Before we get to talking about the Senate, I want you to consider the following possibility. Imagine that tomorrow night the following incumbent Republican governors will lose their bids for reelection:
1. Paul LePage of Maine
2. Tom Corbett of Pennsylvania
3. Rick Snyder of Michigan
4. Scott Walker of Wisconsin
5. Sam Brownback of Kansas
6. Rick Scott of Florida.
7. Sean Parnell of Alaska
I don’t think this far-fetched in the least, and on a good night Governor Nathan Deal of Georgia might lose, too.
Now, don’t get me wrong. There are Democrats running for governor who might lose tomorrow, too, including in deep blue Connecticut and Massachusetts. But before we start talking about an epic beat down for the Democratic Party, we should keep these governors’ races in mind. A governor has a lot more power than any individual senator. And losing control of big states like Florida, Pennsylvania, Michigan, Wisconsin and (perhaps) Georgia would be a very big deal.
So, keep your eyes on these races because the Republicans are at risk of suffering a significant bloodbath tomorrow even if things go as expected in the Senate races.
There are two things that are most important to keep in mind about these Senate races. The first is that pretty much every seriously contested race in the country is polling at or near the margin of error, meaning that even the night before the election it is very hard to be certain about who will win almost any of these contests. The second is that the polls show a small but quite decisive advantage to the Republican Party, and the late momentum has been almost all in the GOP’s direction.
There’s a background to these elections that I am not really going to get into, but the basic facts are that the third of the Senate up for election tomorrow is the only third that comes from states that Mitt Romney carried in his 2012 presidential bid, and this is a midterm election with low turn-out which favors the Republicans, and this is the sixth-year of a Democratic president’s term, which also favors Republicans. Democratic senators running for reelection in Romney states are at a very severe disadvantage, and the ones running in blue states aren’t likely to do as well as they did in 2008. This is baked into the cake, and it’s a major reason why the outlook for the Democrats is so bleak.
With all that said, let’s get started. The Democrats and their affiliated independents (Bernie Sanders of Vermont and Angus King of Maine) currently control 55 senate seats, while Mitch McConnell’s Republicans control 45. Because Joe Biden, the vice president, breaks ties in the Senate, the GOP must net six Senate seats in order to win control of the upper chamber.
I’m going to go in alphabetical order and ignore races where the outcome isn’t in the remotest doubt.
Alaska
For simplicity, I am going to use the RealClearPolitics average of polls, which may not be weighted in a way most favorable to Democrats. In Alaska, the Republican challenger is favored by 2.4%. The Begich campaign has an outstanding ground game that could erase Dan Sullivan’s advantage and I doubt we’ll be certain about the winner of this race for a week or more. It could come down to less than a thousand votes separating them, and I consider this a coin-flip. But I am an optimist, and I think the revolt against the Republican governor and the unstable behavior of Rep. Don Young indicate that the state’s GOP is in a bit of disarray.
PICK: Mark Begich (D)- Democratic Hold
Arkansas
I expected Mark Pryor to hold this seat despite all the disadvantages he needs to overcome, but the RealClearPolitics average of polls now shows Tom Cotton with a 7.1% lead. That seems insurmountable, but the focus down there seems to be on whether or not the Cotton campaign has been advertising on gay hook-up sites and everyone says that Pryor’s turnout machine is very impressive. There’s a hot governor’s race here, too, and the Democrats are desperate to hold on. The problem is that the polling momentum has swung dramatically Cotton’s way in the last days of the campaign. I think Pryor could still pull off a shocker, but that is what it would be.
PICK: Tom Cotton (R)- Republican Pick-Up
Colorado
I never expected Mark Udall to be in jeopardy in this race. That he looks unlikely to win at this point is the biggest shock of the campaign for me. The RealClearPolitics average of polls has Cory Gardner with a narrow 2.5% advantage, but he’s been leading in every recent poll. This could be another coin-flip with the potential for a recount. But I just have a bad feeling about this race. Democrats have been arguing with the polls here for two months, and I don’t think that is a good sign. Perhaps the pollsters are undercounting Latinos, overly discounting the impact of voting by mail, and underestimating the Democrats’ ground game. A 2.5% polling advantage is not a commanding lead. But, in this case, I am betting that the coin-flip goes against us.
PICK: Cory Gardner (R)- Republican Pick-Up
Georgia
Incumbent Republican Senator Saxby Chambliss is retiring and he will be replaced by either professional outsourcer David Perdue or political legacy Michelle Nunn. To avoid a January run-off, the winner needs to win an outright majority, which appears to be a challenge for either candidate considering the closeness of the polls and that there is a Libertarian candidate pulling down three percent of the vote. But this is another contest where the late movement is in the Republicans’ direction. I think there is a very good chance that Michelle Nunn will win a plurality of the vote tomorrow, but that won’t win her the seat. I am not going to make any predictions about the outcome of runoff elections because it is too early to do that, but Democrats have struggled historically in runoff elections in Georgia.
PICK: Michelle Nunn (D)- Runoff
Iowa
This is another seat that I thought would be safe for the Democrats but it hasn’t turned out that way. Republican challenger Joni Ernst has been consistently leading by the narrowest of margins and has a 1.9% advantage in the average of polls. What is going to pull her over the top are the coattails of the popular incumbent governor, Terry Branstad. Bruce Braley is a better person than a candidate, and it’s upsetting that he is headed to a loss against a certifiable lunatic, but that’s what it looks like to me. If Joni Ernst doesn’t win this race, things may not look so bad for the Democrats.
PICK: Joni Ernst (R)- Republican Pick-Up
Kansas
I’ve written a lot about this race, and it really comes down to whether moderate Republicans are going to follow through on a revolt against the conservative establishment of their state. I think they will, and I think both Governor Sam Brownback and Senator Pat Roberts will go down to defeat. I also think Greg Orman made a deal with the Democrats that he would caucus with them if their candidate dropped out of the race. But the average of polls shows only a 0.8% advantage for Orman at this point, so nothing is certain.
PICK: Greg Orman (I)- Democratic Pick-Up
Kentucky
This one hurts. This election has been hotly-contested all year, but it seems to be slipping away from Alison Lundegan Grimes at the end. McConnell has a 7.2% advantage in the average of polls, which is just too big to overcome in a state where the Democrats do not have a great ground game.
PICK: Mitch McConnell (R)- Republican Hold
Louisiana
I think Mary Landrieu will win a plurality of the vote tomorrow, but not a majority, and she will head to a runoff election in which she will not be favored. She’s the queen of winning close elections, though, so don’t count her out.
PICK: Mary Landrieu (D)- Runoff
New Hampshire
This race has become a nail-biter with Jeanne Shaheen holding a paltry 0.8% advantage in the average of polls. Honestly, Scott Brown is a better politician, but he’ll have to pull an upset to win. We’ll know the results here early, and if Shaheen loses you should just go to sleep and hibernate for the next year or so because it’s going to a bloodbath.
PICK: Jeanne Shaheen (D)- Democratic Hold
North Carolina
The polls keep showing Kay Hagan narrowly ahead but I am not so sure. Like with Shaheen, if Hagan goes down then we’ll be in for a world of pain. I think Thom Tillis is just awful enough to lose, and that is what I am predicting.
PICK: Kay Hagan (D)- Democratic Hold
The other races of interest are taking place in Montana, South Dakota, and West Virginia where I expect Republican pick-ups and in Michigan where I expect a Democratic hold.
So, where do these election predictions put us?
We have Republican pick-ups in:
1. Arkansas
2. Colorado
3. Iowa
4. Montana
5. South Dakota
6. West Virginia
We have a Democratic pick-up in Kansas.
And we have runoffs in Georgia and Louisiana where the Democrats will not be favored.
If I am right about these predictions, here is how things will look prior to the runoffs and after all the recounts and challenges are done.
The Republicans will have 50 seats, and the Democrats will have 47 seats. Greg Orman of Kansas will presumably caucus with the Democrats, but that is not assured. Control of the Senate will be the Republicans’ unless Nunn and Landrieu win their runoff elections and Orman agrees to caucus with the Democrats.
In that case, the split would be 50-50 with three independents caucusing with the Dems and Joe Biden breaking the tie.
It matters hugely what happens in Alaska, Colorado, and Iowa. And we can’t lose New Hampshire or North Carolina, either. If I am wrong about Udall and Braley losing and right about Begich winning, then things look quite a lot better.
But my prediction is that the GOP will reach 50 seats with a very good shot at reaching 51 or 52 after the runoffs are completed.
Thank God that the governors’ races look more promising.
So, when states elected a bunch of RWNJ governors, they threw out a few Democratic Senate seats as well, but now that they have figured out that these RWNJ governors suck, they’re going to toss out even more Democratic Senators?
At the end of the day in statewide and presidential elections, Iowa is generally, if just barely, sane. (GWB as an incumbent won in 2004 by just over 0.69%.) Braley’s not a great candidate, but Ernst is a RWNJ and this is the beloved and liberal Tom Harkin’s seat.
The all mail-in ballots in CO and DEM ground game will save Udall.
It doesn’t look good for Pryor, but the X factor in that race is the minimum wage increase that’s on the ballot. That and the flexibility in the PPACA that allowed AR flexibility in expanding Medicaid is what he should have run on. Clinton stumping for him was at best useless. (They’re going to elect one of the House impeachment guys governor.)
Landrieu and Nunn will have a few more weeks to rip off the “moderate” masks of their opponents. And it’s still possible that Carter will have a second shot at Deal.
Landrieu might want to consider working this into her campaign::
<
On the issue of Clinton in all of these southern states it seems as if none of the old Clinton magic remains. I thought they had the magic touch to bring everything back into the fold. And where’s Hillary? Has she campaigned at all? I know there’s going to be a theme coming out of this election of the Clintons coming to save the Democratic Party and it kind of ticks me off.
There is no Clinton magic — never has been except in the minds of the Clintons and their diehard fans. To maintain that fiction, they had to claim that Gore is stupid and/or a liar. That had only Gore used Clinton on the stump in 2000, he would have won. Except early in the campaign wherever Clinton appeared on his behalf, he polls numbers in that area dropped significantly. Also, Clinton was out there for Gore in AR and we know how well that worked out.
Hillary has been out there in a few places of late — such as LA and NH. Bill’s been many places this — and Grimes’s poll numbers flatlined precisely when Clinton made his first KY appearance. Coincidence? Who knows, but it clearly didn’t help her.
One bit of evidence the stat polling guys use to support their claim that polling bias is random is that McAuliffe was projected to win by 6 in VA and only won by 2. But then they overlook the fact that Warner and Kaine improved their polling averages. Kaine by 4.1%. And that was against a former governor and senator that only lost his senate seat in 2006 by a very narrow margin. Whereas, McAuliffe was running against a RWNJ and the sitting GOP governor was under investigation for corruption.
I believe that in Colorado the Clintons, Michelle Obama, Elizabeth Warren, and others have been helpful. Not sure how wise it was for them to campaign in Kentucky, though I don’t know Kentucky very well.
Also, I recall Bill Clinton campaigning a lot for John Kerry and we know how that race turned out.
How many of those states were +10 or better for Romney? 4?
Colorado and Iowa were stronger than the national average for Obama in 2008 and 2012. MT re-elected a DEM governor in 2008 and senator in 2010 and elected a DEM governor in 2012. The Democratic Party as the federal and state level practically conceded MT, SD, and WV before the election cycle began — because, well that required some heavy lifting. Don’t know why having won so easily in 2008 that Pryor has been flat footed with the AR electorate since then.
Pryor ran unopposed in 2008 because the GOP assumed Hillary would be the nominee and winning in Arkansas would be hopeless in the Senate election.
??? Did you mean that with Hillary as the nominee, a GOP Senate candidate in AR wouldn’t have a chance? No chance that there wasn’t some backroom dealing between Bill and others in AR that the state would only get goodies from a second Clinton POTUS if they sent a DEM to the Senate?
Geez, that’s depressing. Some pretty slim reeds of hope.
I’m ready for your epic rant now.
thanks for this, BooMan
Yet more motivation as I head out to vote. Thanks, BooMan.
I’ve worried that the Democrats have had a lot of success fundraising off of fear (“Majority leader Mitch McConnell will steal your baby!”) but that this has depressed Democratic enthusiasm.
I think Republicans vote off fear, but Democrats vote off hope.
I “hope” I’m wrong.
I am particularly interested in the races in which there has been substantial and sustained GOTV efforts and strategy. Three states are signals in this:
North Carolina D+1 69% probability of D win, range: D+6 to R+4
Georgia R+2 70% of R win, range: D+4 to R+6
South Dakota R+12 >99% of R win, range: R+5 to R+18
What I will be watching for is how far, if at all, these move from where the last reading from FiveThirtyEight has them spotted.
The campaigns in these three states in particular have had the goal of expanding the population of likely voters who actually show up and vote for the Democratic candidate. I am curious as to whether comparison between polled estimate and actual vote can separate the noise from the effect of human action enough to determine whether intensive GOTV is an effective strategy going forward.
FiveThirtyEight is a good model to use for comparison because it uses extensive historical data to drive a simulation model that incorporates the latest polling for each race. And that simulation model does (a large number?) of multiple runs that it then displays as a box diagram of estimated value and 90% probability range. Everything within that 90% probability range is noise. A result outside that range requires an explanation. In the case of these three races, the first candidate for explanation is effective GOTV efforts, the most dramatic of which would be a surprise Weiland win in South Dakota.
The other race I’m watching is the Kentucky race. FiveThirtyEight has that as R+6, 98% probability, range: R+1 to R+10. The McConnell campaign over the weekend released some very blatant voter suppression literature. I read that as the fact that McConnell’s GOTV campaign had exhausted all possible additional voters. I’m not sure whether that will show up in a comparison of the results between the FiveThirtyEight model and the actual vote, but it might.
I have a problem with the social physics metaphors that used often without thought in the political polling field commentary. One that really bugs me is “polling momentum”, what the Kennedy campaign called way back when the “Big Mo”. To be useful scientific concepts must be explanatory even as metaphor. This one isn’t. It says nothing about political strategy and is subject by campaigns as mistaking noise for signals of voting behavior. The fact that a poll trends in the last moments of the campaign means that public opinion is moving off the fence. And media can shape that even if it does not translate into actual votes. There are probably some other behavioral effects being presumed with “polling momentum”.
A good discussion of the practical use of polling data would be appreciated. There are a lot of folks here who have an appreciation of the gifts and limitations of polls. I think it would be an interesting and helpful thread.
Yep the governor data is encouraging even if the paired legislature data is not. John Oliver did a good look at state legislatures and ALEC. We are going to see some more attempts to make an unrepealable mess in several more legislatures if the current legislative indications are correct. By far the most interesting race for a Democratic hold will be Hickenlooper in Colorado and also his legislature.
Back in High School when they still had Civics courses that weren’t propaganda, we were told about “bandwagon effect”, the phenomenon that, even though there is a secret ballot, people want to be on the winning side. It’s some sort of herd instinct, to run with the herd. I never could understand it and I’m sure it never affected Marie2, either and probably not you. But we are independent thinkers swimming against a stream. The effect seems real.
It’s real — as is defaulting to the incumbent when the challenger hasn’t made a good enough case. Nothing mysterious about it because enthusiasm is contagious for humans. Particularly those not consistently involved.
Reminds me of two lovely men that I once worked worth. Both SF Catholic boys — one Irish and one Italian — and 49er season ticket holders as soon as they got their first jobs. Decades of supporting a losing team — until 1981. One was there in the Detroit stadium to watch his team go all the way. The crowds welcoming back their champions were a multiple of season ticket holders and irregular game attendees.
And “throw the bums out” without looking hard at the other bums.
but a couple of things:
1. There is significant confusion about the N in the calculation of these odds. In one calculation there have been thousands of polls since 2004. So the N is very large. But in another there have only been 5 elections. What is clear from the data is the miss tends to happen in the same direction. In 2012 and 2010 it missed Dem strength, in 2008 is was about a wash, in ’06 it understated Dem strength and on ’04 it arguably understated GOP strength. But if you go farther back, you find significant shifts late in at least 2 to 3 points in at least have of the elections. A couple of examples:
1980 – national polling completely understated Reagan’s win.
1982 – A forgotten polling failure, but dems significantly underperformed in senate polling
1996 – Polling significantly overstated Clinton’s margin
2000 – national polling showed Bush ahead almost uniformly.
So these percentages in my opinion are way over-confident, and based on a faulty understanding of the true volatility in politics.
The 1980 election is a poor example. First, Reagan’s polling lead dampened Democratic enthusiasm. Second, that lead also bit into Carter’s numbers that went to the IND candidate. Finally, the election was called hours before polling stations in the west closed. That further reduced participation by Democratic voters and many of those that voted nevertheless went for Anderson because he’d made an effective pitch wrt to federal matching funds if he got 5% of the vote. Anderson also ran well in NE — residual pro-Kennedy and anti-Carter votes?
1996 was the lowest voter participation rate ever. The outcome wasn’t in doubt because Dole was one of the worst candidates ever, but Clinton was never popular.
2000 — up until the conventions, a large portion of GWB’s lead was because people thought it was his father running and after eight years of Clinton he didn’t seem as bad as he had in 1992. (The nostalgia rebound among the uninformed electorate.) In the closing days of that election, the polls got it right — particularly in FL.
I just voted and turnout here is crazy – this bodes well for us.
Hope your enthusiasm is wildly contagious in KY. I might have caught a bit of it way out here on the left coast late yesterday when for the first time, I wondered if KY had had enough of McConnell.
Same as you but I have Nunn losing of its a runoff, and Iowa and Colorado switched.
I had to vote provisionally because my registration change didn’t go through. Really annoying. I await a time when this country eliminates voter registration or at least joins the advanced world where it’s automatic. While waiting for the election officer to get my provisional ballot I heard some old white guy say voter ID laws are a solution searching for a problem that doesn’t exist. I turned around and gave him a thumbs up.
See? Old white guys aren’t all bad.
I may be in a different state, I suspect, but I’m another of those old-ish white guys who has said the same thing more than once. 🙂
My Chicago and Suburban Cook County People..
EVEN IF YOU ARE NOT REGISTERED, YOU CAN REGISTER AND VOTE TODAY!!!!
UNTIL 7 PM TODAY.
Here is the link to the Chicago Locations where you can Grace Period Register:
http://www.chicagoelections.com/en/grace-period-registration-and-voting.html
Suburban Cook County Locations where you can Grace Period Register:
http://www.cookcountyclerk.com/elections/registertovote/Pages/ElectionDayRegistration.aspx
yes and all jurisdications in IL have this option
My husband, son, and I voted at 8:00 am and there had been thirty-five voters ahead of us. The weather here in SW Ohio is currently sunny and cool, so hopefully that will encourage better turnout. It’s discouraging to keep hearing how badly the Democrats will fare, but I’m always going to cast my ballot.
So if/when the GOP wins the Senate we can expect a really bad two years coming up. There will absolutely be government shutdowns (plural) in the first year while the GOP fights to overturn ACA. The voters are telling them now that there is no penalty to be had for shutting down government in the first year of a two year term.
Furthermore, they HAVE to pass bills to overturn ACA and HAVE to try to shut down government to get it overturned. Why? First, their base really does believe that ACA is worse than the Holocaust to exponent of Slavery. Second, any Senator who doesn’t go along will be primaried and will lose that primary – period. Ted Cruz will lead that effort. Third, the filibuster won’t work. If the Dems try one the GOP will simply rewrite the rules and blame the Dems (for which Politico will agree).
So it’ll be Obama trying not to cave on his signature accomplishment versus the GOP-led Congress, with the country as hostage.
McConnell has already said that repeal will be the first order of the Senate under him. How many bills will come to his desk before Obama caves? With the government shut don and the nation in peril, he WILL sign say the hundredth bill to hit his desk.
If they shut down the government, that would be fantastic politically for the Democrats.
People don’t understand which party actually controls the House, because people are poorly informed and incurious, but a GOP led Congress that burns the Capitol down with regularity will be excellent news for John McCain…oops, sorry, the Democrats.
Personally, I’m starting to hope for a 50-50 split, so that Biden can break important ties, but McConnell and Boehner have to perform some tasks of governing.
but after tonight it is going to be obvious that the GOP has not path to 270. The Democrats are going to win every Senate Seat in states that Cook rates as Strongly, Leaning, or Likely Democratic with the exception of Collins in Maine.. Those states comprise 253 electoral votes. Beyond this, Dems will win two states that the GOP has to have in Virginia and North Carolina. They will likely win in NH, and may an incumbent governor out in another state they have to have: Florida. They really are only showing strength in Iowa and Colorado among the 7 States Cook rated as tossup.
All of this in a midterm election where turnout will be done, and with an unpopular President.
The case of the emerging democratic majority will be stronger after tonight that it was before it.
My predictions:
We split IA and CO
We narrowly lose in KS
We hold NC
There will be two runoffs, one in GA and one in LA, and we are likely to lose both.
So about New Hampshire. I was at a Hillary Clinton event on Sunday in Nashua – he first entry into the state since 2012.
She did not come close to filling the room.
Think about that. I think it may mean bad things for Shaheen, but it also says something about the NH primary.
To borrow the line from Death of a Salesman, maybe she is liked but not well liked.
very interesting. thanks for your reports overall
If the final results pan out about the way you and a lot of others have said (and the projections are solidly data driven), I expect that the GOP will overstate their significance. It will be an election where the map was favorable for them to begin with, and where GOP constituencies tend to disproportionately show up. What you’re projecting though, is about what I would project as well – basically no real surprises in terms of the overall balance of power in the Senate. Maybe the get out the vote efforts will have had some impact, and give us all a few surprises. Not holding my breath, of course. I know better.