a database of who is armed because this would make it too annoying/difficult for said law-abiding/responsible people to acquire said weapons.

The problem is that the result of the second argument negates the antecedent of the first. I wouldn’t have a problem with Canadian-style checks and widespread ownership, though living currently in the UK, where even the vast majority of the police aren’t armed, I also see that legal gun ownership isn’t exactly a prerequisite for a non-tyranny or for crime prevention. But the US has a system where those who are NOT responsible and should NOT have guns are nonetheless allowed legally to acquire guns– and that’s what is crazy.

28 Mike Hunter December 27, 2012 at 9:35 am

You are incorrect. We have a required waiting period for those who buy guns.

Also the right to keep and bear arms is constitutionally enshired in America. As a U.S. citizen the government can’t prevent you from exercising your civil rights; as long as you are a sane and law abiding citizen, just because a subset of people think you’re irresponsible.

When you go to cast a ballot, protest, or write a blog no one can prevent you from doing those things because they disagree with the way you live your life. The same legal principals apply to all of the civil rights enshrined in our constitution.

You may disagree with that, but thats the system we as a nation have decided upon. If enough Americans think the constitution should be changed then there is a process to do that.

29 Mark December 27, 2012 at 2:25 pm

The first amendment is not unlimited: defamation (libel/slander, with actual malice when referring to public persons), obscenity (criteria based on prurient interest and lacking social value; child porn), and some forms of commercial speech (“This snake oil cures cancer!”). The first amendment is also not unregulated: the government can set the time/manner/place of protests.

The second amendment is similarly subject to interpretation by the courts and regulation by Congress. You may own a gun, but the manner and conditions of that ownership can be regulated by government. In a republic, the citizenry elect representatives to negotiate the terms of those regulations. And recent events may shift the median voter’s definition of reasonable regulation.

30 John December 27, 2012 at 4:18 pm

I don’t think the median voter theorem really fits the existing landscape. We’ll have two coalitions within which a median voter position can exists and between these two nearly disjoint sets of positions a political decision will be made. I think the suggestion that the outcome reflects some median social position highly suspect.

If things worked as suggested by the Median Voter Theorem as voter turn out declined a corresponding decline in public satisfaction with politicians and government action should not be present.

31 Urso December 27, 2012 at 11:15 am

It’s also a brilliant bit of victory by definition — responsible gun owners can *never* commit any unjustified homicides or crimes of violence, because once you do that, you’re not a “responsible” gun owner.

Put otherwise, everyone is a responsible gun owner until all of a sudden he isn’t.

32 Alexei Sadeski December 27, 2012 at 11:42 am

No it’s not.

Responsible simply means people who own guns legally, and if carrying their weapon, that they are carrying with proper license.

Such responsible gun owners will commit crimes from time to time. They will also prevent crimes from time to time. Restricting our sample to only legal owners and carriers will likely change the social cost of gun ownership into a net benefit.

33 Urso December 27, 2012 at 1:29 pm

“Restricting our sample to only legal owners and carriers will likely change the social cost of gun ownership into a net benefit.”

Right. Hence, victory by definition.

34 Tom B December 27, 2012 at 2:31 pm

By that logic, should we therefore anticipate when a crime will be commited and arrest the soon to be guilty party so that the crime doesn’t get committed? Kinda like Minority Report, right?

Truth is though, that most gun “legal” owners don’t commit crimes (with their firearms, I should qualify that), it’s been estimated that the number falls into the less than 1% range. Even Obama in his speech last week correctly acknowledged that, and he’s certainly no fan of firearms ownership so you have to admit that there’s a great deal of validity in that.

35 GiT December 27, 2012 at 5:50 am

The vast majority (80-90%+) of illegal guns come from people who legally purchased them and then illegally sold them. Of course, such gun traffickers make up a small proportion of all legal owners (I’ve seen stats that say about 3% of owners, or 8% of dealers). So the bulk of illegal gun use traces straight back to… legal gun owners. Stolen guns only account for 10-15% of illegally held guns.

I don’t know of data on the incidence of various gun-events conditional on whether or not the person in possession of it was within or outside the bounds of the law, but I doubt suicides and accidental death or injury occur primarily with illegally owned weapons, and I’d guess that a certain subset of criminal activity (domestic abuse, argument escalation, &etc) are not hugely different in frequency between legal v. illegal owners, though others (robberies, crime-related murders, &etc) most likely are.

36 Mike Hunter December 27, 2012 at 9:38 am

Just curious where did you get your stats? Could I have a link?

37 GiT December 27, 2012 at 11:12 am

I feel like I’ve seen them pop in a few different places, but here’s the latest place I saw them:


38 Willitts December 29, 2012 at 5:20 pm

Yes, CO2, if harmful, is a public bad. Bullets are a private bad for the victim.

Why can’t people understand this distinction?

39 prior_approval December 27, 2012 at 4:44 am

Putting a price on needless death.

The culture that is America.

40 Morten December 27, 2012 at 5:57 am

Rather: Putting a price on human life. Economists.
(Which IMO is necessary and useful to evaluate e.g. traffic regulation, health care etc. where life-and-death decisions should be made on a rational basis.)

41 Brian Donohue December 27, 2012 at 7:22 am

Yes. Dismal, but there it is.

42 JVA December 27, 2012 at 7:47 am

Now we only have to set up a market for these externalities and we will have optimal allocation of gun violence!

43 dead serious December 27, 2012 at 7:56 am

Tax the living hell out of gun manufacturers and imports. Increase jail time for illegal gun possession.

44 Elizabeth December 27, 2012 at 11:26 am

Shall we also “tax the living hell” out of vehicle manufacturers? Or perhaps the manufacturers of cleaning products?

45 msgkings December 27, 2012 at 11:48 am

This bullshit again. Guns, if used correctly, result in severe injury or death. Cars, cleaning products, or any other childish product you can think of, do not.

46 Elizabeth December 27, 2012 at 11:52 am

This fecal matter does not fit your thesis, so therefore is obviously irrelevant? Your statement somehow prevents the misuse of common household items, which have caused more deaths than firearms?

47 msgkings December 27, 2012 at 12:01 pm

No, this fecal matter is bullshit because it’s bullshit, not because it doesn’t fit ‘my thesis’.

If you can’t see the existential difference between an item designed to kill and items that might be dangerous when used incorrectly, then you are not a serious person and literally incapable of having a discussion with.

48 Willitts December 27, 2012 at 12:19 pm

Guns, if used correctly, kill people who need to be killed.

Guns used correctly also spend the majority of their lives completely silent which is a valuable lesson for you, msgkings.

49 msgkings December 27, 2012 at 12:25 pm

Fine, Willitts, I agree regarding the proper use of guns. Can you please tell me where I’m wrong to want to work towards a world where the only guns out there are hunting rifles and 6 shot handguns, and that those guns are much more carefully stored and tracked? Bascially to me the root of the CT shooting isn’t the crazy shooter (there will always be crazy people), nor the fact that we can own guns in America. It’s the idiot mom with the multiple automatic pistols with large magazines keeping them lying around in front of her crazy son who she knew was crazy. Can we agree to work towards fixing that part of it?

50 Terms December 27, 2012 at 1:24 pm


automatic, pistol?

An AR-15 is neither.

51 dead serious December 27, 2012 at 7:49 pm

The social benefits of vehicles and cleaning products greatly outweigh their negative externalities.

52 John December 27, 2012 at 4:28 pm

Which will likely have zero impact on things like what happened in Connecticut.

I heard something recently that said most of the gun deaths, and shootings, were actually bad-guy on bad-guy so not that many of the innocent are actually part of the statistics.

Anyone have those stats?

53 dead serious December 27, 2012 at 6:15 pm

I don’t agree. Gun ownership should be an expensive proposition. Not illegal, but costly.

Here are my proposals:

Tax gun manufacturers so end users pay through the nose for their fix.
Tax ammo manufacturers for the same reason.
Raise gun permit costs to $1,000/year.
Assault style weapons are banned outright. There is no reason in a sane society that anyone needs these.
If you’re caught in possession without adequate permit(s), for a first offense you do 1 month jail time and your driver’s license is revoked for a year. With subsequent offenses, the penalties get more serious.

So you gun nuts get to keep your guns but you pay for the externalities. Just like a tobacco user.

54 roystgnr December 27, 2012 at 9:46 am

Leaving the house (whether in a car, as a pedestrian, etc) entails a statistically significant risk to your life. You take this risk for economically measurable reasons all the time. If you aren’t rational enough to realize what you’re doing, but Americans are, this is not an indictment of America.

55 msgkings December 27, 2012 at 11:44 am

But the statistical significance of the danger of leaving your house is a lot higher in a heavily armed America.

The problem with this debate is the lack of nuance and middle-ground compromise (which I guess is kind of reflecting the entire culture these days). One side is “guns are awesome, gotta protect my family, outlaw guns and only outlaws have guns, etc…” and the other side is “all guns are going to eventually murder someone, all gun owners are nuts, ban everything”.

I’d like to ask the gun-rights supporters here if they have a problem with some simple additional steps that we could take to make shootings like CT far less likely, and far less deadly, including more lengthy background checks, limits on ammo magazine sizes, limits on automatic handguns, requiring 2 adults to vouch for all gun purchasers (Canadian idea), ammo ID that can trace bullets to the guns that fire them, very heavy penalties (both civil and criminal) for crimes committed with your gun even if it’s stolen, etc.

These measures won’t eliminate gun violence, not even a crazy person shooting up a school, but you have to think they would reduce those incidents and make them less deadly. If the CT shooter snapped under these new rules, his mother would probably have secured her guns better (fear of penalties) and even if she didn’t his gun would only have been able to fire say 6 times before he’d have to reload.

Do gun rights proponents have a problem if we work towards having only hunting rifles and 6 shot handguns available? The handgun protects your family, the hunting rifle is for sport. Seriously asking this.

56 Willitts December 27, 2012 at 12:26 pm

The Second Amendment isn’t about hunting, sport shooting, or home defense. The fact that you don’t get this is the reason there is no middle ground in this debate.

Today you would agree to hunting rifles and six shot revolvers. Tomorrow when someone is killed with a hunting rifle or six shot revolver, you’ll come to take those too.

The bottom line is that people like you can’t be trusted in an honest debate about firearms and public policy on firearms. You know you can trust me with my guns because you’re still breathing.

57 msgkings December 27, 2012 at 12:45 pm

Wow, Willitts, didn’t pick you for an internet tough guy. Disappointing really. But thanks for not shooting me!

So much fail here. In a world where the black government helicopters are always one hair’s breadth away from crushing our liberty and only the 2nd is there to protect us, we can’t possibly hope to change the role of guns in this nation.

I guess maybe this is one of those debates like abortion where there’s no middle ground after all. Also very disappointing.

58 Alan December 27, 2012 at 12:39 pm

“I’d like to ask the gun-rights supporters here if they have a problem with some simple additional steps that we could take to make shootings like CT far less likely, and far less deadly,”

Shootings like the one in CT already are astronomically unlikely. All “solutions” to violence are sold as “Will decrease violence” but rarely, if ever do. The “gun free zone” was promised to “stop violence” and it didn’t. The banning of guns in England and Australia were supposed to stop violence and didn’t.

All gun control proposals assume results that they can’t deliver.

59 msgkings December 27, 2012 at 12:48 pm

Not looking for a ‘solution to violence’ nor a ban on guns. Just want some common sense improvements in the culture.

And Australia’s recent gun restrictions HAS reduced gun violence dramatically in that country.

60 Ronald Brak December 27, 2012 at 7:20 pm

Crikey mate! How can you be so negligent when it comes to the facts of Australian gun violence? It’s not hard to look up, you know:


Tighter gun control laws accelerated the decline in Australian gun deaths. Also, there have been no mass shootings since the gun laws were made tighter. And also note that guns aren’t banned in Australia, they are just more strictly regulated. Are you going to take back what you said, or are you just going to claim that babies eaten by dingoes increased by the same amount gun deaths went down or something? In other words, or you going to admit you were wrong, or are you going to lie?

61 T. Shaw December 27, 2012 at 1:04 pm

See John Allen Paulos’ book, Innumeracy

Without a familiarity with the workings of large numbers, people can irrationally react to terrifying incidents, especially when propagandized by evil statisticians and politicians.

The author cites an example: fear of flying and terrorism. Airline terrorism deaths have been a media theme. Approximately 85,000,000 body cavity searches later . . .

Here is the math: in 1985, 17 Americans died in air terror. In that year, 28,000,000 Americans traveled by air. Ergo the chances of being killed by air terror were 1:1,600,000. Compare 1:1,600,000 to 1:5,300 killed by car crashes.

Evil journalists and (“evil” would be repetitive) politicians cry, “You are a mass murder waiting to massacre school children!”

In 2012, so far (what?) 50 were killed in assault rifle massacres. Your odds are: 50 in 310,000,000 or 1:6,200,000.

How do lying, vile statisticians and politicians get away with it?

Answer: Public schools consign most Americans to innumeracy: mathematical ignorance/illiteracy.

“But, but . . . if it happened to YOU the odds would be 1:1.”

Here is another symptom of innumeracy: the tendency to personalize (hint: it’s irrational and wrong). The only instances wherein personalization works are death and taxes: you are 1:1 lilely to die, and you can’t avoid taxes, either.

Innumeracy shows itself in pseudosciences like the economics and gun control superstitions.

Isaac Asimov wrote: “Inspect every piece of pseudoscience and you will find a security blanket, a thumb to suck, a skirt to hold. What have we to offer in exchange? Uncertainty! Insecurity!”

And, liberty!

Case in point: innumeracy and the pseudoscience behind assault rifle and gun bans, and, fully demonstrated herein, economics.

62 BintheCT December 27, 2012 at 2:07 pm


63 Claudia December 27, 2012 at 3:02 pm

Neato, I finally understand the conservative vice http://marginalrevolution.com/marginalrevolution/2006/08/the_conservativ.html If those libtards and economists (?) weren’t innumerate, we wouldn’t be having this silly argument over guns. Look numeracy is a serious issue, but it’s kinda second order here. I am sympathetic to your points, but not in all settings. Some deaths are more powerful, more horrible than others…that’s not wrong, it’s human. Social sciences like economics care about valuations not just head counts.

64 Jan December 27, 2012 at 4:57 am

The policy question is not solely legal firearms (versus illegal arms). All illegal firearms started out legal somewhere up the supply chain. Whether they were stolen from a truck en route to the weapons store, sold to an unscrupulous character by an otherwise law-abiding gun owner, purchased by a felon at a gun show (no background check), or taken from someone’s house by a burglar, all illegal firearms start off on the right side of the law.

The authors do find evidence that gun prevalence leads to elevated homicide rates through transfer of guns to from legal to illegal owners, rather than through increased gun misuse by otherwise legal owners.

65 Rahul December 27, 2012 at 7:02 am

One point I’d make is that as a society, I don’t think we care so much about the net social cost of guns but more about the particular sort that disturbs us more. i.e. Say there was some intervention that left total-gun-related-deaths the same (or even increased them) yet reduced the number of mass-shooters to zero I think society would accept this readily.

My point is, net-social-costs are *not* a metric most people use to judge gun policy. 100 people lost to a mass-shooter influence policy way more than a 100 killed in home-invasions, robberies or drive by shootings.

66 Mike Hunter December 27, 2012 at 9:42 am

This is very true. The number of homicides has actually fallen by 73% since 1992. So we have a lot LESS murders despite more lax gun control. Remember the assault weapons ban was introduced during the clinton era. In addition many more states offer concealed weapons permits now then in the 90’s.

67 David Wright December 27, 2012 at 5:01 am

After all the emotionalism and preening of the past week, it’s very nice to see someone with a data-driven, quantitative approach. And yes, I would say exactly the same thing if the elasticity had come out -0.2.

68 Andrew' December 27, 2012 at 5:13 am

“the formula indicates one additional homicide per year for every
10,000 additional handgun-owning households.”

69 Andrew' December 27, 2012 at 5:17 am

So, take out the suicides, the scumbags, and the government fails, and my end probably comes to $50. Where do I send the check? Not that my guns are actually the problem. So, why not let’s randomly internalize everything else starting with immigration. Then being black, etc.

70 prior_approval December 27, 2012 at 7:27 am

Well, here is a bit of information about those suicides –

‘Suicide is still the leading cause of firearm death in the U.S. In 2009, the U.S. firearm suicide total was 18,735 an 11% increase from the 2006 national gun suicide number of 16,883.’ – http://www.ichv.org/facts-about-suicide-and-gun-violence/

Maybe you could send your check somewhere useful, like a suicide prevention organization, unless that firearm death total leaves you as cold as their bodies.

71 Willitts December 27, 2012 at 12:29 pm

Yes, and the presumption is that the person with suicidal ideations wouldn’t commit suicide by the next most convenient means.

72 steve December 27, 2012 at 1:02 pm

“Yes, and the presumption is that the person with suicidal ideations wouldn’t commit suicide by the next most convenient means.”

The Israeli data supports the idea that suicides are often impulsive, and having a gun present increases the risk of a successful suicide happening. So, there will still be suicides absent guns, but with them we will have more.


73 Shanghaied December 27, 2012 at 2:53 pm

“The Israeli data supports the idea that suicides are often impulsive, and having a gun present increases the risk of a successful suicide happening. So, there will still be suicides absent guns, but with them we will have more.”

Looking at rates of suicide and gun ownership, I don’t gun ownership leading to a higher suicide rate. The US has 148x more guns per capita than Japan, yet Japan’s suicide rate is 2x higher. Clearly, it isn’t the presence of a gun increasing the risk of a successful suicide.

Other countries (South Korea, Belgium, Finland, China, Poland, France, Russia, to name a few) have substantially fewer guns per capita, yet have comparable or substantially higher suicide rates. If anything, more prevalent gun ownership appears to correlate with a lower suicide rate, not higher.

Maybe Andrew should send his check somewhere useful, like the NRA, unless suicide prevention leaves you as cold as their bodies.

Country breakdown (all data from Wikipedia; links at bottom)
Suicide rate (# suicides /100,000 people) vs gun ownership rates (#guns/100 people):
South Korea: 31.2 suicides vs 1.1 guns
Japan: 23.8 suicides vs 0.6 guns
China: 22.23 suicides vs 4.9 guns
Russia: 21.4 suicides vs 8.9 guns
Finland: 16.8 suicides vs 32 guns
Poland: 15.4 suicides vs 1.3 guns
France: 15 suicides vs 31.2 guns
Romania: 12 suicides vs 0.7 guns
United States: 12 suicides vs 88.8 guns
Ireland 11.8 suicides vs 8.6 guns
India: 10.5 suicides vs 4.2 guns
Singapore: 10.3 suicides vs 0.5 guns


74 Barkley Rosser December 27, 2012 at 5:53 pm

Japan has pro-suicide cultural values. Under certain circumstances it is considered honorable to commit hara-kiri aka seppuku, aka suicide, an old samurai tradition properly done by slicing one’s stomach open with a samurai sword (sharpest in the world) and holding one’s intestines while one honorably dies. Of course, today, few use that method. Last time I was in Japan, a Diet member who had voted against his faction chief’s orders hanged himself in his living room.
In most Christian sects, suicide is a sin. These facts explain the cross-national suicide differences.

In the US, where there is little cultural variation regarding suicide, the link between suicide and per capita gun ownership is overwhelming. Of the top five states in per capita gun owhership, four are also in the top top five in gun suicide rates and in the top three in overall suicide rates, while among the lowest five in per capita gun ownership, four are in the lowest five in gun suicide rates and three are in the lowest five in overall suicide rates.

75 JVA December 27, 2012 at 6:00 pm

That wikipedia page lists Afghanistan as having 4.6 guns per 100 residents – less than Bahamas or Cuba.

76 Brian Donohue December 27, 2012 at 7:33 am

Here’s a thought: if the whole ‘cold, dead hands’ crowd chipped in along these lines, y’all may be able to implement some of them low-cost protections for schools you were spitballin’ the other day. Hmmm, hmmm?

77 Brian Donohue December 27, 2012 at 7:36 am

“Nobody made a greater mistake than he who did nothing because he could only do a little.”

78 Alexei Sadeski December 27, 2012 at 5:16 am

I’m not convinced that FSS is a reliable proxy for county by county firearm ownership rates.

79 Some Random Economist December 27, 2012 at 8:23 am

I’m convinced it’s a terrible proxy. Most gum deaths, suicides and homicides, are committed by men. As a result, the error term in their proxy is correlated with how well off men are in the area. This would bias their estimates upward.

80 DocMerlin December 27, 2012 at 5:39 am

“The research reported here was supported by a grant from the Joyce Foundation”

Yikes. If pro-gun research had been funded by the NRA, many would speak up about it.

Second of all, his (interpolated!) time series is problematic.

“We ran panel regressions of GSS-based
estimates of gun prevalence against two proxies, FSS and the subscription rate to Guns and Ammo, the proxy used by Duggan (2001). The estimated coefficients of our GSS. measures on FSS are in every case significantly positive, and are especially strong when year fixed effects are omitted, while the subscription rate to Guns and Ammo performs less well and in some cases yields a negative coefficient estimate.”


Another yikes is that % urban has a large effect that reduces the homicide rate. But don’t urban areas have much higher crime and homicide rates? It seems something else is happening here.

Yet another yikes is that he includes log(Robberies) and log(burglaries) as explanatory variables for the dependent variable Log(Homicides), but he doesn’t include log(Assaults) as a dependent variable . I suspect a large amount of endogeneity there.

81 Andrew' December 27, 2012 at 6:15 am

They say something about using non-homicides to show that people don’t gun up in response to high crime/murder rates in their area (reverse causality). What do they mean there?

82 prior_approval December 27, 2012 at 7:30 am

They mean this – ‘Suicide is still the leading cause of firearm death in the U.S. In 2009, the U.S. firearm suicide total was 18,735 an 11% increase from the 2006 national gun suicide number of 16,883.’

And do note that a gun has yet to prevent a single suicide by shooting the shooter, regardless of whether the gun was legal or not.

83 mofo December 27, 2012 at 8:43 am

You keep flogging that horse, do you have any real point?

84 Brent December 27, 2012 at 10:36 am

I don’t think there is any good evidence that gun control will affect the suicide rate. In the US suicide rates have been more or less stable since the 1950’s. Also, the rate of suicide in the US is much lower than in countries with much lower gun prevalence.

85 GiT December 27, 2012 at 5:55 am

The paper was also discussed earlier in the cycle at the RBC blog:


86 Chip December 27, 2012 at 6:06 am

Not a gun owner myself but once you get past the justifiably powerful revulsion to mass killings of children, I don’t find the arguments for gun control very persuasive.

England banned guns and gun crime doubled over the next ten years. For me, living in London presented frequent encounters with thuggery whereas I never experience this kind of thing in the US.

Kids are five times more likely to die in swimming pools but no one really cares. The US president has killed 176 kids in Pakistan alone with personally approved drone attacks but no one really cares. The US government sent untraceable guns across the Mexican border and killed over 200 people and no one cares.

And if we did get a ban in place, would Chicago, DC and the inner cities where the vast majority of these homicides take place really become safer? Are they using less drugs since the War on Drugs began?

87 anon December 27, 2012 at 9:03 am


88 DJS December 27, 2012 at 1:00 pm


89 steve December 27, 2012 at 1:22 pm

” For me, living in London presented frequent encounters with thuggery ”

Where did you live in the US? Are you comparing large urban area with large urban area?

“England banned guns and gun crime doubled over the next ten years.”

Yes. Gun related homicides are up to 0.07 per 100,000 in the UK. They are at 3 per 100,000 in the US. As to crime in general, they do report things differently, they are Brits.

“While the number of crimes involving firearms in England and Wales increased from 13,874 in 1998/99 to 24,070 in 2002/03, they remained relatively static at 24,094 in 2003/04, and fell to 21,521 in 2005/06. The latter includes 3,275 crimes involving imitation firearms and 10,437 involving air weapons, compared to 566 and 8,665 respectively in 1998/99.[”


90 BintheCT December 27, 2012 at 2:11 pm


91 Ricardo December 27, 2012 at 3:53 pm

I’m not sure whether you intended to allude to the fact that it is already illegal to own handguns in DC.

92 ssh December 27, 2012 at 7:03 am

“. The estimates utilize a superior proxy for gun prevalence, the percentage of suicides committed with a gun, which we validate.”

Why is that superior, and superior to what? Why not just use gun ownership rate?

” we estimate the elasticity of homicide with respect to gun prevalence as between +0.1 and + 0.3″

In other words. the log(% all suicides by gun) is correlated with log(homicide rate). I don’t even know what that is supposed to mean. I like how they rename that ” gun prevalence” so it sounds like something more meaningful.

“the average annual marginal social cost of household gun ownership is in the range $100 to $1800.”

I have no idea what that is even supposed to mean. Is $1800 high for a(n) “average annual marginal social cost” of something? Are “social cost(s)” usually paid in dollars? And keep in mind, it is NOT “household gun ownership” they measure, but “% all suicides by gun”.

93 RPLong December 27, 2012 at 7:21 am

What is the social cost of freedom?

94 prior_approval December 27, 2012 at 7:53 am

Thousands and thousands of bodies, every year, apparently, if gun ownership is seen as freedom. ‘Suicide is still the leading cause of firearm death in the U.S. In 2009, the U.S. firearm suicide total was 18,735 an 11% increase from the 2006 national gun suicide number of 16,883.’

See that increase? The price of freedom. And here is something even more interesting about what the cost of freedom is –

‘hose persons with guns in the home were at greater risk than those without guns in the home of dying from a homicide in the home (adjusted odds ratio = 1.9, 95% confidence interval: 1.1, 3.4). They were also at greater risk of dying from a firearm homicide, but risk varied by age and whether the person was living with others at the time of death. The risk of dying from a suicide in the home was greater for males in homes with guns than for males without guns in the home (adjusted odds ratio = 10.4, 95% confidence interval: 5.8, 18.9). Persons with guns in the home were also more likely to have died from suicide committed with a firearm than from one committed by using a different method (adjusted odds ratio = 31.1, 95% confidence interval: 19.5, 49.6). Results show that regardless of storage practice, type of gun, or number of firearms in the home, having a gun in the home was associated with an increased risk of firearm homicide and firearm suicide in the home.’


I prefer the freedom represented by being less likely to die from a firearm myself, but then, I no longer live in the U.S.

95 RPLong December 27, 2012 at 8:39 am

The freedom to drive an automobile comes at the cost of many more thousands of bodies per year. Your fear is a childish irrationality. Every tool is dangerous. Loss of digits is no argument against table saws.

96 prior_approval December 27, 2012 at 8:52 am

Well, I’m a motorcycle rider myself.

Which means that I have a license, involving a certified course in how to ride, passed a test in terms of traffic regulations, and am also tested regularly in terms of vision.

My motorcycle is also registered, and is inspected on a regular basis.

Of course, the same applies to any other driver and motor vehicle operating legally on the public roads of the Commonwealth of Virginia.

In contrast, I need none of these things to buy a gun – and yet, a gun is a piece of machinery designed with essentially one overwhelming goal in mind. The same applies to motor vehicles, though the goal is different. The difference is when a motor vehicle is involved in fatalities, it is a failure of either the owner or the machine. When a gun kills someone, it is generally the intent of the owner and essentially the reason for the gun itself to exist.

We attempt to deal with the deaths caused by motor vehicles (which obviously include people killing themselves with them) in a wide variety of ways, and the success in reducing fatalities is measurable.

And yet, a certain group of people reject even having this discussion, seemingly conflating freedom with the ability to kill.

97 anon December 27, 2012 at 9:06 am

seemingly conflating freedom with the ability to kill.

Are you really that obtuse?

98 RPLong December 27, 2012 at 9:51 am

Why do people keep making this argument and others like it? Are you under the impression that someone who opposes gun regulations is in full support of motorcycle regulations?

First, you have to establish why the current level of motorcycle regulations is an improvement on a regulation-free scenario and I have to agree with you. ONLY THEN can you use those regulations in your attempt to convince me of gun control.

99 mofo December 27, 2012 at 10:14 am

Point of fact, you need a license to *operate* a motorcycle in public not own one. You also need a license (and in most places special training, background checks, finger printing, permission from the local police) to carry a firearm in public. If motorcycles were like guns, you would need to undergo a background check (and, often, a waiting period) just to buy one, and could be denied for a number of reasons that have little to do with you being a safe motorcycle owner.

And if gun owners seem like “people (who) reject even having this discussion” its because the overwhelming majority of anti-gun people have as their explicit goal the total ban of firearms, and are willing and even eager to distort the truth to achieve that purpose. Do you think that you would be so willing to have a discussion about motorcycle deaths if the discussion always wound up with a group of people calling for a total ban on motorcycles, irrespective of the facts?

100 Urso December 27, 2012 at 11:24 am

“overwhelming majority of anti-gun people have as their explicit goal the total ban of firearms, ”

Objectively untrue, but it’s amazing how many gun-rights advocates seem to genuinely believe this. They really do work themselves into a lather on this issue, and let their imaginations run rampant with fantasies of Obama sending Marines knocking down the front doors of every gun owner in the country.

101 mofo December 27, 2012 at 8:47 am

Oh, if only we knew what the national suicide rate was in 2006 and 2009! You should figure out what that is and repeat it over and over, because that will prove whatever point it is you are trying to make.

102 prior_approval December 27, 2012 at 9:01 am

I repeat it because you seem to miss the point – so without the number, let me emphasize the text ‘Suicide is still the leading cause of firearm death in the U.S.’

The reality remains that suicide, not crime or accidents, is where guns kill the most people. And yet, everywhere I have posted that fact is one someone has simply seemed unaware of how guns turn living people into bodies in the U.S.

The leading use of guns involving fatalities in America is suicide.

Let us at least use a factual basis when talking about the social cost of guns, and not the fever dreams which seem to permeate so much of this discussion in the U.S.

103 RPLong December 27, 2012 at 9:49 am

This has got to be the worst argument against guns I’ve ever heard. Do you know what the definition of the word suicide is?

104 mofo December 27, 2012 at 10:29 am

And what does the firearm have to do with it?

105 Alexei Sadeski December 27, 2012 at 10:07 am

You assume the consequent.

Also note that many studies have shown that loose gun regulation doesn’t lead to increased suicide rates – folks substitute other methods when guns are unavailable.

106 The Snake December 27, 2012 at 8:59 am

20 dead children. Freedom isn’t free!

Freedom is overated in the USA. We have plenty of freedom. I don’t mind that I cannot do 200mph on the highways, and I would not mind placing a ban on assault rifles and hand guns. I’d gladly sacrifice those freedoms in an effort to prevent another large scale massacre.

107 RPLong December 27, 2012 at 9:46 am

The freedom that you voluntarily wave is not the one you’re willing to sacrifice. What we are actually talking about are the freedoms other people enjoy that you think they ought not.

Let’s change the subject to a freedom you personally hold dear. I’ll make the case for why my life is made no worse off if you give up that freedom. Then we will see which side of freedom you are on.

108 The Snake December 27, 2012 at 10:49 am

1. I did not say that I though others ought not enjoy their assault rifles and hand guns. Rather I was pointing out what freedoms I personally would be willing to sacrifice for the greater welfare and prosperity of civilization. We all have our free will constrained in some manner for that purpose, although that point is often overlooked by Libertarians.

2. Regarding a freedom I hold dear: Well, I sure find seizing things I like the look of a lot of fun. I’d like to get me a band of merry men and take a few of the houses in the nice part of town. Problem is, government is always trampling on my freedoms!

109 RPLong December 27, 2012 at 11:07 am

Your point #2 destroys whatever credibility you imagine you have when you refer to what libertarians overlook in your point #1.

110 The Snake December 27, 2012 at 12:47 pm


On the contrary, sir. Your inability or reluctance to argue the point says more about the credibility of your position.

111 Norman Pfyster December 27, 2012 at 10:11 am

Banning guns would not prevent large scale massacres, in case you were wondering. No more than banning drugs prevents their use.

112 The Snake December 27, 2012 at 10:58 am

No, but guns sure make it easier though! They are really efficient for killing things.

My point is a ban would *reduce* – not prevent – the number of large scale massacres. Just as making pot illegal decreases the use of pot.

113 Elizabeth December 27, 2012 at 11:40 am

Do you have numerical evidence to support your claim about marijuana usage?

Restrictive firearm laws only apply to those willing to obey the law in the first place. Do you believe that fewer firearms in hands of law-abiding citizens removes firearms from non-law-abiding citizens?

114 msgkings December 27, 2012 at 11:57 am

Yes. If that CT mom had fewer (or none) or better secured or smaller-magazined guns, her crazy son might not have been able to kill so many small children.

The lawbreakers get the guns somewhere. Usually from law abiders.

115 Elizabeth December 27, 2012 at 12:11 pm

Do you believe that, without firearms available, the compulsion within that young man to commit mass murder would have gone away?

116 msgkings December 27, 2012 at 12:20 pm

No, I believe that if he didn’t have easy access to his mother’s arsenal, or if she didn’t have one, or if the only guns he could possibly procure were simple 6 shot handguns, that the incident would have been far less deadly if it occurred at all. It’s pretty common sense to want to make guns harder for crazy people to get, and to make them less capable of firing multiple bullets quickly. Most of these crazies impulsively snap and grab whatever’s easy to come by and go do their killing. Make the guns harder to access and make them less deadly, and a lot of innocents are spared.

But you’ve proven upthread to be completely untainted by common sense. Can a more reasonable gun rights supporter engage my argument? I’m not advocating outlawing guns. I want to try to work towards an equilibrium where the only guns out there are hunting rifles and 6 shot handguns, and those guns are a lot harder to get and taken care of much more carefully than that CT shooter’s idiot mom did. Please tell me the problem with that goal.

117 The Snake December 27, 2012 at 12:33 pm

@Elizabeth 11:40AM

1. Do I need empirical evidence on this one? Are all prohibitions useless because people would not commit those acts anyway?

2. There is the contextual aspect to gun control – that there are firearms already in circulation – to which your second point speaks. We cannot simply legislate and have certain guns disappear off the face of the planet. Yes, it would take years to have effective gun control, and it would take resources, like gun buy-back programs. To answer your question: No. But I’m not convinced that it is an argument against gun control.

118 Elizabeth December 27, 2012 at 12:40 pm

Common sense is not, in fact, common. It is trained into one by social norms. What you assume to reasonable, I believe to be unconstitutional. Freedom involves daily personal risk, which our forefathers assumed, knowing they would be killed, messily and publicly, if they failed to stand firm against our nation’s enemies. Why should I personally assume less risk for myself, when so many have died to provide me with this freedom?

119 The Snake December 27, 2012 at 12:42 pm

Elizabeth @12:11PM

No. Do you really think he could have killed so many people, so quickly, without the use of a firearm?

120 T. Shaw December 28, 2012 at 9:12 am

So, why isn’t NBC’s David Gergory in jail for violating one of DC’s 16,387 gun laws?

That high magazine capacity he absconded into that TV studio may have killed someone!

I know! Because laws only are used against the ruled class.

Anyhow, these statisticians made their study employing similar scientific methodologies as RE appraisers deployed during the RE bubble: “As Instructed.”

They started with the answer: guns kill! (Firm grasp of the obvious.) They simply needed to calculate, distort, fabicate, formulate, omit, and throw around fifty-cents words, graphs, pie-charts, and tables to support the narrative: ban guns.

And so, after weeks of waving 20 tiny, bloody shirts, and 24/7 lies, the truth does not comes out.

No so-called assault weapon was fired in the Sandy Hook massacre.

Gallup Poll: NRA More Popular Than the Lying, Vile Media: NRA: 54% positive 38% negative; while 60% of Americans have no or little trust in the lying, vile media.

121 Elizabeth December 27, 2012 at 2:16 pm

Poison and arson come to mind.

122 Elizabeth December 27, 2012 at 2:17 pm

My apologies. This was meant as a response to The Snake @ 1242.

123 msgkings December 27, 2012 at 4:46 pm

Also he could have nuked the whole town!

124 Cyrus December 27, 2012 at 11:26 pm

In other parts if the world, those wishing to kill significant numbers of others as part of their suicide, and who maintain that idea long enough to plan for and act on it, detonate themselves in a crowd.

125 Slocum December 27, 2012 at 7:43 am

Hmmm — this seems like a problem:

The estimates presented below are based on panel data for 200 counties that had the
largest population in 1990, or a subset of those counties, for the period 1980 to 1999. We
also present estimates based on state-level panel data. The 200 largest counties accounted
for 74% of all homicides in the United States in 1990.

Choosing the 200 highest population counties would tend to exclude precisely those rural counties with high rates of gun ownership and low homicide rates, no? And then, of course, there’s the question of causality — are homicide rates high because guns are prevalent? Or are people more likely to buy guns when they live where homicide rates are high? Given how easily guns are moved across county lines, why would we expect to see much of a correlation between household gun ownership and gun ownership by criminals? Also — would anybody be surprised to see also a correlation between homicide rates and, say, prevalence of bars on windows? But would anybody suggest that there is a ‘social cost’ of barred windows?

126 Chris December 27, 2012 at 8:09 am

Choosing the 200 highest population counties would tend to exclude precisely those rural counties with high rates of gun ownership and low homicide rates, no?

It would exclude those counties, but that’s not going to bias the estimates. Any time-invariant county characteristic like that is going to be soaked up in the county fixed effects. That’s why you use a panel regression.

127 Brent December 27, 2012 at 10:08 am

It’s difficult to believe that excluding high gun owning, low homicide counties from the analysis is not going to bias results that are attempting to correlate gun ownership with homicide.

128 Chris December 27, 2012 at 10:55 am

Do you understand how a fixed effects panel regression works? It’s perfectly fine if you don’t, but your question seems to indicate you do not. Let me make it clear: Identification of the parameter that’s being estimated is not based on a correlation of gun ownership and homicide rates. Time-invariant place specific factors like “high gun owning, low homicide counties” are in the fixed effects.

129 Brent December 27, 2012 at 11:49 am

True. I don’t know much about fixed effects panel regression, nor does it seem necessary for this discussion. The authors are making some very broad statements derived from some very limited data. I don’t buy it – I guess that’s skepticism from being a scientist for the last fifteen years.

By the way, the authors are explicitly correlating gun ownership with homicide rates:

“each additional 10,000 gun-owning households leads to around 6 additional crime-related gunshot injuries”

Perhaps I’m missing something, but this whole paper is really setting off my BS detector.

130 Chris December 27, 2012 at 12:04 pm

I don’t know much about fixed effects panel regression, nor does it seem necessary for this discussion.
Really? Understanding how a paper does its work isn’t necessary for discussing whether its conclusions are valid?

the authors are explicitly correlating gun ownership with homicide rates
No, they aren’t, it’s not that simple. It is absolutely possible in that regression for areas with high gun ownership to have high homicide rates AND STILL have the effect of more guns on homicides to be negative.

I don’t buy it – I guess that’s skepticism from being a scientist for the last fifteen years.
Skepticism is all well and good, but that doesn’t warrant a blanket dismissal of papers that you admit to not understanding. As a scientist, you wouldn’t like it if people did that with quantum mechanics or general relativity or evolutionary theory.

Like all papers, it has its flaws, but a shotgun approach of point out “mistakes” which simply are not being made is irritating.

131 murks December 27, 2012 at 7:55 am

Who needs guns when you can buy drones? Ah, those silly Americans.

132 doodad December 27, 2012 at 8:03 am

Maybe I am missing something here…

The paper mentions the majority of people killed by guns are already engaged in criminal activity (likely an understatement). It doesn’t mention that had these particular victims lived, they would continue to suck up resources through income redistribution, government social programs, incarceration, and the constant need for policing.

So on a warm weekend in Chicago when half-dozen gangmembers are killed, it is likely each of these individuals would otherwise cost the state between $20,000-50,000 per year, for life. This doesn’t include externalities either. For example, each of these victims likely disrupted every highschool class they bothered to attend. Additionally, how many were basically out-of-wedlock baby machines, thus repeating the cycle? For some of these shooting victims, it wouldn’t surprise me if they were costing society >100k annually.

I’m not saying we should rejoice when these people are shot. But from a strictly economic POV, how is this not a net positive?

133 prior_approval December 27, 2012 at 8:41 am

‘The paper mentions the majority of people killed by guns are already engaged in criminal activity’

No, that is incorrect – ‘Suicide is still the leading cause of firearm death in the U.S. In 2009, the U.S. firearm suicide total was 18,735 an 11% increase from the 2006 national gun suicide number of 16,883’

134 Paquette December 27, 2012 at 9:13 am

Isn’t suicide illegal?

135 Alexei Sadeski December 27, 2012 at 10:09 am

No, you are incorrect.

Paper only deals with homicides – suicides aren’t included in their numbers here.

136 Nanoprof December 27, 2012 at 9:11 am

Because a new gang member takes their place, if the social forces that create/support those gangs persist.

137 Thomas December 27, 2012 at 9:41 am

We should outlaw those social forces.

138 doodad December 27, 2012 at 10:38 am

If Americans were truly interested in stopping those social forces, they would simply pay certain individuals to stop reproducing. In the case of habitual criminals or the most egregious deadbeat fathers/mothers, implement some sort of compulsory sterilization. It doesn’t stop it overnight, rather in 15 to 18 years, but it will stop it.

Of course, if a public figure suggests this, then here come the hand-wringers, fretting about eugenics and Nazism.

139 Cyrus December 29, 2012 at 4:35 am

This looks like the”lump of labor” fallacy applied to a black market. Any evidence that the fallacy becomes true when the labor becomes illegal?

140 GiT December 27, 2012 at 11:23 am

It’s always nice to see a little justification of social cleansing in the morning. You know, the babies of the poor are really just a big drag on everyone else; why don’t we just eat them?

141 Mac December 28, 2012 at 11:22 am

Instead of eating them, why not let the mother murder them prior to birth?

Liberals don’t seem to have a problem with this concept in the name unencumbered casual sex or freedom from the financial burden of caring for a child, as long as it’s the responsible party (the mother) that’s relieved. If it’s some blameless party (the tax payer) then it’s a societal duty, and the burden must be bared for someone else’s irresponsibility.

It’s ironic that when 20 six year olds are murdered by a nutcase, liberals are demanding a further curtailment of the Second Amendment of the Bill of Rights, while the murder of 1.2 million unborn children is celebrated as a victory for women’s privacy (under the due process clause of the 14th Amendment.)

142 doodad December 27, 2012 at 8:55 am

Correction: the bulk of homicide victims that are killed by guns are already engaged in criminal activity. Once you include homicide victims who are on the periphery of criminal element, that percentage likely skyrockets.

143 GiT December 27, 2012 at 11:26 am

Some research (by Hemenway) suggests that most criminals who are shot are in fact being victimized when they are shot – whatever criminal acts they have engaged in were not in process when they were shot. But I guess they deserve it, right? It’s just karma.

144 Ram December 27, 2012 at 8:59 am

IIRC, in the period following the 2008 election, gun sales went through the roof. I don’t know the data well enough, but someone ought to determine whether gun sales consistently accelerate/deccelerate when we elect Democratic/Republican presidents. If so, the partisan identification of elected presidents would be a good instrument for exploring casual relations between gun prevalence and violence.

145 Therapsid December 27, 2012 at 9:20 am

No, it’s not one place to start. It’s a relief that the U.S. at its founding didn’t turn to economists to ascertain the marginal social costs of the other 9 amendments in the Bill of Rights.

This same sort of vaguely utilitarian logic can and will be used to limit free speech. “We discover that the marginal social cost of hate speech is in the range of $200 to $100,000.”

146 Kevin P. December 27, 2012 at 9:39 am

This is a 2006 paper by Philip J. Cook and Jens Ludwig, well known anti-gun “researchers”. Their research was funded by the Joyce Foundation, which has long funded gun control “studies” and other efforts:

A 1999 paper by the National Center for Policy Analysis found a considerable benefit to firearms ownership:

The total social benefit from defensive gun uses is equal to the annual savings from non-deadly gun uses plus the annual savings from defensive gun uses in which the criminal is killed. And any number selected is greater than the cost of firearm violence.72 For instance, even under a best-case scenario for lawsuit proponents, using the NCVS firearms crimes number (915,000), and the Tarrance Group’s defensive gun use statistic (764,000), the net benefit from defensive gun use ( the total benefits less the total costs) ranges from just over $90 million annually to $3.5 billion per year. Using Kleck’s more credible estimate for annual defensive gun use, the net benefit ranges from $1 billion to $38.9 billion annually. [See Table IV.] The annual net benefit of civilian gun use in the prevention of crime is even greater if the NCVS overestimates firearms crimes and the Bureau of Justice Statistics figures (483,000) are more accurate.

147 GiT December 27, 2012 at 11:32 am

The Tarrance Group is a Republican polling organization. Why use their number for defensive gun use? And why compare ambiguous reports of defensive gun use to concrete incidents of firearms crimes, rather than compare justified homicides/legal interventions + alleged defensive use to gun crimes + alleged gun victimization. After all, reports of defensive use taken at face value by those who report them, if submitted to objective review, often turn out to be illegal and/or conflict-escalating; and polling on victimization can pick up more incidents than polling on self-defense in the first place, regardless.

148 GiT December 27, 2012 at 11:34 am

The Tarrance Group is a Republican polling organization. Why use their number for defensive gun use? And why compare ambiguous reports of defensive gun use to concrete incidents of firearms crimes, rather than compare justified homicides/legal interventions + alleged defensive use to gun crimes + alleged gun victimization.

After all, reports of defensive use taken at face value as the honest truth of those recounting them, if submitted to objective review, often turn out to be illegal and/or conflict-escalating; and polling on victimization can pick up more incidents than polling on self-defense, regardless.

149 GiT December 27, 2012 at 11:35 am

Sorry for the double post. MR was deceiving me.

150 Steven L. December 27, 2012 at 9:49 am

When sociologists decry how many people were killed with guns, they ignore the fact that many of those people deserved their fate.

If gangs in East Los Angeles slaughter each other, it’s a big win for the rest of us.

151 Jan December 27, 2012 at 4:45 pm

And it’s an especially big win for all the children who live in the afflicted neighborhoods. Not only do the bad guys take on the tough work of cleaning up the neighborhood, but the kids receive some very good life lessons by witnessing those stupid bad guys killing each other.

152 MikeDC December 27, 2012 at 9:59 am

I will not take seriously any paper that includes suicide by firearm in an argument relating to costs and benefits.

Omit these, run the numbers again, and then tell me the result.

153 Noah Yetter December 27, 2012 at 10:15 am

Absolutely agreed. Suicide is a funndamental human right. That it is not recognized as such is a great tragedy.

154 prior_approval December 27, 2012 at 10:45 am

However, the Swiss do agree with you – they just feel that blowing your brains out is incredibly messy, and not a suitable use of firearms.

‘Legislation on euthanasia in Switzerland permits assisted suicide. For example, lethal drugs may be prescribed as long as the recipient takes an active role in the drug administration. Active euthanasia (such as administering a lethal injection) is not legal.[1] For assisted suicide, the law does not require a physician to be involved, nor does it require the recipient to be a Swiss national. These aspects of the law are unique in the world.’


155 Chris H December 28, 2012 at 2:41 pm

I was wondering if anyone else caught that.

Indeed, going by the first order effects, if decreasing gun ownership decreased the amount of suicides would in fact be increasing the social costs would it not? An economic actor is prevented from engaging in a trade for a weapon that would increase the utility for both the seller and the buyer (as people who commit suicide view death as a benefit not a cost). The first-order costs of suicide are fully internalized by the suicdee. The second order effects are ambiguous and depend upon numerous factors such as the number of friends and loved ones who will grieve, whether that was taken into account by the suicidee (the practice of suicide notes shows that it very often is), the heirs who will receive inheritances, and the enemies of the person who are glad that he died.

If suicide is going to be considered a “social cost” it must first be shown what the average social cost of suicide is (the cost to the individual suicidee cannot be included for the above reasons). Show me a convincing study that the social costs of suicide are on average negative and then I will be willing to consider a study on the social costs of gun ownership to the extent that they use the estimate of the average social costs of suicide. As far as I could see this was not done in this study and therefore that’s a serious methodological flaw given the higher number of suicides involved in gun related deaths.

156 Noah Yetter December 27, 2012 at 10:14 am

So, since my household owns well north of $10,000 worth of handguns, we’re obviously in the clear in terms of net benefits, right?

157 prior_approval December 27, 2012 at 10:41 am

‘I will not take seriously any paper that includes suicide by firearm in an argument relating to costs and benefits.’

The Swiss, a nation with a famously well regulated (not to mention well armed) militia, disagree with your perspective.

‘The issues of gun suicide and Switzerland’s high rate of weapon ownership came under the spotlight again in January. The police chief overseeing security at the annual meeting of the World Economic Forum in Davos turned his service revolver against himself a day before the event began.

His case is hardly unique. From 1996 to 2005, 3,410 suicides, or between 24 and 28 per cent of all those in Switzerland, were committed using firearms.

That percentage trails the United States, it is true, where 57 per cent of suicides involve a gun. But few European countries come anywhere near Switzerland.

And of other English-speaking countries, the highest rate is in Canada, where 19 per cent of suicides were by firearm in 2000. In England and Wales the figure stands at 2.8 per cent, and in Scotland at 1.8 per cent.

Guns are highly efficient: experts say that 90 per cent of suicide attempts involving firearms are successful.’


But then again, maybe the Swiss shouldn’t be too concerned – they would need to double their firearm suicide rate before approaching the level which seems to leave at least many putatively American commenters here absolutely blase.

158 prior_approval December 27, 2012 at 10:47 am

Poor posting placement – this belongs above in response MikeDC.

159 Numbers December 27, 2012 at 11:51 am

Yet up until 2007 the rate of death by suicide in the U.S. was lower than Switzerland.


160 prior_approval December 27, 2012 at 12:06 pm

Suicide by firearms is the point, and the social costs of guns.

It is a quite complicated discussion, of course. Which is not uncomplicated is that rate of suicide in the U.S. by firearms is much, much higher than in other countries. Another uncomplicated fact is that the effectiveness of using firearms when committing suicide is much higher than other methods. A final uncomplicated fact is that the rate of gun ownership in the U.S. is much higher than in other countries.

In other words, the U.S. could reduce the total number of suicides per year if firearms were less readily accessible to those interested in committing suicide. As demonstrated by Canada, with a firearm suicide rate a third of the American one, though Canadians have more than easy access to hunting weapons (though not to handguns – obviously, a point which adds complexity to any comparisons or conclusions).

The truly complicated point is how to cover the social costs of guns in the U.S. when looking at the likely thousands of additional dead fellow citizens each year. In other words, having suicide involving a firearm no longer be the leading cause of firearms deaths would be a step in the right direction.

And do note – mental health services directed to reducing the number of deaths caused by firearms would be most effectively directed towards those likely to commit suicide. Which is a situation where it is provably true that the gun the shooter uses has never prevented a single fatality.

161 Elizabeth December 27, 2012 at 12:29 pm

Do you believe that the restriction of firearms decrease the total percentage of suicides?

162 Better Question December 27, 2012 at 1:15 pm

Does he/she believe if all guns could be effectively eliminated would the number of total suicides decrease by the number of previous gun suicides? From below it seems the answer might be yes.

163 GiT December 27, 2012 at 1:42 pm

Suicide attempts with a firearm are much more likely to be successful (something like 90% result in death. Other methods are much much lower). That could just be those who want to go through with it selecting the best available method, but presumably one could compare success rates across countries similar to the US apart from gun ownership. It wouldn’t be surprising if guns turned some portion of what would have ended up as suicide attempts into successful suicides.

164 K.P. December 27, 2012 at 2:34 pm

Canada and the U.S. have almost identical suicide rates (around 11 per 100,000) so if it doesn’t look like stricter gun regulation would amount to fewer suicides, just different methods, more hangings and poisons. However, maybe Canadians really are just way more suicidal than Americans.

165 GiT December 27, 2012 at 3:13 pm

The interesting thing would be to compare success rates for non-firearm suicides. It could be that there are just more attempts with similar success rates for non-gun methods (northerly climates can do that, right?) Do people committed to the project just commit to a more difficult method?

166 Noah Yetter December 27, 2012 at 5:02 pm

You assume that reducing suicides is an unqualified net positive. It is not. As I stated above, suicide is a fundamental human right. Denying people their rights does not create value, it destroys value. If someone wishes to end their life, for whatever reason, the best thing we can do for them is to let them.

167 GiT December 27, 2012 at 5:24 pm

Many suicides are merely attempts; and many attempts aren’t followed up on. But if one uses a gun, you’re quite unlikely to get a chance to reconsider.

168 maguro December 27, 2012 at 8:35 pm

Pretty tough to reconsider after jumping off a tall building, too.

In reality, it is not at all difficult to kill yourself.

169 prior_approval December 28, 2012 at 1:46 am

To ensure that this information is seen as a reply to this –

‘Pretty tough to reconsider after jumping off a tall building, too.

In reality, it is not at all difficult to kill yourself.’

Let’s look at some American numbers. And these are fairly accurate numbers, it must be noted – though university suicide statistics are fudged in terms of trying to cover up student suicide, the police/coroner do keep accurate records on the bodies they handle. And when it comes to suicide by firearm or leaping off a building or bridge, it is fair to assume those numbers are reliable.

So let us see how suicide looks in a city with many tall buildings and very strict laws concerning gun ownership – why speculate when facts are available, after all.

And a trigger warning – NYC smugness oozes through the factual information.

‘Many adjectives are frequently used to describe New Yorkers, but “cheerful” and “optimistic” are not among them.

And yet from the health department comes news that the city’s suicide rate is about half the national average — about 6 deaths per 100,000 people here, compared with 11 in the rest of the country.

The health commissioner, Dr. Thomas A. Farley, attributes much of the difference to the city’s relatively strict gun laws: he noted that firearms were the chosen method in only 12 percent of suicides in New York in 2010, compared with 51 percent nationwide in 2009. In the three states with the nation’s highest suicide rates — Montana, Alaska and Wyoming — nearly two-thirds of suicides were by firearm. (New Jersey, the least suicidal state, has very strict gun laws.)

But it’s not as if New Yorkers lack access to means of self-destruction. There are many tall buildings here (jumping was the second most common suicide method, after hanging), and a 468-station subway system that runs round the clock (death by train was fifth, behind guns).

Perhaps the city is, in its own odd way, a more uplifting, life-affirming place to live than America at large. Or at least a harder place to justify taking your life.’


Strange – almost as if one could draw the conclusion that easy access to guns increases the actual number of people who kill themselves, as in NYC (suicide involving firearms was in place four, just ahead of the distinctly urban method of being killed by subway). Oddly enough, the Swiss seemed to have made the same conclusion in terms of suicide prevention and firearm effectiveness.

170 Chris H December 28, 2012 at 3:00 pm

“Many suicides are merely attempts; and many attempts aren’t followed up on. But if one uses a gun, you’re quite unlikely to get a chance to reconsider.”

Have you considered that the effectiveness of guns might be well known and therefore people who use guns as suicide helpers are not in fact doing a cry for help. Now it’s true that it’s impossible to reconsider once one has decided to commit suicide, but technically speaking that’s true for a LOT of interactions we humans have. When you sign a labor contract which assess a penalty for quitting early you can’t return to the pre-contract state. You have made a choice you cannot fully reverse. Indeed taking this further this is really true no matter what interaction you undertake. A complete reversal of a decision to return to the pre-decision state is not actually possible, the only difference is how close you can approximate that pre-decision state, and even then there are some decisions besides suicide where even coming close to that is extremely difficult. Take renouncing US citizenship for instance. If you renounce your citizenship that is a permanent act that can’t be regained. This doesn’t mean we shouldn’t ever allow people to make that choice.

171 GiT December 28, 2012 at 3:05 pm

Yes, I have considered that, as insinuated in posts above.

172 maguro December 27, 2012 at 11:25 am

I assume the baseline here is some kind of gun-free utopia and not a situation where the tens of millions of unregistered/illegal weapons remain in circulation, mostly in the hands of criminals and other bad actors. What would the marginal social cost of the latter scenario be?

173 prior_approval December 27, 2012 at 12:11 pm

Most likely, thousands of lives saved every year as the rate of suicide by firearm decreased to the Swiss rate (9,000 Americans no longer dead from suicide by firearm) or the Canadia rate (12,000 saved Americans).

Obviously, approaching UK rates is unlikely, even though it would save 18,000 Americans a year.

Suicide is the leading cause of death through firearms – all other scenarios pale before this simple reality. Legal/illegal, crime/self-defense, accidents – these are literally second rank considerations when counting the bodies involving firearms, day after day.

174 maguro December 27, 2012 at 12:30 pm

Suicide rates in Canada and the US are pretty much identical, and Switzerland’s is considerably higher than either. So, uh, no.

175 Shanghaied December 27, 2012 at 3:11 pm

@prior_approval: What makes you think suicide rates would go down? As gun ownership declines, we may become more similar to other countries with low gun ownership and far higher overall rates of suicide?

As gun ownership declines, our suicide rate could rise to French levels (resulting in 9,000 more dead Americans every year), or Belgium (16,800 more Americans dead), or Japan (35,400 more Americans dead) or god forbid South Korea (58,800 more Americans dead as the suicide rate jumps from 12 suicides/100,000 to 31.2 suicides per 100,000).

176 prior_approval December 27, 2012 at 11:20 pm

This is where the discussion gets complicated, of course, since suicide has many components, where culture plays a large role, as does such things as geographical location.

But after a classroom of children was killed, a mjaor proposal seemingly put forward by those opposed to any other attempt to reduce the amount of death through firearms in America was for mental health services provided by the government to be increased. (Proposing that more weapons be distributed would not reduce the amount of death through firearms, as accidents would likely increase by a couple of deaths a year if 100,000s of weapons were distributed in schools. And the purpose of those weapons would be to kill someone in the event of mass murder.)

If such services were effective, then I would expect the suicide rate to go down. If the rate went down, then it is equally fair to assume that firearms would be involved in less death.

The latest American statistics for suicide by firearm had a 2000 body increase from 2006 to 2009 – it is reasonable to assume that rate means that currently, 20,000 Americans have killed themselves with a firearm this year.

If one wishes to reduce the amount of death caused by firearms, a discussion about how to best deal with why suicide in the U.S. involves firearms in 57% of the cases, and how to best reduce that rate, would be a fine discussion to have with people who seemingly don’t understand that the leading use of guns killing people in America is to commit suicide.

What bothered me is how the ‘social cost’ of the link excluded suicide, simply using it as a baseline. That is a level of callousness that is unimaginable to me.

Unfortunately, this discussion is unlikely to happen – see the reactions here alone to the most basic reality of firearm use in America.

It isn’t heroic defense against mass murderers – those events are extremely rare. It isn’t to defend life, liberty, or property from those using guns to take life, liberty, or property – that is a tiny fraction of the events where gun use leads to death.

The single largest use of guns in America which leads to death is suicide. A statistic, apparently, to calibrate a scale to continue the same tiresome discussion while ignoring what guns are used for most in America – the death of the people pulling the trigger.

177 Bill December 27, 2012 at 12:02 pm

The purpose of the Second Amendment is to arm people in order to prevent future tyranny. They need the tools to do this.

The term “Well Regulated” in the Second Amendment meant “Well Manned and Equipped ” in 1791 as was determined in the 1939 United States v. Miller case after referencing the autobiography of Benjamin Franklin. The concept of Government Regulation, as we understand it today, did not exist at the time.

United States v. Miller also determined that the term “Arms” refers to “Ordinary Military Weapons”. American Citizens have the right to Keep and Bear, which means Own and Carry, any weapons that a soldier carries into battle. That includes past, present and future weapons. A Militia consisted of armed volunteers willing to fight with their personal arms and not under government control.

To limit the Second Amendment to muskets would be the equivalent of limiting the First Amendment to writings in quill pens.

Liberty is worth the risk of death!

178 GiT December 27, 2012 at 3:14 pm

So unrestricted access to RPGs, rocket launchers, artillery platforms, tanks, fighter jets, &etc follow naturally, I trust? I mean, they’re all just ‘armaments.’

179 Urso December 27, 2012 at 4:48 pm

“[My] Liberty is worth the risk of [your] death!”

180 The Anonymouse December 28, 2012 at 12:08 am

Very clever.

But your liberty is also worth the risk of my death, as myself or any other veteran has demonstrated.

181 T. Shaw December 28, 2012 at 3:34 pm

You people (think you own me) are demanding, “What is the ‘gun community’ going to do about this (Sandy Hook) tragedy?”

Well, I dunno.

What is the gay community going to do about this Penn State gay child molestation crisis, and about these 2,5000,000 newly AIDS-infected people in 2011?

What is the Muslim community going to do about 10,000 Islamic terrorist massacres since 9/11/2001?

I could go on about punishing innocent people for the acts of maniacs,

182 P S December 27, 2012 at 12:03 pm

There are already 300 million guns in private hands. Tons of ammunition and magazines. Private citizens already own more guns then the US Military. Government will not be able to disarm people without a bloody civil war. We will not allow ourselves to be passively disarmed.

183 dead serious December 27, 2012 at 7:56 pm

Then we’ll just have to work to limit and tax the hell out of the sale of ammunition.

184 steve December 27, 2012 at 1:36 pm

Since Lott’s work always comes up, nice piece by pro-gun Robert Ehrlich on the problems with Lott’s studies. Best if you are mathematically inclined.



185 Les Hardie December 27, 2012 at 1:38 pm

Imagine large capacity magazines have been banned.If someone is intent on mass murder, his inability to use such magazines will slow him down very little, because it takes only a few seconds to eject a spent mag and insert a fresh one. If no one is nearby to shoot back, he has all the time he needs to finish the task. If someone is nearby who can shoot back, the fraction of a second it takes the shooter to reload mags can be the difference between more victims and stopping the shooter. Thus having armed guards or teachers with CCW in schools makes sense.

186 john q public December 27, 2012 at 3:47 pm

Insurance companies would charge you more for life and homeowners insurance, if they could. The NRA took care of that, made it illegal.

187 GiT December 27, 2012 at 5:27 pm

This is the sort of stuff that pushes the NRA towards the edge of toleration – giving itself special insurance exemptions, banning federal research on gun violence, getting itself special exemptions to donor disclosure rules, lobbying against DC getting a house vote because of their gun laws. Stick to arms regulations.

188 ecurb December 28, 2012 at 5:25 am

People like prior_approval and msgkings are why I stopped even visiting this blog. The comments have gotten even worse since then, and there is even less actual content on the blog itself.

Chris, I understand why time-invariant county characteristics don’t undermine the study in the way Doc suggested. However, surely selectively restricting your panel data allows you to produce results for only those elements of the panel that you use?
Say we have a panel that consists of a cross-sample of the population. We wish to determine the correlation between rock concert attendance and deafness. Using a fixed effects panel regression allows us to ignore the unobservable differences between individuals. However, if we use only the tallest 20% of the sample, we will mostly be examining men. Should rock concerts affect the hearing of men and women differently (or tall people and short people, for that matter), we will not be able to apply our estimate to the whole population.
Similarly, an increase in handguns over a period of time may have a dramatically different impact on the homicide rate in cities vs rural areas. The estimate only gives results in marginal homicides for the 200 most populated counties.
Am I completely wrong?

189 prior_approval December 28, 2012 at 6:22 am

Such a nice compliment – I’m sure you have already found another echo chamber more to your taste.

190 BobMendon December 28, 2012 at 8:44 am

I agree that data was highly selective and based on a single statisitcal source. Meanwhile wider data is available but was ignored.

191 T. Shaw December 28, 2012 at 1:22 pm

I freaking love statistics!

What is the social cost of the NYC subway system?

So far (three more days to run!) in December 2012, two men were murdered by being pushed in front of NYC subway trains. Ergo, odds are 1:2,650,000 (5,300,000 daily weekday ridership) you will be killed by a NYC subway train. In 2012, so far (what?) 12 (Sandy Hook was all handgun) were killed by assault rifles. Your odds are: 12 in 315,000,000 or 1:26,250,000.

Suicides by NYCTA subway, 343. They legally could not get a gun in NYC.

I truly love statistics!!

192 ecurb December 28, 2012 at 3:57 pm

I must say that I would really like a stats-minded person to correct me if I’m wrong here. My last class was quite some time ago, and my current job doesn’t give me any practice.

193 ecurb December 28, 2012 at 5:26 am

Ugh, not sure why all my paragraph breaks vanished there…

194 BobMendon December 28, 2012 at 8:42 am

Although this study was published in a peer-reviewed publication, that does not speak for the validity of the study or it’s methodology. If that were presented by a grad student, they probably wouldn’t get a very good grade. Pay particular attention to where assumptions were being made and also where the data was obtained from. I was most surprised that the CDC WISKARS database was not referenced and that there seemed to be a glaring lack of controls. Nice try but come back when you are ready to change your methodology.

195 Borepatch December 30, 2012 at 8:34 pm

BobMendon is entirely correct: the paper is junk. It relies on a proxy of questionable accuracy (which the paper does not discuss in any but the most cursory manner). At the same time it ignores much better direct measures:

1. Background checks are submitted to the NCIS by gun stores at the time of purchase, and so directly measure firearm availability.

2. Firearms manufacturers’ SEC 10-K filings annually break down sales figures for investors.

3. States report how many applications for Concealed Carry licenses are submitted and approved. This may be the most accurate measure of “prevalence” because it doesn’t measure firearms that are left in people’s homes but rather the number of people who wish to have a firearm on their person at any particular time.

Cook & Ludwig either chose not to use these and instead substitute a proxy of questionable utility, or they did not realize that this data is available. The first suggests bias and the second suggests incompetence. However, we need not draw either conclusion to recognize that their study is worthless in that it is falsified by hard data: the murder rate is sharply down over the last 15 years while the number of firearms that are lawfully owned are sharply up over the same period.

The authors dismiss John Lott’s “More guns, less crime” thesis pretty much out of hand, but there is simply no getting past the fact that there is *no* correlation between increased firearms ownership and the murder rate. That’s the *best* *case* view of the data from their perspective, and it runs completely counter to their conclusion.

Net/net, this is another study to be filed in the “Junk Science” drawer along with Bellesiles, Perrin, Kellerman & Reay, Loftin, Koop & Lundberg, et al.

196 Floccina January 1, 2013 at 10:02 pm

Of course the social cost might be balanced by the private benefits to gun owners.

Comments on this entry are closed.

Previous post:

Next post: