America's 1st Freedom, Nov. 2010
By Dave Kopel. More by Kopel on anti-Second Amendment groups.
"Gun Laws Matter."
So says the title of a new report from the Legal Community Against Violence (LCAV), an organization of anti-gun corporate lawyers that was profiled in the March issue of America's 1st Freedom ("California Gun-Ban Groups Shred Your Gun Rights," p. 34).
According to the report, the more stringent the anti-gun laws, the lower the gun death rate. But if LCAV's methods have any validity--which is very questionable--they actually suggest that more gun control goes hand in hand with more robbery.
Let's take a look at LCAV's data: On one side of the ledger, they score states based on a 25-item list of anti-gun laws.
According to LCAV, among the best practices for states to lower gun death rates is to prohibit concealed carry, prohibit open carry, register ammunition and firearms, ban so-called "assault weapons" under the broadest possible definition, ban magazines holding more than 10 rounds and require current owners of such magazines to surrender them, and ban .50-caliber firearms--a gun banners' wish list.
By LCAV criteria, the best states are, in order, California, New Jersey, Massachusetts, Hawaii and Connecticut. The worst states (that is, the most pro-Second Amendment states) are Montana, Arkansas, Maine, Wyoming and Kentucky.
Next, LCAV looks at the 2007 gun death rates for all 50 states. The data are from the Centers for Disease Control. (See www.cdc.gov/injury/wisqars/fatal.html for a very handy website for data queries on all sorts of causes of mortality from 1999 to 2007.)
Then LCAV produces a list showing that many of the states with the most repressive anti-gun laws have the lowest rates of gun deaths. And states with the highest rates tend to have less repressive laws.
LCAV lists the states with the highest and lowest household gun ownership rates, and again finds that states with higher gun ownership rates tend to have higher rates of gun death.
Is there anything wrong with the LCAV report?
Well, not if you read a story about a woman shooting a home invader intending to rape and strangle her, and you think, "Oh, how tragic. Another gun death."
The CDC data counts all deaths from any cause--including justifiable homicide. So in the CDC figures, a shooting caused by an armed robber, by a policeman who shoots an armed robber or by an ordinary citizen who shoots an armed robber all count equally as "firearms mortality."
It stands to reason that states that allow individuals to carry firearms for protection, and to have firearms readily available in the home in case of an emergency, would have higher rates of justifiable firearm homicide. From my point of view, a justifiable firearm homicide is a good thing, because it means that an innocent victim of a violent criminal attacker was able to defend him or herself. But the anti-gun groups have always been opposed to defensive arms, so from their point of view, justifiable homicide is just another bad thing.
Not only does LCAV treat justifiable self-defense as a social evil, but the group also ignores all other anti-crime benefits of firearms. If a gun is used to drive away a robber without firing a shot, or if a robbery is never even attempted because the criminal is afraid that the victim might be armed, that's a great social benefit. But the "Gun Laws Matter" report does not attempt to find out if gun laws matter in reducing crime.
That said, the LCAV report does contain a germ of validity. In the United States, the large majority of firearm deaths are suicides. There's extensive social science evidence showing that reducing the number of firearms will reduce the number of firearm suicides--but will not reduce the overall suicide rate; people simply choose an alternative method. A thorough analysis of the studies can be found in Gary Kleck's books "Targeting Guns" and "Point Blank". Or for a summary of some of the key data, visit www.guncite.com.
In terms of ranking gun control, LCAV methodology is not very good. Let's just presume for a moment that gun control laws are usually quite effective. According to the LCAV gun control scoring system, if a state bans all .50-cal. firearms, the state gets three points. Hypothesizing that the law was completely successful in eliminating all .50-cal. firearms in the state, the law would have no statistically discernable effect on gun crime, because .50-cal. guns are almost never used in crime.
In contrast, the majority of guns used in crime are handguns. Yet if the state requires a license to own a handgun, the state receives only two points from LCAV. If gun control were effective, then a law restricting handgun ownership would be vastly more important than a law about .50-cal. guns.
According to LCAV, a state loses one point if it has enacted a "Firearms Freedom Act" (FFA), by which the state legislature declares guns that are manufactured in the state, and which never leave the state, are not subject to federal gun control laws. But these laws have symbolic thus far had zero effect on federal gun-law enforcement; the federal Bureau of Alcohol, Tobacco, Firearms and Explosives has stated that it does not believe that state legislatures have any power to restrict its enforcement authority. So far, no court has ever given a Firearms Freedom Act any force.
Since FFAs have not led to any reduction in the enforcement of any federal gun law, it makes little sense for LCAV to expect that an FFA would have any effect on gun use or misuse.
Thus, the LCAV gun control scoring system puts too much weight on laws that are aesthetically important to anti-gun advocates rather than on laws that (at least to a believer in gun control) might be expected to have significant real-world benefits.
But for the moment, let's accept the validity of LCAV's scoring system. Even then, when we dig into the data, we find that support for gun control is not nearly so strong as the LCAV claims.
My research intern, Shelby Lane, and I created a spreadsheet that listed all the states with their LCAV rankings for gun control. Next, we acquired the FBI's Uniform Crime Reports data for 2008, and ranked every state in various categories of crime. We also ranked the states according to gun ownership rates and gun death rates. (The LCAV report does not provide the actual LCAV score for a state's gun control severity, so we likewise had to use state rankings for all the data we tested.)
Unlike LCAV, we also included data for the District of Columbia and Puerto Rico, so we had rankings in 10 variables from 52 jurisdictions. (The spreadsheet is available on my website, at davekopel.org/2A/mags/LCAV.htm.) Because the gun control laws of both D.C. and Puerto Rico are much more severe than California's law, we listed them both as "0" on the LCAV gun control severity ranking. (That is, they were the most restrictive jurisdictions.)
Next, we did a statistical test for correlation. Here's how correlation works: imagine that you have a list of 12 basketball players on a high school team, listed in order of height. You also have a list of how many points each player scored during the season. Suppose that the tallest player scored the most points, the second-tallest player scored the second-most points, and so on. Because there is a perfect relationship between height and scoring, the correlation coefficient would be 1.0.
Now suppose that there's no relationship between the two variables--for example, between the players' heights and their grade point average. Then the correlation coefficient would be 0.0.
Finally, suppose that there's an opposite relationship between the two variables. Maybe we test how fast the players can crawl through a narrow tunnel. The tallest player is the slowest crawler, the second-tallest player is the second-slowest crawler and so on. Thus, the correlation coefficient between height and crawling speed is -1.0. The greater the height, the less speedy the crawling.
We tested correlations using Spearman's rank correlation coefficient. Instructions for how to calculate the Spearman coefficient are at www.mnstate.edu/wasson/ed602calccorr.htm.
Table 1 on p. 28 shows the results of testing the Spearman correlation coefficients for the LCAV rankings of gun control. Table 2 shows the results for testing the Spearman coefficients for household gun ownership rates. (Table 2 does not include Puerto Rico, since we did not have data to assess its gun ownership ranking.)
As the tables show, there's one relationship that's much stronger than any other: the more severe the gun controls, the lower the household gun ownership rate. The correlation coefficient is +.74. This is a very strong relationship; the more gun control a state has, the fewer families will own guns.
This makes intuitive sense, and it's important. Sometimes advocates of gun control laws claim that the laws will not harm legitimate gun ownership. But the data show that when you pile law upon law upon law, you make legal gun ownership so difficult that some people give up. It becomes more and more cumbersome for people to try out the shooting sports.
For example, suppose you live in New Jersey and have a friend who is interested in learning about guns. So you take your unloaded .22 rifle over to his house and show him the basics of firearm operation, such as how to switch the safety on and off. Under New Jersey law, both of you have just committed felonies. In New Jersey, sharing a gun, even for a moment, is against the law unless there is paperwork.
In order for the gun safety demonstration to be possible, the friend would have to fill out a Certificate of Eligibility form, and then the owner would have to fill out another Certificate of Eligibility form to get the gun back. Both the gun owner and the friend would have to keep permanent files of the Certificate of Eligibility forms recording the gun "transfer" that took place. As for a handgun in New Jersey, it is completely illegal to take it to a friend's house, even if the friend just looks at the handgun and never touches it.
So it's no wonder that the states with the most extreme gun control laws--such as New Jersey or Massachusetts--have the lowest gun ownership rates.
What about the gun death rate? There was a correlation of +.36 between lower gun death rates and LCAV-favored anti-gun laws. Likewise, there was a correlation of +.48 between lower gun death rates and lower household rates of gun ownership. This is to be expected because, as explained above, reducing gun ownership will reduce suicide by gun, but without changing the total suicide rate.
A more accurate name for the Legal Community Against Violence might be the Legal Community Against Guns. After all, the group does not seem to pay much attention to non-gun violence.
However, if we're interested in the full picture of the costs and benefits of guns, then it makes sense to consider violence rates. For example, guns might increase the robbery rate, since they make the robber more powerful. On the other hand, guns might reduce the robbery rate, since potential victims can use guns to deter or defeat robbers. Which effect is greater?
We found a negative correlation of -0.46 between robbery rates and LCAV anti-gun laws. There was also a negative correlation of -0.55 between robbery and low rates of household gun ownership. In other words, states that had fewer robberies tended to have fewer anti-gun laws, and tended to have higher rates of gun ownership.
For most other types of crime, we found smaller relationships with gun control laws or gun ownership rates.
For murder, aggravated assault, property crime and burglary, the correlation coefficients were between -0.13 and +0.12. These numbers are much closer to 0 than to +1 or -1, so they don't indicate much of a relationship between crime rates and gun ownership or gun laws. The figures do not support the hypothesis of the anti-gun lobbies that restrictive laws make people safer.
Overall, there was a -0.19 correlation between LCAV-favored laws and the total violent crime rate, and a -0.26 correlation between low household gun ownership and violent crime. In other words, the more guns, and the less gun control, the lower the violent crime rate.
However, any results from state-level comparisons should be interpreted with caution. State-level studies, while not necessarily illegitimate, can overlook important differences within states. For example, in the state of Illinois, the city of Chicago has a very low rate of lawful gun ownership, a very high violent crime rate and very restrictive gun control laws. In contrast, downstate Illinois counties have many fewer gun control laws, much higher gun ownership and much less crime. When a study just looks at states, such as LCAV did, it overlooks important intra-state differences.
The FBI, which compiles the Uniform Crime Reports, cautions that mere rankings "provide no insight into the many variables that mold the crime in a particular town, city, county, state, region or other jurisdiction." As the LCAV report acknowledges in an endnote, "There are numerous other factors that may affect gun death rates that are not analyzed in this report such as enforcement of gun laws, urbanization, age and income distribution."
Thus, the most sophisticated studies of gun control effectiveness, such as those by Professors Gary Kleck or Carlisle Moody, take into account many different demographic and other variables. The National Academies of Science and the Centers for Disease Control both conducted meta-studies of the whole body of sophisticated research about gun control. Both studies concluded that there was, at present, no persuasive evidence that gun laws reduced gun misuse or other crime.
In sum, because LCAV's "Gun Laws Matter" consists only of simple rankings of states, the report makes little contribution to understanding the benefits or harms of gun control. If mere rankings of state are valid, then the data indicate that fewer anti-gun laws are associated with lower violent crime, especially robbery.
However, there will be plenty of readers of the LCAV paper who will just take LCAV's claims at face value, and LCAV will help motivate them to support restrictions on Second Amendment rights and to advocate for those restrictions--as voters, as journalists or as government officials.
The impact of LCAV should not be underestimated. LCAV representatives appear on television programs such as PBS's Newshour and are quoted in leading newspapers such as the Los Angeles Times, Chicago Tribune and Boston Globe. Although anti-gun grassroots groups such as the so-called "Million Mom March" barely exist these days, there are plenty of other anti-gun groups like LCAV that continue to wage a broad-front campaign against Second Amendment rights.
Spearman rank correlation coefficient of LCAV state gun control severity ranking with lower rates of:
Total violent crime: -0.19
Murder and non-negligent manslaughter: -0.11
Robbery: -0.46
Aggravated assault: -0.06
Property crime: +0.08
Burglary: +0.12
Gun death: +0.36
Household gun ownership: +0.74
Spearman rank correlation coefficient of lower rates of household gun ownership with lower rates of:
Total violent crime: -0.26
Murder and non-negligent manslaughter: -0.13
Robbery: -0.55
Aggravated assault: -0.08
Property crime: -0.09
Burglary: -0.04
Gun death: +0.48
