Catch-22? Mass Shootings and Gun Control Legislation

The Issue:

It has become conventional wisdom that gun sales spike every time an all-too-common mass shooting occurs in the United States. However, while gun sales do spike after certain mass shooting events, they do not seem to respond in the same way to others. Our research shows that the political discussion that sometimes follows these events appears to play a contributing role in higher-than-expected gun sales. Moreover, this greater exposure to guns can have the unintended consequence of increasing the incidence of accidental gun deaths.

Attempts to prevent firearm tragedies by restricting gun access can lead to additional firearm tragedies — if they lead to increased gun exposure through increased sales and increased handling of guns.

The Facts:

• Gun sales do not dramatically increase following most mass shootings, but only after a few of them. In particular, the Sandy Hook school shooting in Newtown, CT, on December 14, 2012 and the San Bernardino terrorist attack on December 2, 2015 are the only two mass shootings that generated large spikes in gun sales, over and above the gun sales you would normally expect (see chart). Those two shootings led to short-term increases in firearm sales of 3 million and 1.7 million guns, respectively, after adjusting for trends over time and the normal seasonal variation that takes place in gun sales. We are able to identify a handful of other events that generated a noticeable, albeit smaller, increase in gun sales, but none of them were mass shootings. They include the period leading up to January 2000 (Y2K – 300,000 additional guns sold), following the September 11th terrorist attack (400,000 guns), and President Obama’s election in November 2008 (500,000 guns). In contrast, other mass shootings that generated large media attention such as the events at Columbine High School in 1999, the Virginia Tech shooting in 2007, the Aurora movie theater shooting in 2012, or the Orlando Pulse nightclub in 2016, to name a few, are not associated with similarly noticeable spikes. Our numbers on gun sales come from background check data that the FBI collects whenever a gun is sold through a licensed dealer. While these data are not perfect measures of gun sales (guns sold at gun shows are not counted, for instance), they are highly correlated with other measures of these sales — such as tax revenue from firearm sales, per capita gun ownership, and the net total of guns manufactured, imported and exported — and are widely used to track the volume of gun sales.
• One notable feature of the increase in gun sales following the Sandy Hook and San Bernardino tragedies is the length of time over which gun sales remained elevated. The duration of the sales response coincided with the duration of the political debates regarding gun control legislation that took place at those times. After Sandy Hook, President Obama delivered a dramatic speech (“these tragedies must end”) on December 16th, 2012, created an inter-agency gun violence task force on December 19th, 2012, and proposed new gun control legislation on January 16th, 2013. Gun sales on December 19th, 20th, 21st and 22nd were all among the top-ten days of gun sales since data collection began, with over 600,000 guns sold on those four days alone. The spike in gun sales dissipated following the April 17th Senate vote that rejected the legislation and effectively ended the discussion of changes to firearms policy. Similarly, a month after the San Bernardino attack, President Obama issued executive actions intended to reduce gun violence on January 4th, 2016, which extended the surge in gun sales.
• A similar pattern is present in Google search terms reflecting interest in purchasing guns around the times of the Sandy Hook and San Bernardino shootings and the gun policy focus that followed. We analyzed the level of online search activity recorded by Google for expressions that include the words “buy gun.” These data represent an index value, where the peak in search activity is defined by the value of 100 and all lower values represent the volume of searches relative to the peak. For example, a value of 50 represents half of the search activity that was observed at the peak. Over relatively short periods of time (around five years), these data are available on a weekly basis, so they allow us to observe interest in gun purchases at a higher frequency than the background check data. The peak in activity occurred immediately following the December 2012 Sandy Hook school shooting. The greater frequency of searches extended through April 2013, when the political debate ended. In these weekly data, we can also see a second spike that occurred following the January 16th proposal of specific gun control legislation. The same pattern occurred after San Bernardino: The initial spike that occurred immediately after the shooting was followed by a second spike that occurred the following month after President Obama issued his executive actions (see chart). Although the Pulse Nightclub shooting in Orlando generated a larger weekly spike in search activity, it quickly dissipated. We also observe a noticeable spike in search activity after the recent shooting in Las Vegas, but it also faded very quickly. Neither of these short-term spikes in search activity translated into a noticeable increase in sales, as noted above. Each red dot in the figure represents another mass shooting in this time period (DC Navy Yard, Charleston, Roseburg, OR, and Sutherland Springs, TX, are not individually labeled). In each case, no meaningful increase in search activity related to purchasing guns occurred. These data make clear that the aftermaths of Sandy Hook and San Bernardino were unique insofar as they each led to a longer duration of interest in purchasing guns and to a substantial spike in gun sales.
• The increase in gun exposure that followed from the Sandy Hook event and the associated political attempts to restrict access to guns, had the unfortunate and unintended consequence of increasing the number of accidental gun deaths, particularly among children. In our recent research we find that the aftermath of the Sandy Hook school shooting led to the accidental firearm deaths of 60 individuals, including 20 children, over a four-month period. That represents the same number of children and considerably more adults than died in the shooting itself. In our study, we document that the call for new gun control legislation contributed to a large increase in gun exposure: Americans showed increased interest in purchasing new guns as well as in handling guns for the period that followed Sandy Hook until the moment that gun control legislation failed to pass. In addition to the spike in gun sales in the five months following the Sandy Hook incident and for Google searches for the term "buy gun", we found a similar increase in search terms for "clean gun". Both the presence of new guns, as well as having existing guns removed from storage, may plausibly increase the risk of accidents. Indeed, we find that there was a noticeable spike in the rate of accidental firearm deaths for children above the expected trend during the five-month window following the Sandy Hook event. The increase in these accidental deaths was larger in those states where the spike in gun sales per capita was larger.