Private Guns, Public Health

This is David Hemenway's response to criticism of his book Private Guns Public Health by Kevin Baker at The Smallest Minority.

September 27, 2004

I was asked to respond to what is claimed to be a critique of my book by Kevin Baker. I have neither the time nor inclination to have a detailed response to the many assertions and arguments he makes, many of which are wrong or misleading.

It turns out that Baker isn't really discussing my book Private Guns Public Health, but a magazine article about it. Unfortunately it seems that Baker may not have read my book (or the hundreds of journal articles that the book summarizes). It does seem silly for him to accuse the journalist who tried to reduce a 300+ page book and 3 hours of interviews into 3 interesting pages of text, as engaging in "bait-and-switch" tactics or not sufficiently discussing what Baker would have liked discussed.

I will talk about one issue, to illustrate the type of problem found in Baker's discussion.

A dozen case-control studies all find that, in the U.S., a gun in the home is a risk factor for "violent death" (i.e., homicide, suicide or unintentional gun death). Some of the other risk factors accounted for in one or more of these studies include age, gender, community, living alone, education, alcohol illicit drug use, depression medication, and psychiatric diagnosis. Ecological studies also find that, across U.S. states and regions, higher levels of household gun ownership are associated with higher rates of homicide (due to higher gun homicide rates), higher rates of suicide (due to higher gun suicide rates) and more unintentional gun deaths. Some of the other risk factors accounted for in one or more of these studies include poverty, alcohol consumption, unemployment, urbanization, divorce, education, violent crime, major depression, and suicidal thoughts.

Massachusetts, where I live, is a state with (relatively) low levels of household gun ownership, strict gun control laws, and low rates of violent death. I remarked to the journalist, who lives in Massachusetts, that I was glad I lived in Massachusetts and that "It's nice to have raised my son in Massachusetts, where he is so much safer" than most other states. Baker took this quote, asserted that I live in Boston, which I do not, and made comparisons to violent death in parts of Arizona, a state that has more permissive gun laws than Massachusetts.

So, let's compare Massachusetts and Arizona. Here are data from 1999--2001, the most recent time period available, easily obtained from the CDC WISQARS website.

Number of Deaths and Mortality Rate Ratio, 1999-2001
  Arizona
pop: 5.154 million
Massachusetts
pop: 6.356 million
Mortality Rate
Ratio, Arizona v.
Massachusetts
Homicides 1,374 501 3.4
  Gun 909 218 5.2
  Non-Gun 465 283 2.0
Suicides 2,317 1,244 2.3
  Gun 1,433 330 5.4
  Non-Gun 884 914 1.2
Unintentional Gun 47 6 10.0
Total Gun Deaths 2,460 565 5.4

In other words, a resident of Arizona is over 5 times more likely to be murdered with a gun, commit suicide with a gun, and be unintentionally killed with a gun than a resident of Massachusetts. Arizona may be nicer than Massachusetts in many ways (e.g. climate) but it's difficult to understand how Baker can suggest that Arizona is a safer state in terms of gun deaths, or violent deaths.

In general, Baker seems to believe if he can find an anomaly, then the general associations scientists find between guns and death is disproved. It is analogous to his finding that Abel smokes but Cain doesn't, and Cain has heart disease but Abel doesn't, and believing that this proves that smoking does not really cause heart disease. Or believing that the fact that Japanese smoke more than Americans and have less cancer shows not only that cigarettes don't cause cancer, but may well be protective. But such anecdotal evidence shows only what everyone knows, that there are many factors affecting the likelihood of heart disease or cancer, and smoking is only one of those factors. It is not the only factor. Nor is gun availability the only factor affecting homicide or suicide---but the evidence is quite strong that it is one important factor.

What makes Baker's arguments even more questionable is that his claimed anomalies are often specious. One can find states with more guns and a lower homicide rate than Massachusetts (look for very rural states, since virtually all crime, including homicide, is much higher in urban areas), but Arizona is not one of them. There are many other examples. Baker also says that about half the households in Finland contain guns. While a UN report did say that, the information appears to be incorrect. Probably the best source for comparative gun ownership is the International Crime Surveys that found that in 1989 23% of Finnish households contained a gun, in 1992 it was 25%, and in 1996 it was about 26%.

Discussions of firearms on Baker's and many other internet sites, seem primarily to be debates, where each party tries to find evidence to support his already held point of view. These are interesting exercises, but they add little to science, and I am not very interested in them. There seems to be a surprisingly lack of curiosity as to what really is happening in the lives of 300 million American, or the 5-6 billion people on the planet. We can't rely on news to tell us much. We should rely, not on anecdotes, but on good scientific studies, where the goal is to find the truth rather than support for what one already believes.

David Hemenway

Tags

More like this

Eugene Volokh writes: Martin Killias's "International Correlations Between Gun Ownership and Rates of Homicide and Suicide," 148 Can. Med. Assoc. J. 1721, 1723-24 (1993), purported to show that "the proportions and the rates of homicide and suicide committed with a gun as well as the overall rates…
A Harvard School of Public Health Press Release describes a new study by Miller, Hemenway and Azreal: In the first nationally representative study to examine the relationship between survey measures of household firearm ownership and state level rates of homicide, researchers at the Harvard Injury…
circe wrote: In homes with guns the homicide of a household member is three times more likely to occur than in homes without guns. New England Journal 1993;329.1084-1091. David Friedman writes: You may also have noticed that the death rate in hospitals is much higher than in hotels. So if only…
In another post we pointed out that the number one cause of death in people aged 1 to 44 is unintentional injury. But some injuries are intentional, about half directed to other people and half self directed. Among people aged 15 to 44 intentional injury represents the second leading cause of death…

It is incontrovertable that living in a big city significantly increases your risk of becoming the victim of homicide.

First, why is this so? There's the scientific question.

Second, shouldn't we "decrease access" of citizens to big cities? That is, shouldn't we ban big cities? That follows using the logic of those wanting to ban guns.

It doesn't seem to work that way here. Sydney has about 4 million people and a murder rate of 1.5 per 100,000 population, about the same as the rest of the state.

Tim, American big cities are the cancerous result of 1960s social experiments, and may be phenomena that are unique to the USA. Also, I'd be curious to know what the percentages of minorities and illegal immigrants are in Sydney. Blacks, hispanics, and especially illegal immigrants account for far more than their share of violent crimes in American cities.

That aside, I don't understand the use of suicide statistics in arguments about gun control. The victims in those cases are also the perpetrators, unlike homicide victims who had no choice in the matter.

thanks sarah,

yeah, limit my comment to just the USA.

I've still got to post this same response at my site, with comments, but I haven't got time right now to do it justice.

However, as Hemenway noted, I was not critiquing his book, I was critiquing a Harvard Magazine review of his book. My issues were more with author Craig Lambert (any relation?) than with Hemenway (though I did critique what Lambert quoted), since I haven't read Hemenway's book, just Lambert's article. I thought I was quite clear about that in all three parts. (There are three parts, not just the one you link to.)

Why? Does the world end at the US border?

Sarah - Sydney's one of the most multicultural/racial cities you'll come across - with 33% of the population born overseas. This number doesn't include the children and grandchildren of people born overseas. Most of NSW's overseas pop. lives in Sydney. Overall 25% of Australia's pop was born overseas.

Actually, what sarah is refering to is our unique cultural differences. We have a large group of people who have been marginalized by the unintended consiquences of what Sarah calls our social experiments of the 60's. It really has nothing causaly to do with race, just that things fell along those lines for other reasons.

I bet England and other countries have similar problems with their socio-economic groups. I bet if you look closely, many, if not most, criminals are on the dole. The welfare state is tremendously destructive to people's futures.

I don't understand the mortality rate ratio. 47 divided by 6 equals 10? All of the ratios seem off to me. I'd appreciate a clarification if anyone has it.

Brian:

It's the mortality rate per capita, eg 47 people out of 5.154 million vs 6 people out of 6.356 million. Still, I get 9.7 rather than 10.0 for that one; all the others seem about right. Maybe it's a formating thing; 10. (ie, to 2 sig figs, which is the level of accuracy at which all the other numbers are quoted), would be correct, but putting 10.0 implies that the number is correct to 3 sig figs, which it isn't. Either 10. or 9.7 would be ok, 10.0 isn't.

I think "Post hoc, ergo propter hoc" is the correct phrase, isn't it, ben? Relation does not prove causality, especially not when talking in a comments-thread from this 'blog.

If it's true that most criminals are on the dole, then it may be reasonable to conclude that poor people are more likely to commit petty crimes --- after all, they need the money. The dole may not come into it; look at America, where crime is far worse than it is here, and social services are also well below our standards. I assume you'd object if I were to say that welfare reduces the crime rate, however --- so why make the opposite claim yourself?

alright, I retract my statement because I cannot back it up without hogging up 90% of Tim's bandwidth. Even then, my statistics and experts would compete with everyone else's statistics and experts and nobody would be convinced of anything. Totally pointless.

I'll just say this: The problem in the USA, I believe, especially among African Americans, is that the welfare state has unintentionally abrogated the need for a two-parent houshold in black families on the dole. This has even become an acceptable way of life, something young women in these communities plan on. It has also delayed large segments of the black population from integration into mainstream society. I conjecture that both of these things are very bad.

I believe that welfare keeps more people in poverty than would otherwise be poor. I conjecture further that this is a major problem associated with welfare; poverty itself is not the problem.

The truth exists, but when you think you've found it, try to prove it, and good luck :)

Ok, while it is true that, statistically, Massachusetts is generally "safer" than Arizona in terms of gun violence, it is not true that it is Massachusetts' strict gun control laws that make this so. Let me show in one way why this is so.

If MA's strict gun control was the cause of its high "safety" then shouldn't it follow that all states with equally strict gun control should be, more or less, equally safe? Well, brady campaign gives an A- rating on gun control to both MA and California. Let's look at the firearm homicide rates for MA and CA (using link given above for data, and excluding hispanics from the data to simplify things, e.g. white = white and non-hispanic)
Massechusetts(white) 13 total, 0.25 crude rate (black) 38 total, 10.2 crude rate
California(white) 273 total, 1.7 crude rate (black) 524 total, 22.3 crude rate
Hmmm, doesn't seem to be any causal relationship between gun control and firearm homicide between CA and MA, how about that?

Now lets look at my home state, Washington, given a nice fat D- rating by the brady campaign:

Washington(white) 57 total, 1.2 crude rate (black) 22 total, 9.8 crude rate

hmmm, nearly half the firearm homicide rates of "strong gun control" California, but somewhat higher than Mass. Why is this?

So, sorry Dr. Hemenway, MA's "strict gun control laws" are not the cause of MA's relative safety in terms of firearm violence.

Sydney's one of the most multicultural/racial cities you'll come across - with 33% of the population born overseas.

OK, but what is informative is the breakdown of race/nationality of these overseas populations. How well do they integrate with Australians? Does Sydney have a significant immigrant gang problem? This matters.

For example, Asians in the U.S. integrate with whites far better than blacks or hispanics. They also account for comparatively far fewer violent crimes.

Ben, I think you have misunderstood Hemenway's comparison. It was a response to Kevin's comparison, not a proof that MA's gun laws cause the difference. You really need to look at all the states and control for a whole pile of other factors if you want to investigate that question.

I know there are a whole pile of factors. That wasn't my point either.

Mark, you said, "look at America, where crime is far worse than it is here". By what measurement?

Certainly America has far higher gun crime, and we've got higher homicide rates, but other crime? No. And that's a common misconception.

From this International Crime Victimization Survey table, Australia has a considerably higher level of crime than the U.S. does. Some of the data:

Victimisation in the year preceding the survey: percentage victimised once or more (prevalence rates)
Based on eleven crimes standard across sweeps:

Australia1989 26.1
1992 28.6
2000 30.0

USA1989 28.9
1992 26.1
1996 24.2
2000 21.1

Let's look at some specific crimes, like burglary:

Australia 1989 2.2
1996 3.7
2000 3.9

USA1989 3.9
1992 3.6
1996 2.6
2000 1.9

Or robbery:

Australia 1989 0.9
1992 1.3
2000 1.2

USA1989 1.9
1992 1.5
1996 1.3
2000 0.6

Australia does NOT have lower levels of "crime" than the U.S. does. We DO kill each other far more often.

"By what measurement?" Why, by that "common misconception".

You've got me there, Kevin. I was wrong about the comparative crime rates. Thanks for setting me straight...

The other misconception is that gun laws have somehow changed things.

In fact, as far as the records go, Americans have ALWAYS been pretty bloodthirsty, while, for example, Englishmen never have. Our homicide rates, long before anybody passed anything resembling "gun control" in either country, have been in excess of five times higher than theirs. Now, after a 20th Century in which England went from no gun control to the strictest gun control in the free world, their homicide rates have crept up. Ours have oscillated wildly, largely associated with Prohibitions (alcohol and drugs), but always comparitively high. Homicide is the "American Disease" apparently, but guns aren't the cause, just one mechanism. As of 2000 (prior to the recording of Dr. Shipman's multiple homicides, or the 58 Chinese that suffocated in a shipping container)the homicide rate for England & Wales was about 1.6/100k. In the U.S. it was about 5.6. The ratio had dropped to about 3.5:1, mostly because our homicide rates have dropped precipitously, but the fact remains that England's rates have continued to inch up, and a larger percentage of those homicides are committed with firearms - largely handguns - each year.

What did skyrocket in England was violent crime, particularly after the mid 50's when it was made illegal to carry any sort of "offensive weapon" in public. Correlation does not equal causation, but the coincidence is striking. England appears to be getting a handle on it now, maybe, but that depends on how much you trust their data reporting I guess.

However, here we're told that "the number of guns" is the cause of, or at least a major influence on our high gun-crime rates, yet for over a decade our violent crime rates, including homicide, have been declining and all the while we add between two and three million firearms annually to the number already held by private citizens.

It's not as simple as it appears at first glance.

Just to let anybody who cares know, I got permission from Dr. Hemenway to reprint his letter (the one Tim has posted here) and I'll be doing that with a (typically excruciatingly long) reply hopefully this weekend.

I'm sure that somethight else happened in England during the mid-1950's - the importation of American pop culture and the rise of 'Rock and Roll'. Correlation does not equal causation, but the coincidence is striking.

Yeah, we exported it over there, then re-imported it back here. I guess we outsourced our Rock-n-Roll too, call the politicians!

Kevin: I have to admit my own biases against your views. Even so, your 'rebuttal' was underwhelming and even a little embarrassing. Several times you inartfully sought to deflect the argument elsewhere and--at the end--you basically stated mob opinion should and ought to trump hard science.

BTW, comparing crime rates--not gun-related crimes--is essentially meaningless. Perhaps the greatest contributing factor in crime is the economy. OTOH, gun-related crime requires a firearm.

If Kevin were to look at recent crime rates in the US he'd notice an uptick (due to the economy not performing as well as in 2000). Yet, there are certainly more guns in the US now.

Ah, Jade, we meet again!

Nice that you admit your own biases. Thanks. Hmm, "mob opinion should and ought to trump hard science." Well, you can read it that way, I suppose, but what I said was,

The goal of reducing death by gunshot is noble. The path to it is wrong, and I'll fight that path as strongly as I possibly can because it's wrong. It's wrong because it doesn't address the actual underlying causes. It's wrong because it's been proven a failure. And most importantly, it's wrong because it violates the fundamental law of the United States.

Now, what that means is that I believe that Dr. Hemenway (and other gun control enthusiasts) have chosen "reducing the number of guns" as the path to "reducing gun violence" and that they have done so despite "hard science" that indicates that "reducing the number of guns" is at best a waste of time and at worst possibly counter-productive. Further, with the significant exception of the Violence Policy Center, all of the gun-control enthusiasts deny that "reducing the number of guns" is the actual goal, or that "reducing the number of guns" means any kind of bans or confiscations. In short, they lie. Therefore it is necessary for we, "the mob," to ensure that they are unsuccessful at foisting this "solution" on us, which is (though you disagree) in violation of the fundamental law of the land.

All a matter of interpretation, I suppose.

I agree with you re. the economy being the largest contributing factor - AFTER traffic in illegal substances. Yet the fact remains that each and every year we get between two and three million new firearms in the hands of private citizens, and our gun crime rates don't go up. Our gun suicide rates don't go up, and our gun accident rates don't go up. All of which casts a bit of a pall on the idea that "the number of guns" is the problem.

The problem is behavior - or, as I termed it, "culture," and that behavior is concentrated in a small minority of the overall population, a small minority in the gun-owning population, and in a population that will ignore, in the main, any "gun control" laws designed to disarm them. Dr. Hemenway dismisses this idea and treats gun owners as a homogeneous whole. Perhaps because he cannot scientifically quantize behavior?

Kevin: As Dr. Hemenway has pointed out, using facts, the presence of a gun leads to more gun-related crime. You, instead, opt to chalk it up as a byproduct of 'culture.'
This is nonsense; you may as well have said 'fairy pixie magic dust.'

I fear you also conflated gun-related crime with crime. The point is that gun-related crime, by definition and necessity, requires a gun. Crime doesn't necessarily. Therefore, drawing inferences between crime and guns is far more difficult and tenuous than making the same inferences between guns and gun-related crime.

You also unilaterally declare gun control a failure with no evidence whatsoever. The fact is we've never had meaningful gun control in this country. The fact is there hasn't been a time when any felon couldn't purchase the firearm of choice easily. Moreover, there hasn't been a time when any gunowner has had to demonstrate even the slightest degree of competence or proficiency.

"You also unilaterally declare gun control a failure with no evidence whatsoever."

Read Under the Gun: Weapons, Crime, and Violence in America. Or, read the 2003 CDC report that stated that lead Jointogether.org to state "a review of 51 published studies on the effectiveness of gun laws, the Centers for Disease Control and Prevention (CDC) says there is inadequate research to draw any conclusions".

If "gun control" was effective, don't you think at least one of those 51 studies would have found unequivocal proof?

Oh, silly me. Of course not, because:

"The fact is we've never had meaningful gun control in this country."

Uh-huh. The 1934 NFA? The 1968 GCA? The Brady Act? The D.C. gun ban? Chicago's ban? None of those are "meaningful." Check.

"The fact is there hasn't been a time when any felon couldn't purchase the firearm of choice easily."

Excellent point! And there won't be. That's my point. Guns aren't yachts. The market is flooded. The product is simple to manufacture and easy to transport, and it's highly lucrative. All "gun control" seems to do is make the black market for the product more profitable.

It's too late to close the barn door, the livestock's done escaped, and they aren't going to be rounded up again.

Attacking the gun crime problem through the vector of "GUN CONTROL!" is therefore destined to FAILURE. Instead, the problem must be attacked through the BEHAVIOR (excuse me, the "pixie magic dust") of those who will use guns criminally. And surprisingly, that seems to have worked. A decent economy seems to affect that behavior, doesn't it? So does incarceration. There must be other ways, but no one pursues them. Hell, no one contemplates them. Certainly no one researches them. It's easier to try to pass gun-control laws.

"Moreover, there hasn't been a time when any gunowner has had to demonstrate even the slightest degree of competence or proficiency."

And, if I have anything to say about it, there won't be for simple possession. That's the difference between "a right" and "a privilege." With the former, you don't have to petition the government for permission.

I realize this is a difficult concept for you, but it's a crucial one to grasp. The importance of the right to arms is such that it's one of the few the Founders actually enumerated in the Bill of Rights. So of course those twenty-seven words can't mean now what they meant when they were written, right? Let's just violate one of the fundamental laws of the land because it's too costly, and it's too hard to convince the requisite portion of the population to overturn it. Besides, it's not like we aren't violating most of the rest of the Bill of Rights anyway, right?

not to mention that we have SERIOUS drug control. Hell, it's a "war." Has it worked? Why, in the face of that, would you expect that controlling guns from us law-abiders is going to keep guns out of the hands of criminals?

Kevin: As I noted, there's never been meaningful gun control in this country. The evidence is plain: there is no bar for anyone, anywhere in the US, to get the firearm of his or her choice. Further, since there are no standards whatsoever for gun ownership--mentally challenged, substance impaired, blind and physically-challenged people, etc. have complete and unfettered access to guns.

WRT the 2A; it doesn't mean what you think it does. A fact, I'd remind you, every court at every level has asserted in my favor--not yours.

Ben: Illogical argument. But let's entertain it for a moment. Do you believe the drug problem in this country would be worse, the same, or better if we were to make all drugs legal and allow untrained and unscrupulous people to sell them wherever they wished?

JadeGold: all I meant was that full prohibition is a total failure. There are ways of decriminalizing drugs that will end the violence associated with the trade without encouraging widespread indulgence.

Now, you say

there is no bar for anyone, anywhere in the US, to get the firearm of his or her choice

this is simply false. Unless by "firearm of choice" you mean, weapon of choice subject to restrictions, which, shock of shocks, implies a "bar" on ownership. Among the outright bars on ownership, I may not, even though it may be my firearm of choice, legally aquire or own a fully automatic M60 machine gun in the state of Washington (there are, if I recall correctly, 41 states where this is legal, provided you have the federal class III license, but Washigton is not one of them). Then you go on to say

there are no standards whatsoever for gun ownership--mentally challenged, substance impaired, blind and physically-challenged people

which is easily, demonstrably, false. For instance, from the Oklahoma (fairly low on gun control) state firearms law

It is unlawful for any person to knowingly furnish any firearm to any person who is: a felon; a delinquent; under the influence of alcohol or drugs; or is mentally or emotionally incompetent, unbalanced or disturbed.

That would make them have incomplete and fettered access to guns, no? I believe that this is the case for all the other states, and there's probably some federal regulation to that regard too. Washington laws with regard to illegal possesion can be found here.

Frankly, I cannot fathom why you thought what you posted above was true.

Ben: You didn't answer the question; I'll repeat it: Do you believe the drug problem in this country would be worse, the same, or better if we were to make all drugs legal and allow untrained and unscrupulous people to sell them wherever they wished?

Merely saying full prohibition is a total failure doesn't make it so. Not any more than clicking the heels of your ruby red slippers will return you to Kansas.

Ben, have you ever been to a gun show? I have. I'll guarantee you can buy (assuming you have the cash) any weapon in the place--with no questions asked. What's more, if you don't see what you desire, you can fairly easily find someone there who will procure it for you for a price. Again, with no questions asked.

Frankly, to point out a few gun statutes and say this bars (or even significantly mitigates) illegal possession is akin to pointing out the US Tax Codes and claiming nobody cheats on their taxes.

and so how will adding more gun laws change anything if the current ones cannot be enforced?

The drug problem itself? I do not know. However, the problems associated with trafficing would certainly be better.

"Do you believe the drug problem in this country would be worse, the same, or better if we were to make all drugs legal and allow untrained and unscrupulous people to sell them wherever they wished?"

The drug problem would be different,how different I cannot say with any degree of certainty (and neither can you,) but it would probably be considerably worse in terms of users ODing, and better in terms of drug dealers killing each other. For how long however? I, for one, believe things would eventually improve and stabilize, but the interim period would be pretty bad.

Drug abuse is another "pixie-dust" behavior that we keep trying to affect through the wrong vector.

However, had drug prohibition not started, I am convinced that conditions today would have been better than they are.

If you're interested in my opinion on this topic in more detail, read It is Not the Business of Government.

Ben: I asked you what you believed WRT my drug question. Apparently, you now assert you have no beliefs on that matter. Yet, despite being confronted with empirical, scientific evidence WRT guns and gun-related crime, you fully believe the evidence is wrong with no supporting rebuttal or evidence of your own. Amazing.

Er, Jade, I thought I just crushed your demonstrably false statements above. Please defend or retract those comments.

Which empirical, scientific evidence WRT guns and gun-related crime exactly would you like me to rebut with evidence of my own (that I previously attacked without evidence)? And WRT that, at least I'm make it plain when I offer a conjecture that I am not in a position to defend appropriately at the moment, for whatever reason. I think there is a place for conjecture and hypothesis in these sorts of discussions, but they ought to be labeled as such, so that the party puting them forth doesn't look too ridiculus when they get thrown back in their face.

Er, Ben, you might wish to reread this thread. If you want examples of good, old-fashioned, make-stuff-up-as-we-go-along apocrypha--Kevin Baker's attack on Dr. Hemenway is as good as it gets. Of course, Poor Kevin gets a heavy duty dose of reality in the form of Dr. Hemenway's scientific and empirical rebuttal.

Jade, you don't care to address incorrect statements of yours that I pointed out?

I suppose I win.