The Statistical Case Against the NRA and America’s Relationship with Guns

The most recent school shooting in Santa Fe, Texas, warrants an update of a post I wrote recently about the statistical case against the arguments put forth by the National Rifle Association in favor of virtually unfettered gun access in America.  That this latest school shooting didn’t even make the front page of our San Francisco Chronicle may be a tragic indication of how numb our country has become to mass gun murder.   Although the rhetorical divide today seems paralyzing, having degenerated to a Left-Right shouting match, this is too serious an issue to not continue to try to allow reason, logical argument, and statistics to play a role in the debate over gun ownership in our country:

Although it receives a significant amount of its funding as a corporate lobbyist for the gun manufacturing industry, the stated principles of the NRA are derived from the Second Amendment of the U.S. Constitution dictating that “A well-regulated Militia, being necessary to the security of a free State, the right of the people to keep and bear Arms, shall not be infringed.”  Thus the NRA’s opposition to virtually any form of gun control rests on the argument that unfettered rights to gun ownership is necessary to protect us from tyranny, foreign or domestic.

This is an intriguing principle to stand on, but at least to some extent principles need to be supported by data.  So let’s consider the statistical validity of the following proposition: “The democracy of the United States would be put at risk if we regulated the ownership of handguns, assault rifles, and other firearms commonly used to commit homicides.”

But how do we know what the United States would be like if it had stricter gun laws?  While of course we cannot answer this question exactly, we can create a reasonable counterfactual for what America would look like if it had gun laws similar to other countries that are like it in other ways.  Today this is done in econometrics through the use of a “synthetic control,” in this case a statistical hybrid of other countries that pretty much looks like our country, except without so many guns. I have not carried out this exercise.  But almost certainly what this synthetic control would look like is some combination of other industrialized, upper-income, English-speaking former colonies of Britain.

Based on this kind of reasonable counterfactual, an America with more restrictive laws does not fall apart.  It probably looks a lot like Australia or Canada. Recently, the Economist, not exactly a left-wing publication, recently downgraded the United States from a “full democracy” to a “flawed democracy” (see green map).  In other words, the assault rifles don’t really seem to be helping our democracy much these days.  It is unlikely that assault rifles are responsible for our democracy’s recent downturn either, which is probably due more to the adult-leaders-behaving-as-teenagers problem (see paragraph #1), but the idea that the extraordinary proliferation of handguns and assault rifles in our midst (see blue map) has made our democracy exceptional, of higher quality, or any more robust than the rest of the western world is almost certainly false.

Moreover, what would have to be true for our high rate of gun ownership to make our democracy stronger would be some characteristic about the United States that put it at a democratic disadvantage relative to these reasonable control countries, such that we would have had a far weaker democracy if we didn’t have so many guns, but that gun ownership has made our democracy less disadvantaged than that of Canada or Australia.  It is very hard to identify such a characteristic.  Therefore it is reasonable to conclude that our high rate of gun ownership has not caused our democracy to be stronger than it would have been in the counterfactual.

This confusion of cause and effect could be overlooked if it didn’t have such tragic consequences.  In an economics framework, think of guns as capital.  The purpose of capital, of machines, is to make people more productive. Assault rifles are machines—killing machines. They give people horrifyingly greater productivity at killing other people.  And there is nowhere in the world that people possess as much killing-machine capital than the United States, and that data show that we are exceptionally “productive” at using it.

Firearm related homicides are from 5 to 20 times higher in the U.S. than comparable countries like Australia and Canada—countries that are similar in tradition, history, ethnicity, culture, income, and politics (see chart).  Similar in all except one pertinent respect: gun ownership, which is about 3 to 20 times higher in our country than in these otherwise similar countries.

At this point you may respond, “Now professor, you are confusing correlation with causation.”  Yes, of course this is always the key distinction. The differences in outcomes that we observe (gun homicides) across differences in gun proliferation are produced by two phenomena: the causal effect of gun proliferation, and any outside and separate factors that may be correlated with it.

By comparing the United States to countries like Australia and Canada, we are already controlling for a bunch of other related factors. Because the difference in gun deaths between the United States and its synthetic control countries are so outlandishly huge, and because the proliferation of guns in our country is similarly so much greater, the statistical probability that there is some third variable apart from firearm proliferation that is responsible for this difference in firearm deaths is virtually zero.

After every school shooting, the NRA likes to put the responsibility on the mental health system, as if people in the United States are somehow more prone to horrendous acts of mass murder than people in other countries. But the proportion of mentally ill people is about the same in the U.S. as in these control countries, and it would be crazy (no pun intended) to think that differences in mental health are the main reason for America’s outlier status in death rate by firearms.  The only variable that makes any logical and statistical  is the huge difference in our rate of gun ownership and the concealable and military-grade weapons that our society allows virtually anyone in the general public to possess.

Some may respond by appealing to academic studies that have showed mixed results on the effects of marginal changes to gun laws.  In some cases they seem to work, and in others they don’t.  Unfortunately these studies are famously susceptible to a host of issues related to time trends, crime waves, and a host of other statistically confounding factors. Moreover, it is easy to see how small changes in the legality of certain types of weapons could have little or no effect at all.  The Washington Post estimates that there are 357 million guns in our country, more than one per every American.  With guns so widely available, it is no wonder marginal changes in gun laws have had little effect–their ubiquity in our country simply makes it too easy for criminals to obtain them illegally.  If someone wants to possess a certain type of gun illegally, you have to make them work a little harder for it by making them scarce and expensive as in other countries, not available for a few dollars on the street.   As a result, gun law changes at the margin are likely to have very different effects than more substantial changes.

The only way for the catastrophically high rate of firearm-related killings to begin to decline in our country is to make major inroads in reducing our country’s killing-machine capital.  We can do this and still retain the right for citizens to hunt and protect their homes, but not at the expense of a firearm death-rate that so vastly exceeds that in similar countries with more standard gun laws. If it takes teenagers to finally convince us of this, then let it be so.

Follow AcrossTwoWorlds on Twitter @BruceWydick.

• ← For Black History Month: A Review of Uncle Tom’s Cabin
• The Behavioral Economics of Police Stops →