Unexpected Early Voting Results: Good tides for Hagan in NC

There were two structural–as opposed to political–reasons to worry about Senator Kay Hagan’s (D.) chances of winning reelection this midterm. The first? It’s a midterm election! This means lower turnout due to less educated voters foregoing the election. The quasi-technical term “less educated voters” usually means young voters and minority voters–the people who just so happen to vote for democrats.

The second? New election legislation in North Carolina has dramatically changed the voting landscape. One consequence is that the first week of early voting was cut off, which means that there are fewer days to use North Carolina’s very popular one-stop voting mode (in 2012, over 40% of voters returned their ballots before Election Day). It seemed unlikely, given this change, that early turnout would be as high as it could be, which would, in yet another way, hurt Hagan’s chances.

So: Bad tidings for Hagan, who also faces a tough challenge from her republican opponent Thom Tillis. 538’s forecast for this race has Hagan winning, but only by a hair, and the vote share is well within the margin of error.

Bad tides, maybe. But the first few days of early voting may tell a very different story. Check out figure 1, which presents the proportion of democrats, republicans, and unaffiliated voters who turned out in both 2010 and 2014. Two points stand out. First, voters in NC are turning out early at a much faster rate this year than in 2010. So much faster, in fact, that, with eight days until Election Day, the proportion of democrats to vote early this year is already the same as the proportion to vote eight days out in 2010. Again, that’s with five fewer days to vote (I made the same point in my previous post).

Now, republicans and unaffiliated voters aren’t too far behind, so clearly voters, in general, aren’t too perturbed by the change in the number of early voting days. But let’s consider what it means that (a) democrats are turning out at a fast rate this year, and (b) that their fast turnout rate is faster than the rate republicans and unaffiliated voters are turning out.


Figure 1

First, it means that the anticipated decline in turnout is not, at least initially, coming to fruition. Whatever the reason for this result (and we’ll get to that soon), it bodes well for Hagan. Higher turnout, most likely, indicates that more democrats are turning out this year, which means more votes for Hagan.

Second, it means that democrats in particular are turning out at particularly high rates this midterm than in past midterms. Note, again in figure 1, that republican turnout was higher throughout the whole early voting period. That hasn’t been the case this year at all.

But what explains this change? Sure, figure 1 appears to bode well for Hagan, but if an explanation for the results is that (for ex.) all the democrats are turning out early in the early voting period, but that the democratic turnout will plateau over the next few days as republican turnout increases, Hagan’s hopes may not be so hot.

I have a theory. To understand it, check out figures 2 and 3. Figure 2 presents the proportion of democrats, republicans, and unaffiliated voters to turn out in 2012 across the whole early voting period. Figure 3 presents the same data for the first few days of early voting in 2010.


Figure 2


ncabs_proportion_inperson_byparty copy 2.oct28

Figure 3

Above all, compare the first few days of figure 2 to all of figure 3. I’m not saying their identical, but the first few days of early voting in 2012 and 2014 look more similar than the first few days of early voting in 2010 and 2014. To me, this suggests that voting may be more consistent with turnout results from the most recent presidential election rather than the most recent midterm. Presidential elections, as I noted above, are much more popular, receiving more attention statewide and bringing more registered voters to the polls.

But this just demands that we ask yet one more question: Why does it seem that 2014 is more like the most recent presidential election than the recent midterm? Check out EVIC over then next few days as I try to find out.