Of the 123 million people eligible to vote in the country's primaries so far, only 18 million have bothered to do so -- 14.6 percent, according to analysis from the Center for the Study of the American Electorate reported by the Associated Press. That's not all registered voters, mind you, but every American who could vote. But didn't.

We were curious, though, if that held up for some of the most closely watched races in the country. The AP is deliberate about noting that campaign spending is soaring even as more than half of primary states saw record low turnout. But that spending is soaring only in select races.

We looked at turnout for nine federal races: Senate primaries in Kentucky, Georgia, North Carolina, Nebraska, Mississippi, and Iowa; House primaries in Idaho-2, Virginia-7; and the California governor's race. Cutting to the chase, turnout was up slightly over the 2010 primary in three races and down significantly in five. (We couldn't determine turnout in Mississippi in 2010 by the time this post was filed, but it seems almost certain that the 2014 run-off between Sen. Thad Cochran and Chris McDaniel topped it.) The states that saw turnout fall saw it fall, on average, by 6.35 percent. Where it rose, it rose by 3.6 percent.

Time for caveats. This is a very small sample size from all of the country's primary races. But we can still learn some interesting things.

For example, there doesn't appear to have been much correlation between how much money was spent in the race and how much turnout changed.


But races that saw an increase in turnout were more likely to be close. This makes sense, of course; more interesting races where votes make more of a difference are more likely to get people to go to the polls. (The dotted line is the trend.)


Perhaps the secret, then, is to have more races be close? Just spitballing here.

Since we had the data, we went ahead and looked at the relationship between the margin of victory and how much was spent. We looked at two scenarios, spending by candidate in the nine races and all spending, including outside money. There wasn't much of a relationship between how much you spent and how big you won.

Then we took out the two outliers -- Virginia-7, where Rep. Eric Cantor vastly outspent David Brat, and the California governor's race, where Jerry Brown romped without spending anything at all. In these, there was more of a relationship between how much was spent and how big the margin of victory.


But again, small sample size. Overall, though, it's hard to be skeptical about the Center for the Study of the American Electorate's findings. What gets people to vote is a close, interesting race. In a lot of places, that simply doesn't happen.