The Moneyball/politics intersection

Republicans kept telling me before the election that low turnout was going to favor them in the 2014 election. I didn’t doubt that, though I wasn’t as willing to predict four incumbents getting unseated.

Once the election happened and I started asking why things turned out the way they did, all heckfire (We’re a family newspaper.) broke loose when I wrote a story that included turnout as a factor. Some Republicans seemed downright offended. I got a nice letter saying I should have mentioned discontent with President Barack Obama, and that is probably correct to some degree. But midterm elections almost always go against the sitting president. The last time one didn’t was 2002 in the post 9/11 flag-waving era that favored President George W. Bush. Before that it was 1998 after Republicans impeached President Bill Clinton.

Those two elections were freakish, because the last mid-term election that didn’t go against the president before 1998 was in 1934. The last second-term mid-term election that favored the president was in 1822, when James Monroe’s Democratic-Republican Party picked up 34 seats in the House while the Federalists lost eight. (The census upped the number of seats in Congress by 26.) Mid-term discontent with the incumbent president is a given, so much so that I likely just neglected to bring it up.

Another letter writer was not so kind, saying I was trying “to rationalize the landslide victory of the Republicans as merely a result of poor turnout.” He also wrote, “Any reasonable person would assume that changes in voter turn out would affect both parties.”

I would agree with the last sentence. It would affect both parties. And, in fact, my story pointed out other possible factors besides turnout, one of which I will reference later. But there is a reason the common perception is that low turnout, the perception that was repeated to me by Republicans before the election, is bad for Democrats. True, it’s not the only one. The mid-term elections are a case in point, because Republican presidents are hurt by them even in low-turnout years. In 2006 Bush lost 30 seats in the House and six in the Senate.

On Facebook I announced I would be diving deeper into the numbers to see if there was evidence to back up the theory that Democrats were more harmed by low turnout than Republicans. Mick Sheldon responded, “I am a numbers person also. I think it’s from liking baseball so much.”

I’m a baseball fan, too. So it gives me great joy to offer you this quote to prepare you for what I have found.

“Know what the difference between hitting .250 and .300 is? It’s 25 hits. Twenty-five hits in 500 at bats is 50 points, okay? There’s six months in a season, that’s about 25 weeks. That means if you get just one extra flare a week, just one, a gorp … you get a groundball, you get a groundball with eyes, you get a dying quail, just one more dying quail a week and you’re in Yankee Stadium.” — Crash Davis, Kevin Costner’s character in ‘Bull Durham’

When I dug into Kitsap County precinct data it showed me there was some evidence that turnout hurt Democrats, but not enough to be the sole factor. A word of warning: I am an amateur statistician. I invited people to tell me a better way to study this and that invitation is still open. First I’ll give you some stats, but after that we’ll refer to an expert on this stuff.

The first thing I tried to do was to characterize precincts by whether they leaned Democrat or Republican in 2012. Where more voters picked Obama and Jay Inslee for governor over Republican Mitt Romney for president and Rob McKenna for governor, those were Democratic precincts. I then compared the turnout numbers in those precincts from 2012 to 2014.

In red precincts, the ones that voted Republican two years ago, the turnout in 2014 was 66.85 percent what it was in 2012. In Democratic precincts the turnout was 65.76 percent.

That’s a difference of 1.09 percentage points, which I’ll agree doesn’t seem that significant. To arrive at just how much a difference it makes requires taking those percentage points and creating a multiplier that will get applied in a couple of ways. I see weaknesses with both ways. But hey, we’re kind of spitballing here.

One way is to apply the Democratic multiplier to both Democratic and Republican numbers in a race in Democratic precincts, then the Republican number to both party numbers in the GOP precincts. There is an easier way, but we’ll get to that in a moment.

Republican Ed Wolfe won his county commissioner election by 1,265 votes, but in the precincts I counted he won by 974. If you multiply using the formula I described earlier, he actually beats Democrat Linda Streissguth in those precincts by 1,316 votes.

The other and equally as imperfect way to do it is to just multiply Democratic totals by the Democratic multiplier and Republican numbers by the Republican multiplier. It doesn’t take a five-tool brainiac to figure out that will benefit Democrats. The question is by how much.
The answer is Wolfe wins by 892 votes instead of 974. The new numbers wouldn’t have made enough of a difference in the legislative or prosecutor races either. Again, turnout is a likely factor, but it’s not the only one.

Mark Smith, professor of political science at the University of Washington, said turnout is one of three factors and at least on a national level it’s clear who was turning out to vote made a difference. But the quality of candidates is huge. He said Republicans fielded good candidates nationally and Kitsap County Republican Party Chairman Chris Tibbs made that case locally.

Secondly, on the national scene the map favored Republicans, Smith said. Democrats won U.S. Senate seats in 2008 in traditionally Republican states. Those seats were likely to swing back to the red six years later. That tide will swing back, he predicted, in 2016 in states where Republicans won in traditionally blue states in 2010.

Where turnout is a factor traditionally is when you start dividing the electorate into different groups. In 2012 voters aged 18-29 made up somewhere around 19 percent of the electorate, according to exit poll data presented by the Washington Post. In 2014 they represented about 12 percent. Those voters favored Democrats over Republicans in congressional races 55 percent to 45 percent, according to New York Times data compiled by Edison Research. Black and Hispanic voters favored Democrats in large numbers in both elections, but their turnout was markedly reduced this year as well.

The Republican-Democrat percentages were reversed for voters aged 45-59, but those voters made up about 42 percent of the overall votes in 2014, compared to about 38 percent in 2012. Smith said the fact that older voters are more likely to vote in mid-terms should not necessarily be seen as a knock on younger voters. “The longer you’ve been participating in politics the more you understand the system,” he said. “When you don’t have as much experience with the political system you’re not as attached.”

Given time those younger voters will eventually buy homes, become less mobile and more stable. Whether they continue to vote Democratic is another question.

Finally, if you look at the New York Times data it is clear that there was a red tide that swept nearly every demographic. Even though the youngest voters favored Democrats by 10 points this year, that margin was 22 points in 2012. Assuming that trend played out locally as well, that would explain to a large degree the departure of four Democratic incumbents. Next election could be a whole new ballgame.

In the end of all this data mining and expert commentary we’re left with the same conclusion. It’s nearly indisputable that turnout was a factor. But it’s equally as unquestionable that the nationwide red tide and the quality of local candidates also played a role.