Editor's Note: (Phillip Levine is the Katharine Coman and A. Barton Hepburn Professor of Economics at Wellesley College and a research associate at the National Bureau of Economic Research. Robin McKnight is associate professor of economics at Wellesley College and a research associate at the National Bureau of Economic Research. The views expressed in this commentary are their own.)
(CNN) Despite conventional wisdom, gun sales do not always spike after mass shootings. In fact, as two economic researchers, we observed no spike in sales following this year's two mass shootings -- Las Vegas, the worst in modern US history, and Sutherland Springs, Texas, the fifth worst.
However, following the Sandy Hook school shooting in 2012, there was a significant spike in sales -- 3 million additional guns sold from December 2012 to April 2013, according to our calculations based on FBI criminal background check data. And after the San Bernardino terrorist attack in 2015, there was also a spike -- 1.7 million additional guns sold from December 2015 to February 2016.
So why the discrepancy in sales? Based on our research, going back to 1998, dramatic spikes in sales occur when a mass shooting takes place in a political environment where it is plausible that federal restrictions on firearms sales may result.
When President Barack Obama's administration prioritized reducing gun violence, there was a real fear of gun restrictions among gun owners. At other times, and certainly now under Donald Trump's administration and with a Republican-controlled Congress, there simply isn't the political will to put in place gun control measures.
In conducting our recent research, we used background checks, which are required before a gun is purchased from a licensed dealer, as an indicator of gun sales. While they are not perfect, they are highly correlated with several other indicators of gun sales -- tax revenue from firearm sales and number of guns produced, for instance -- and are widely used as indicators of gun sales.
In an ordinary month, about 2 million firearms are sold in the US. But guns sales show a strong seasonal pattern, with a peak each year in December. Sales also have been trending upward over time. To accurately determine whether gun sales in any period have increased beyond what would be expected, it is important to adjust for these trends and seasonal patterns.
Based on our analysis of background checks over the past two decades, only a handful of events generated increases in gun sales (again, after adjusting the data for trends and seasonal patterns). Most of those increases were relatively minor. For example, we observed small spikes leading up to January 2000 (Y2K -- 300,000 additional guns sold in November and December of 1999) and following the September 11 terrorist attack (400,000 additional guns sold from September through November of 2001).
Barack Obama's election in November 2008 also generated a modest spike (500,000 additional guns from November 2008 through January 2009).
Large spikes in gun sales occur in response to a legitimate risk that some controls on gun access will be implemented, even if those efforts are ultimately unsuccessful. The Sandy Hook school shooting provides the most striking evidence in support of this hypothesis. Following the shooting, Obama delivered a dramatic speech on December 16, 2012, created an interagency gun violence task force on December 19, 2012 and proposed new gun control legislation on January, 16, 2013. The legislation was ultimately voted down on April 17, 2013, by a Republican Congress.
Interest in buying guns, based on Google searches that include the words, "buy gun," also closely tracks these events. These searches jumped up after the first three dates and dropped after the fourth date. Data on gun sales further supports this view. According to the FBI, gun sales on December 19-22 -- the days immediately following Obama's announcement -- were among the top-10 days of gun sales since data collection began, with over 600,000 guns sold on those four days alone.
The shooting in San Bernardino led to similar calls for gun control legislation. Hillary Clinton, who was the leading candidate for president at that time, tweeted, "We must take action to stop gun violence now." President Obama made a plea for a "sense of urgency" about gun violence and, in early January 2016, announced specific executive actions that were aimed at reducing it. Google searches related to purchasing guns jumped after the shooting itself and again after Obama's announcement of executive actions.
Other events, such as the recent tragedies in Las Vegas or Sutherland Springs, engender no such fear of firearms restrictions. President Trump's stated unwillingness to support new gun legislation reduces the fear of restricted access, even following tragic events like these. We see this in the absence of a spike in gun sales and a smaller and shorter duration spike in searches for terms related to buying a gun after these events.
Spikes in gun sales are important, in part because they generate the possibility of greater accidental deaths. Newly purchased guns may be less likely to be stored properly, increasing exposure in those households. Our research shows that, indeed, the spike in gun exposure in the aftermath of the Sandy Hook shooting led to the accidental firearm deaths of an additional 20 children and 40 adults from December 2012 through April 2013.
The irony is that the desire to prevent gun tragedies by introducing gun legislation leads to more tragedies. It is a Catch-22 that we would all benefit from resolving.