What on Earth is this?
Let's get this done.
The vast majority of Americans pic.twitter.com/23ND36tFFm
— Hillary Clinton (@HillaryClinton) May 20, 2016
FiveThirtyEight's Ben Casselman put it succinctly on Twitter: "This isn't remotely how Venn diagrams work." (His tweet, as of writing, has more retweets than the Clinton campaign's.)
He's right. A Venn diagram is meant to show the overlap between two groups of things -- ideally, to scale. One problem with the Clinton diagram, for example, is that "gun owners" in this context is a group that exists entirely within the universe of "Americans." What's more, if the overlap is meant to be "supports universal background checks," the overlap should cover 90 percent of the circle of "Americans."
Part of the reason the Clinton team opted for this sham diagram was probably because figuring out how the "gun owners" group overlaps with "Americans" is itself tricky. The Washington Post's Chris Ingraham looked at the gun issue in October, noting conflicting data from the General Social Survey and Gallup. The former set of data figures that about 32 percent of households have a gun; Gallup's estimate last fall put the number at 43 percent.
But that's households, not people. For 2015, the Census Bureau estimated that there about 124.6 million households in the United States, with an average household size of 2.54 people. Since the surveys asked if a gun was in the household, and since some households have guns owned by more than one member, we will play it safe and use the larger Gallup estimate for gun ownership. In other words, let's assume that 43 percent of people in households own guns -- meaning about 43 percent of Americans. (Just for kicks, we did this the hard way: taking 43 percent of 2.54 people times 124.6 million households.)
The Venn diagram, then, should have looked something like this.
(We didn't use circles, because we didn't feel like doing advanced math to calculate the size of the curved overlap.)
Notice how this shifts what the graph looks like -- and the percentages. More than 90 percent of non-gun-owning Americans support universal background checks -- using the campaign's numbers -- because less than 90 percent of gun owners do.
But should we use the campaign's numbers? Probably not. Or, at least, not without context.
It is true that the vast majority of Americans support universal background checks. After the mass shooting in Oregon (which took place shortly before Ingraham's post above), we looked at support for the issue. In September 2015, Quinnipiac University found that more than 90 percent of Americans supported the idea. But it also found that more than 70 percent of Republicans and a plurality of independents opposed stricter gun control laws.
How do we reconcile those two things? One way is by recognizing that a lot of people seem to think that background checks are already universal -- or close enough. More importantly, though, it shows why Congress has been reticent to act. Passing "universal background checks" might be popular, but opponents of the measures (read: the NRA) simply have to point out to voters that this is a "stricter gun control law" -- and suddenly, more than two-thirds of Republicans will not want to see this pass.
Not really the point here. The point here? That diagram was bad, and the Clinton staffer who made it (Bill?) should feel bad.