Election Predictions, October 2014 Edition

2014 Senate   AoSHQ Decision Desk

I have personally avoided writing too much about the 2014 election cycle for a simple reason:  there hasn’t been much to say.

One can go back and read what I wrote in January, and little has changed.  Structurally, this is an election that favors the GOP, with battles being fought on friendly territory.  GOP should, by any reasonable measure, pick up enough seats to take the Senate.  Democrats are facing headwinds because Barack Obama is not popular, and Obamacare still lacked any traction among the populace.  And the GOP was nominating higher quality candidates than in past cycles.

I think most of this remains true.

What has changed? Well, Obamacare is slightly less of an issue today than 9 months ago.  Some liberals have argued that the issue has completely shifted.  I don’t believe that is the case.  What I do believe is that the issue is ‘baked in’; meaning that those people who have made up their minds on the issue have already picked which side they support.  Obamacare issue ads are unlikely to move the electorate at this point.

The bigger issue has been the plummeting of faith in the presidency of Barack Obama.  His approval numbers are around 40%, which is similar to George W. Bush’s numbers in late 2006.  On issues as wide-ranging as the economy, immigration, and foreign policy, Republicans are now favored over Democrats. That is a shift even when compared to earlier this year.

One common refrain has been, “Why haven’t we seen a GOP wave yet then?”.  It is a legitimate question, which actually has legitimate answers.  In 2010, the wave only really began in late September.  Likely, most people simply aren’t paying attention until then.  Furthermore, unlike past years, the number of seats that can potentially switch is much smaller, especially when talking about the House.  Simply put, even if there was a wave, it is hard to move immovable objects.

So where are we, with little more than a month to election day?


Not going to spend a ton of time on this, but worth a few comments.

In my home state of Ohio, Gov. John Kasich is going to shellack the Democrats by approximately 20 points.  A remarkable recovery for a man who was hovering below 50% approval just a year ago.

Wisconsin should once again be close, but Gov. Scott Walker again holds a consistent small lead over his Democrat challenger. This election looks a lot like the last two races, where Walker looks to be in trouble, but pulls it out in the end.

The Florida race between Gov. Rick Scott and Fmr. Gov. Charlie Crist has been back and forth all cycle.  I wouldn’t be a bit surprised if a few thousand votes makes the difference in the end.

Sam Brownback is in serious trouble in Kansas, though he is within the margin of error. If he can convince enough Republicans to give him another chance, he could pull it off. My guess right now is that he loses.

Martha Coakley is once again running a poor campaign in Massachusetts, and her GOP challenger Charlie Baker is taking advantage.  Coakley probably leads, but not by much.  Could be a photo finish.

Gov. Nathan Deal has struggled in Georgia against Democrat Carter, and the polls have shown it. My guess is Deal pulls it out in the end with a last-minute conservative surge helping pull him to the finish line.

Michigan Gov. Rick Snyder has led virtually the entire way against Democrat Mark Schauer, even in the blue state.  That bodes well for Snyder.

Illinois has been a lot of fun to watch.  Gov. Pat Quinn is under investigation for multiple offenses, not to mention corruption charges.  Illinois has one of the worst records economically over the past four years, and the state is still under a mountain of debt.  Even still, Republican has struggled to pull ahead consistently.  This is a tough race to call, another true tossup.

PREDICTION: If forced to, would predict Democrats to pick up 2-3 Governors seats overall, with the at-risk GOP seats most likely PA, ME, and KS. If interested in each individual race, Larry Sabato‘s run down is excellent.


Easiest prediction:  The GOP will hold the House of Representatives.

Considering the generic ballot, and the structural realities, it is frankly impossible for Republicans to have a huge wave.  Why is that?  Because we are basically in a scenario where we are living with the previous wave, the 2010 midterm election.  That election basically showed what a realistic high water mark is for the GOP. In 2012, the GOP lost 8 House seats.  The most likely result is the GOP wins 5-8 seats in this cycle, basically reaching the high water 2010 mark again.

PREDICTION:  Gain of 5-8 House seats.


All the real fun is with the Senate.

The Senate prediction models (538, NY Times Upshot, Washington PostRealclearpolitics, Huffington Post, Wang, Larry Sabato, and the new AoSHQDD) have been all over the board.  I think one take away?  If you see a model having huge swings, it is best to ignore it until right before the election, because its predictive value is very, very low.

In the middle of September, there appeared to be a suddens shift toward the Democrats, with several models showing the likelihood of Democrats holding the chamber to be better than even. That lasted for about 48 hours.  The reality is nothing fundamentally changed, but poll variables were shifting the dynamics.

Models such as those by Professor Wang deviated widely, which led 538’s Nate Silver to take pot shots at him. For example, on September 25th, Wang’s model had Begich as a 99% favorite in Alaska. The next day, it gave him a 23% chance.  Such deviations are signs of a poor model.

In any case, the larger issue is nothing has really changed that much, but there have been some state by state variability.  Here are the races I think that are most important.  Please note that I am no longer even discussing Montana, West Virginia, and South Dakota, which are likely locks for the GOP.  I also think Kentucky and Georgia have basically trended away from the Democrats.  Barring any ‘black swan’ even, Republicans should hold those seats.  Same with longshot seats in Virginia and Oregon, where Democrats have largely locked up re-election.

The remaining?


Arkansas is a state the GOP must win to take the Senate.  Tom Cotton has run an up and down campaign against Senator Mark Pryor.  Pryor, on the other hand, has not run a perfect re-election campaign.  One steady truth though: Cotton has held a small but significant lead against Pryor since early summer, currently leading by 3.6 points.



This is a race that is pretty unique, because it has trended away from the GOP.  Sen. Kay Hagan has run a brilliant campaign, largely focused not on her record but the education record of opponent Thom Tillis.  That, along with her significant monetary advantage has allowed her strategy to prove successful.  Hagan has opened up a significant 3.5% lead since early September.  However, unlike other races, like Arkansas, that lead has only been for a few weeks, so is less certain.  But it is significant, and right now Ms. Hagan has the edge.

RATING: Leans Democrat.


Of course because of this state’s strange election rules, it is highly likely this goes to a run off in December, barring a single candidate getting 50% in November, which is highly unlikely.  But in any case, Republican Bill Cassidy has had a solid, steady lead or incumbent Mary Landrieu.  The lead is 5.1% today, and has been similar for months.


4.  Alaska

Alaska is notoriously hard to poll, because of its sparse population.  But there has been some decent polling there in recent weeks, and the news is not good for Democrats.  Dan Sullivan has opened a small, but persistent, lead over Democrat Senator Mark Begich.


5. Iowa

Iowa was considered the ‘firewall’ for Senate Democrats’ hopes to hold the Senate, along with Colorado (see below). Bruce Braley was a unanimous choice as a strong candidate to hold the seat.  However, conservative Joni Ernst has run a strong campaign, attacking Braley on both policy and personal issues.  Surprisingly, Ernst appears to have the tiniest amount of momentum at this point.  Still too close to call.

RATING: Tossup.

6.  Colorado.

Along with Iowa, this was considered the Democrat firewall to hold the Senate.  Cory Gardner has disrupted that strategy.  Gardner is a solid candidate, who has run a clean campaign against incumbent Sen. Mark Udall. Udall has led for most of the year, but recently Gardner has taken a slight lead.  Udall has had several hiccups of late, but he still has a lot of money and a strong ground game. This will grind out until election day.

RATING:  Tossup.

7.  New Hampshire

Honestly…I did not think we would be talking about New Hampshire at this point.  Sen.  Jeanne Shaheen is a relatively popular Senator, with no major scandals.  Fmr. Sen. Scott Brown is a relative usurper, moving from Massachusetts just earlier this year. But key issues, including foreign policy, have made this race competitive. Shaheen still holds a steady lead though, and I presume she will pull it out.

RATING:  Likely Democrat.

8.  Michigan

Of all the races for the GOP, this is by far the most disappointing.  I openly advocated for Terri Lynn Land, but she has run a horrendous campaign, where her messaging has been off, her campaigning has been lackadaisical, and she has allowed herself to become mired in silly controversies time and again.   Unlike every other Republican on this list, she has actually outspent her opponent, to little or no avail. Gary Peters is not a good candidate, but in a blue state, you don’t have to be a good Democrat candidate to beat a mediocre Republican.

RATING:  Likely Democrat.

9.  Kansas

This is a race nobody can honestly predict.  All the fundamentals should mean Sen. Pat Roberts wins re-election.  Independent (Democrat?) Greg Orman leads in several polls, but hasn’t been challenged at all.  Now, I guess the GOP could fail miserably and not call Orman out to task…but even I find that difficult to believe.  The polls are not great in this race, but like Sean Trende has said on Twitter, until I see solid evidence, you have to bet on Roberts.

RATING:  Likely Republican (with little or no evidence to prove either way).

PREDICTION: The Senate is still too close to call, as every prognosticator has suggested.  Charlie Cook this week suggested Republicans have a 60% chance of taking the Senate majority…and I believe that is the most forceful prediction I have seen recently.  The Senate is on a razor’s edge.

Right now, I think Republicans have a significant edge in KY and GA, as I stated above.  I think they lead by a small amount in AR and LA.  I think that Begich is in trouble in AK as well.

If Roberts holds in Kansas, that would give the Senate to the Republicans.  If not, it gives you a 50/50 tie, and gives Biden the Senate for Democrats with the VP tiebreaker.

However, even if Roberts loses, there is a better than coin flips chance that Republicans take either CO or IA.  I think both are literal tossups, and there is at least a 50% chance of taking one of those seats for the GOP.  I also don’t believe NC is lost to the Republicans yet, though I would bet on Hagan if forced to at this point.

What it comes down to is, the worst case scenario for the GOP is a pickup of 5 seats, meaning they fall short of taking the majority because of Joe Biden.  The best case scenario is a 8 seat pick up.

In short…I basically agree with the prognosticators.  Whether Nate Silver, Huffington Post, Charlie Cook…they give the GOP a slightly better than 50% chance of taking the Senate.  I have said this actually since January, and nothing has fundamentally changed.  Or to simplify matters, presuming my assumptions above, Republicans would need to win 2 of the four races in IA, CO, AK and Kansas.  If I were the GOP, I would be relatively happy with that coin flip.

One thought on “Election Predictions, October 2014 Edition

  • October 3, 2014 at 3:19 am

    Excellent run down. I see nothing to disagree with. I do have a clarification, however. Rick Scott of Florida might become a victim of his own success. While every other state went through the economic meltdown, Florida kept cruising along at a slower, but still stable clip.Of course our GOP controlled legislature didn’t/doesn’t hurt. That stability allows people to look around and be concerned about more than how to put food on the table, so Scott’s having trouble with his former support of Common Core and other smaller, but still important issues. This gives the LP, Wyllie, a boost in some cases.

    We also have the marijuana bill (Amendment 2) on the ballot so that’s going to pull more away from Scott, who is strongly against Amendment 2. I think that block would already support Wyllie, anyway, so not sure how that’s going to play out. We passed Charlotte’s Web recently, a very limited bill on medical marijuana, but we both know it’s not about the medical.
    Again, great write up.

Comments are closed.