My Voter Turnout Prediction Sucked—But I Have an Excuse

The figures are almost complete, and there are too few precincts still out to really matter; so why did my prediction miss the mark by so much? I blame it on the weather on Election Day.

Let’s look at the totals. In the Republican primary, 1,333,010 voters cast ballots in the gubernatorial contest. The early vote total was 711,704. I predicted that the early vote total would be 730,000, missing by 18,296 votes (approximately 2.6 percent). I predicted that the early vote would constitute 42 percent of the total vote, but the early vote constituted 53.4 percent of the total vote in the Republican primary. That was off by 11.4 percent. Thus, a much larger percentage of Republican voters voted early than I had anticipated, causing my prediction for the total votes cast to be off by 404,990 votes (approximately 30.4 percent), which is horrible. There is no doubt that the weather was a deterrent to some people voting, but there are probably other factors involved as well.

In the Democratic primary, 546,480 voters cast ballots in the gubernatorial contest. The early vote was 304,136. I predicted that the early vote would be 350,000, missing by 45,864 votes (approximately 15 percent). I predicted that the early vote would constitute 45 percent of the total vote, but the early vote constituted 55 percent of the total vote. That was off by 10 percent. Like in the Republican primary, my miscalculation of the early vote as a percentage of the total vote caused me to miss the total vote by 230,520 votes (approximately 42 percent), which is even worse than my prediction of the Republican primary vote total. Again, the weather was a factor.

I had hoped not to embarrass myself with these predictions. Unfortunately, I was not successful: Mea culpa, mea culpa, mea culpa.