Updating 2014 Senate Election Predictions

This article updates my 2014 Fall Senate election predictions.

My initial predictions were made March 2, 2014 in this article. I concluded that article, saying, “Democrats could have 46 – 53 seats, but I realistically expect 51-52.”

Those predictions, as covered in the previous article, were roughly consistent with NYT predictions. Now we have Washington Post predictions as well, using newer and in some cases further in-depth data. Furthermore, WaPo leans a bit less far to the left compared to NYT, so they may have a bit more of an honest look from the bias perspective as well.

Despite being a left-leaning organization, Washington Post results are even more favorable towards Republicans than I was. Their model shows an 80% chance that Democrats take 49 or less seats with about a 95% chance they take 51 seats or less.

After obtaining general statistical consistency across races, the next step in more rigorous analysis is to look deeper into each race on a case by case basis. The Washington Post was kind enough to share some of their case-by-case research here. Republicans need 51 seats to take the Senate.

The Washington Post found that Republicans have 30 seats which are not up for reelection and will be retained. Democrats will retain some seats as well, with 36 seats actually being up for election. Of the 36, Republicans must take 21. Washington Post found that 17 seats are likely Republican, 11 are likely Democratic and 8 are competitive or toss up.

Their in-depth look at the 8 competitive races concluded with the following:

  • Louisiana and Montana have about a 75% chance to go Republican.
  • Iowa has about a 65% chance to go R.
  • Michigan has about a 60% chance to go R.
  • Arkansas, 55% R.
  • Alaska, 51% R.
  • Colorado and New Hampshire, about 40% to go R.

In short, Republicans are very likely to take at least 4 of the toss-up seats. Specifically, they are likely to take Louisiana, Montana, Iowa and Michigan.

I think the analysis is reasonable, although I would put Alaska a bit higher and Colorado and New Hampshire a bit lower, but I imagine their data is better than mine.

In conclusion, I modify my original prediction slightly, from this:

Democrats could have 46 – 53 seats, but I realistically expect 51-52.

To this:

Democrats could have 46 – 53 seats, but I realistically expect 48-49.


Leave a Comment