Data in Policy Debate: Mass Shootings

“More mass shootings occur in the United States than any other country.”

Depending on how you feel about the gun control issue, you may have one or more of a few reactions to that:

• “The US has a major gun violence problem.”

• “The US has a major mental illness problem.”

• “That’s a bad measurement; the US is the third-largest country in the world.”

• “The number is probably so low it’s not important.”

For policy concerning mass shootings to be a true priority for us as Americans, we would need to believe two things:

1. Mass shootings are a large enough problem (compared to others) to warrant devoted resources to fixing

2. It is fairly likely that these resources would have a significant impact on the problem

We don’t always approach highly emotional policy questions with this cool rigor; let’s use it as an exercise.

One way of looking at the problem would be to compare the incidence of mass shootings in the US to other “peer” countries. If there was a large gap, it might suggest that the US could adopt policies like those other countries in order to significantly reduce the frequency of mass shootings. If there is a small gap, it might suggest that peer countries would not necessarily have an effective model for the US to follow.

We ought to control for population size. If we look at incidents and victims of mass shootings per capita, we see that the US doesn’t fall in the top five.

Take a moment and ask yourself about your reaction: are you feeling acceptance and confidence? Skepticism and doubt? Frustration?

Depending on what “camp” we fall into, we’re more likely to greet certain data with rapid acceptance, and other data with immediate rejection. It’s a very human bias.

We can look at the data in other ways if we’re curious (it helps to understand statistics). Looking at how many mass shootings happen in those top 5 countries, we see that it is always 1 or 2, compared to the US’ 38. The US is more than 38x larger than these countries, which is why the incidents per capita figure drops. But, as the associated article explains, one incident is a tiny sample set, and it’s hard to extrapolate that to the future. Maybe Finland will go another 20 years without a single mass shooting incident? One data point does not create a trend line.

So if we compare the shootings per capita in the US to the entire rest of the OECD put together, we see that the US has 4-5x as many mass shootings per capita as the rest of the OECD.

But this might not be fair, either: there may be some countries in the OECD that have, indeed, a higher likelihood of a mass shooting than the United States; many countries will be above-average. Such is the nature of averages.

Ultimately, it’s very hard to compare the frequency of mass shootings in the US versus other countries because, no matter the country, mass shootings are incredibly rare. With such a small data set, it’s easy to frame the data in whatever way fits the story you want to be true. (Though this article makes a case using simulations that the US probably does indeed have a higher incidence of mass shootings.)

To get a sense of how rare mass shootings are, let’s compare the number of murder victims in the US over the past 5 years from mass shootings to other forms of violence:

Source: disastercenter.com, using FBI statistics

This should be some worthwhile input for our first question about priorities. Should solving mass shootings in the US be a priority? Does something about mass shootings make them a greater problem to focus on than other murders in the US?