The most recent mass killing in Orlando poured more fuel on the fire of gun control debate. Omar Mateen’s attack resulted in 49 deaths and additional 53 injuries, making it the worst terrorist attacks on American soil since 9/11. Regardless of the complicated plot threads surrounding the murderous affray – with Mateen apparently being a regular at the gay nightclub and the FBI finding no evidence of direct links between the shooter and ISIS – gun control advocates pointed to this attack as yet another example of the excessive ease with which anyone can purchase a semi-automatic firearm and go on to cause such carnage.

There are counter arguments: opponents of stricter gun control point to the fact that Mateen was a licensed security guard, who had passed the required psychological test, had no criminal record and had easier access than most to weapons. Additionally, his training and familiarity with firearms allowed him to conduct the attack efficiently, unlike some more novice shooters who fail to maximize the carnage by their inability to clear stoppages or simply shoot straight. Opponents also point to the series of attacks in Europe last year and this year, where determined assailants had little difficulty acquiring all manner of firearms despite far more restrictive gun ownership rights.

There is truth on both sides, and I’ve already mentioned some Common Sense policies that could help detect and prevent some avoidable mass killings. Many of these coincide with the proposals of the Obama Administration, but the key plank in gun control platform revolves around reinstatement of the 1994 assault weapons ban[1]. Supporters of the AWP argue that it is made necessary by spree of mass shootings in recent years and the image of a black-clad Adam Lanza or James Holmes is poster child of the effort. Despite assurances that an assault weapons ban is necessary and that it would prove useful, there has been little discussion on just how effective a ban would be.

For one thing, firearm-related homicides in the US have been in long-term decline. After peaking in the mid-1970’s from a combination of growing urbanization, economic stagnation and increase in drug usage, the homicide rate oscillated between 8 and 10 per 100,000 throughout the 1980’s. Then in the mid-1990’s, the homicide rate dropped by almost 50% to around 5.5 per 100,000.  This coincides with the decade of the 1994 Assault Weapons Ban signed into law by President Clinton. The law expired and was not renewed in 2004; but rather than seeing a resurgence in the homicide rate, it actually fell another 30%, approaching 4 homicides per 100,000. This is despite the fact that there was a surge in the sale of semi-automatic rifles after the ban was lifted and another after Barack Obama was elected President in 2008.

The apparent paradox of more assault weapons and less homicides is at least partially explained by the coincidence in the timing of the assault weapons ban with the passing of “three strikes and you’re out” laws in many states. New York, Texas and Washington already had “habitual offenders” statutes in place before the AWP took effect in September 1994. That same year, 13 more states passed similar laws:

• California, Colorado, Connecticut, Indiana, Kansas, Maryland, New Mexico, North Carolina, Virginia, Louisiana, Wisconsin, Tennessee, and Georgia

The following year, another 10 states joined them:

• Arkansas, Florida, Montana, Nevada, New Jersey, North Dakota, Pennsylvania, South Carolina, Utah, and Vermont

Arizona and Massachusetts revised their penal codes in 2006 and 2012 respectively, for a total of 28 states, including the most populous ones. If we compare the performance of those states with a three strikes rule and those states without, normalizing performance from a “year one” in which the statute was enacted, we can measure the impact on the homicide rate over time. Since most states enacted their legislation in either 1994 or 1995, we have 18 years of data for comparison.

The graph shows that there is a significant difference between the two groups of states; those with the habitual offender laws trend consistently lower than their counterparts. At the end of the observation period, the homicide rate had dropped by 11% more in the three strikes states than in those without; and the difference was on average 8% over the entire 18 year period. That is a very large difference, but it is not the entire explanation. The fact remains that the homicide rate dropped by an average 33% even in the states without the “three strikes” statutes.

Nor is it evident that the introduction of “stand your ground laws” in 26 states since 2005 has had much to do with the continued decline in homicides. Florida started the rush[2] in 2005 with a “model” Stand Your Ground law that became the basis for similar legislation in other states.

• In 2006, 13 states followed Florida including Alabama, Alaska, Arizona, Georgia, Indiana, Kansas, Kentucky, Louisiana, Michigan, Mississippi, Oklahoma, South Carolina, and South Dakota;
• Between 2007 and 2016, an additional 9 states joined the club: Tennessee, Texas (2007), West Virginia (2008), Montana (2009), Nevada, New Hampshire, North Carolina, Pennsylvania, Wisconsin (2011), Ohio (2013) and Missouri (2016);
• Utah already had stand your ground provisions in its books since 1994;
• There are another 8 states that have “Castle Doctrine” provisions in their legal codes that are similar, though not as comprehensive, as “Stand your Ground”: California, Idaho, Illinois, New Mexico, North Dakota, Oregon, Virginia and Washington.

By normalizing the trend from the year the legislation was enacted, and comparing against the performance of the non-Stand Your Ground states as well as the Castle Doctrine states from 2006, we can compare the results. Unfortunately for both supporters and detractors of this controversial legislation, there is no evidence in the homicide rates to indicate that “Stand Your Ground” has had any effect. The performance of the three groups is indistinguishable beyond minor fluctuations in the data.

If stricter penal laws are only part of the explanation and stand your ground is irrelevant, the assault weapon ban itself is not any obvious part of the explanation. The number of total homicides continued to decline after the ban expired in 2004 even as the population expanded by approximately 25 million people in the following decade. The proportion of firearm-related homicides remained constant at 67% of the total, while homicides using rifles stayed at 5% of the total[3]. The vast majority of homicides – and gun crimes in general – were committed by people using handguns; almost exclusively semi-automatic pistols.

But what about mass shootings rather than homicides in general? Surely there is evidence here to support a renewable of the assault weapons ban? In fact, there is some more evidence, but it is far from conclusive. Using a database of mass shootings in the US since 1982 compiled by Mother Jones provides a rather mixed view. For one thing, there is no evidence that the assault rifle ban reduced the number of mass shootings: these occurred at a similar rate in the 12 years prior to the ban as during the ban. It is entirely true that mass shootings have increased in frequency considerably in the 12 years since the ban ended, but it is not at all clear that this has anything to do with greater access to semi-automatic rifles.

An analysis of the types of weapons used in each of these mass shootings reveals some interesting data. Prior to and during the assault weapons ban, from 1982 to September 2004, there were 34 mass shootings. Of these, the following type of weapon was carried by the attacker[4]:

• Semi-automatic handgun, 24 incidents or 71% of the total
• Revolver, 12 incidents or 35% of the total
• Rifle (all types), 17 incidents or 50% of the total
• Shotgun (all types), 10 incidents or 29% of the total

Since the end of the ban, from October 2004 to 2016, there have been 49 mass shootings with the following weapon types employed:

• Semi-automatic handgun, 41 incidents or 84% of the total
• Revolver, 8 incidents or 16% of the total
• Rifle (all types), 17 incidents or 35% of the total
• Shotgun (all types), 11 incidents or 22% of the total

Semi-automatic rifles do make up a higher percentage of mass shootings than of homicides in general; but they are not even the weapon of preference in the majority of mass shooting incidents.

Why has this data been ignored or glossed over? Mostly because semi-automatic rifles provide a much easier political image to sell. Few handguns can compete with an “assault weapon” in terms of intimidation; even the name generates fear. No politician is going to ignore such an obvious and useful tool to get their message across.

Assume gun control advocates could pass legislation completely banning ownership of “assault rifles”; assume further that they succeeded in repurchasing or confiscating 100% of such weapons. Even then, they would have reduced the homicide rate by a maximum of 5% and it is questionable that they would have prevented any mass shootings, since most attackers had access to multiple firearms, including semi-automatic pistols. Supporters of such measures would argue that it would be worthwhile if even one life was saved; but that is a terrible basis for making policy decisions. If you are going to pick a fight on gun control, choose the highest impact fight: and that would be the fight against handguns.

Handguns are the weapon of choice for most criminals, for obvious reasons: even a carbine is almost impossible to conceal on your person without a trench coat, which tends to make one very conspicuous when fleeing a crime scene. Most homicides take place at extremely close range as well, where the greater accuracy of a rifle makes no difference whatsoever. And a 9-mm bullet will kill just as surely as a 5.56 mm rifle round; more so in fact, since most pistol ammunition is of the “hollow point” kind, designed to be expansive to maximize stopping power while much 5.56 ammunition is what is known as “ball” or “full metal jacket” and designed to maintain its shape for better penetration.

I am not suggesting a total ban on handgun ownership; but if you want to reduce gun fatalities, it makes sense to target the weapons that are actually causing the deaths, not the ones that simply “look sinister”.

Sources and Notes

[1] Officially the Public Safety and Recreational Firearms Use Protection Act, part of the Violent Crime Control and Law Enforcement Act of 1994

[2] In fact, Utah had enacted “stand your ground” provisions as far back as 1994, but as no one paid attention at the time, Florida gets the credit

[3] There is a problem with the FBI data set in that the percentage of homicides falling under the category of “firearm not reported” has grown from 4% of the total to 16% of the total. This could cause a significant reinterpretation of the data if it were shown that the majority of these homicides involved semi-automatic weapons, but there is no way of determining this from the data source.

[4] The number and percentage sum to more than 100% because more than one weapon could have been carried by the attacker in any incident. This analysis does not consider whether a weapon was used or not, only if it was carried by the attacker.