The Real Winners and Losers of the Midterm

The 2014 midterms brought with it hundreds of literal winners and losers.  However, in the broader picture, here are our big three winners and losers of the night:

Winners

GOP Governors

Republicans had a much better than expected night Tuesday.  We had projected them to lose three seats, with many of them falling within one or two points of one another.  Instead, the GOP swept nearly every competitive Governor race, and some that we didn’t even see as competitive. 

What this means is that the GOP now control of the Governor’s mansions since 2000.  The pickups include wins in Arkansas, Maryland, Massachusetts, and Illinois.  Arkansas was pretty much a foregone conclusion, but Maryland and Illinois, especially, shocked most political observers.

Many of the other winners were incumbents running for re-election in purple and even blue states.  These Republicans not only will have more influence in their state’s voting, but they also will have a bigger role in the 2016 Presidential race.  For some, that includes running for President, while for others it means more political sway with potential endorsements.  The Governor who looks best coming out of Tuesday is Scott Walker, who has now won three statewide elections in the past four years in liberal-leaning Wisconsin. 

John Kasich

One other Governor who came out of the midterms looking strong is Ohio Governor John Kasich.  Kasich has taken credit for a much improved Ohio economy since he took office, and he won re-election Tuesday night by a whopping 31 points.  True, he had a challenger who was embroiled in scandal and who turned out to be a very poor candidate.  But the scale of his victory is still massive, especially in a state that is one of the most important to Presidential candidates.

Though he would likely have a tough time as an establishment Republican in a potentially crowded Presidential field, Kasich has certainly established himself as a top-tier contender for the Republican Vice Presidential nomination.  Republicans have never won the White House without also winning Ohio, and the temptation to nominate the state’s extremely popular Governor may just win out over choosing for other reasons, such as courting a specific demographic.

Kentucky Democrats

Despite a sea of Democrats losing their seats Tuesday night, those in the Democratic-controlled state house of Kentucky surprisingly kept their majority.  Why is this important?  One reason: Rand Paul.

Paul is widely assumed to end up running for President in 2016, but under rules for Kentucky elections, no candidate can appear twice on any one ballot.  This means Paul cannot run for both his current Senate seat as well as for President at the same time; he would have to choose. 

If Republicans had gained control of the state-house, it is possible they could have signed legislation overturning this election law.  However, with a Democratic-controlled state house and Governor, there is no way it will happen.

If Paul chooses to run for President, it will leave an open seat in the Bluegrass State.  This could provide an opportunity for a strong Kentucky Democrat to run for Senate and have a good chance of winning, especially if Hillary Clinton is at the top of the Ticket.  She is still relatively well liked in the state and would likely want to make a play for winning Kentucky herself.

 

Losers

Pollsters

We’ll have a more in-depth article specifically covering this issue, but in short, the polls were wrong.  Our projections are mainly based on polling averages and they have only about a 10% room for adjustment, which we, for the most part, tilted to the Republican candidate. 

On average, it seems as though pollsters had about a four to five point Democratic bias. Many of them are claiming that the race broke in just the final few days of the campaign towards the Republicans, and that many of their polls were not conducted in that time.  But for a race that was remarkably stable for months, that seems like a thin argument. 

Hillary Clinton

Let’s not kid ourselves; Hillary Rodham Clinton is almost certainly going to run for President.  Considering that, the midterms were not a great omen for her chances.  It’s certainly easy to read too much into this, but of all the states in which she campaigned for the Democratic candidate, the only state in which she seemed to have a positive effect was New Hampshire.  The Democratic Governor there, Maggie Hassan, ended up winning by about six points, beating the spread between the state’s Senate race. 

In contrast, the Clintons put much political capital into helping the Democratic candidates in Florida, Arkansas, and Kentucky, among others.  All of those candidates lost by a much larger margin than many were predicting (again, mostly due to polling error).  Still, it seems as though Clinton’s clout is much diminished from where it was even a year ago and certainly lower than in 2008. 

All this likely won’t deter her from running.  Turnout was terrible in 2014, especially among groups who typically support Democratic candidates, such as Hispanics, youth, and single women.  A run by Clinton will no doubt increase turnout among these three groups.

Martin O’Malley

Unlike Clinton, Maryland Governor Martin O’Malley will likely be deterred from running for President after Tuesday’s results.  Maryland’s electorate overwhelmingly voted for the Republican candidate to replace O’Malley over O’Malley’s Lieutenant Governor, Anthony Brown. 

Brown was largely considered a proxy for a third O’Malley term.  With Democrats not being able to hold on to this deep-blue state, it is unlikely the current Governor will have enough clout within the party to challenge Clinton, much less win a Democratic primary.

Election Projection - The Senate

When we launched our Senate forecast in September, Republicans had an approximately 53% chance of winning a majority in the Senate.  That is now up to about 77%.  The change hasn't come exactly linearly, though there certainly has been at least some trend line, specifically over the past month.  

The GOP's chances rose slightly in early September and then fell again when a string of polls showed Democrats with surprisingly strong results.  However, that high for Democrats didn't last long, and since the GOP's low of 53.9% on October 12th, they have steadily increased, almost daily, their chances of taking a majority in the next Congress.

In that time, what has changed?  Not that much, actually.  Most races have just shifted slightly to the Republicans, which has created an overall shift in our forecast.

In September, our two closest races were Iowa and Colorado.  Those states are now both leaning towards Republicans.  Now, the two closest races are Kansas and Alaska.  Kansas was leaning Democratic about two months ago, and Alaska was leaning Republican.  

Alaska, however, is the exception.  The majority of races have shifted about 10 to 15% towards the Republican candidate.  Below, we'll go into more depth on each race.

Beginning with the race Democrats are most likely to win, New Hampshire's Jeanne Shaheen has held on despite a surprisingly strong challenge from former Massachusetts Senator Scott Brown.  The race has become tight over the past few weeks, but Shaheen still holds a polling edge.  She is expected to win by about 2%, according to our forecast.

New Hampshire represents the only 'lean Democratic' state we have at this point, though North Carolina's Kay Hagan is nearly in that column.  Her seat essentially must be won by Democrats in order for the party to have any hope of winning today.  If Republicans were to win North Carolina, Democrats would only have about an 11% chance of winning, and that would include sweeping many states that they are clearly behind in at this point.  Still, she has a 58% chance of winning tonight, and we project her to win by a margin of 1.2%.

Next in the tossup column is Kansas.  Independent candidate Greg Orman could end up ruining the GOP's chances at a Senate takeover by turning what was once thought of as a safe Republican seat to a true tossup.  He has about a 52% chance of winning at this point. The unknown factor is how the undecided voters in the race will vote.  If they end up 'coming home' to the Republican candidate instead of largely splitting the vote between the two candidates, Roberts will end up winning.  However, we assume Orman will win about 45% of the undecided vote, giving him a mere 0.3% win after it's all said and done.

Alaska is by far the state that could produce the biggest unknowns tonight.  The polls won't be closed there until 1 am ET, and it could take weeks to tally all the mailed in and absentee votes.  Polls (as expected) have been all over the place, with the average giving Mark Begich a very slight lead.  But when factoring in the state fundamentals, Begich actually has less than 50% chance of winning (about 48%).  That is essentially a coin flip.  We project he will lose by about 1% to Dan Sullivan, but the uncertainty is high.  Either candidate could win by about 5%, depending on which polls were closest.

Iowa and Colorado represent the first two 'leaning Republican' states.  Polls have tightened in recent days in both states, but at this point, Democrats need to hope for a slight polling error or larger than expected turnout in order to keep either of these states.

Iowa's Joni Ernst has run a surprisingly strong campaign, while the Democrat, Bruce Braley, has made numerous unforced errors.  The Democrat has only a 35% chance of winning now, down from about 50% in early September.  

Meanwhile, the situation is nearly exactly the same in Colorado.  The Republican, Cory Gardner, is running about 2 points ahead in the polls, but we expect that to tighten slightly on election day due to Colorado's mail-in ballot law, which we see as helping the Democrat slightly.  In the end though, we expect Gardner to hold on to win by about 1.7%.  

Georgia is a state in which the Republican Senator is retiring, leaving an open seat. Democrats are working to finally turn Georgia purple, and it shows.  Michelle Nunn is trailing by about three points currently, which is within the margin of error of most polls.  We project that this race will go to a runoff, however, as the winner will need to win by more than 50% of the vote.  We currently expect David Perdue to get just short of that, 49.9%.  At this point, Nunn should be hoping for a runoff, because there is only about a 5% chance she could win outright tonight, given the current polling.

Louisiana's Mary Landrieu is in much the same boat, although she is most certainly not hoping for a runoff.  We expect her to 'win' with about 46.4%, well short of the 50% she also needs for a runoff.  Her problem, however, is that the vast majority of the votes not going to her main opponent, Bill Cassidy, are going to an even more conservative opponent.  It's unlikely she will pick up many of those voters in a runoff, and her chances of winning eventually are only 14%.

The next three races look to be safe for Republicans.  It would require a huge polling error in any of these states to give an upset for the Democrats.  Arkansas' Mark Pryor was running neck and neck with his Republican opponent for much of the spring and summer, but since then, Cotton has pulled ahead in the polls, and we expect him to win by about 6.2%.

There is a similar dynamic in Kentucky, home of Senate Minority Leader Mitch McConnell. His Democratic opponent was keeping the race close until a couple weeks ago, when she infamously refused to say who she voted for in the last Presidential election.  Now, McConnell has pulled ahead, seemingly taking all of the previously undecided voters with him.  He is favored to win by 7.8%.  

And finally, Republicans had a bit of a scare in early October when a couple polls showed their candidate, Mike Rounds, with only a slim lead against either the Democratic or Independent candidates.  However, after more polls were released and money was poured in the state to help shore him up, it seems like Rounds will safely win, with about 47% of the vote.

So, what does this all translate to?  We now expect Republicans to net five seats tomorrow night, bringing them to 50 seats and leaving the Democrats with 48 seats (including Greg Orman).  This scenario would require that both Michelle Nunn and Mary Landrieu would need to win their runoff elections, an extremely unlikely scenario.  If Republicans win both, they will hold 52 seats (not including Orman) in January. 

Democrats at this point need to hope for the polling to be slightly wrong in two of the three states including Alaska, Colorado, and Iowa.  It's not an impossible scenario, and there's about a 25% chance Democrats could pull off an upset in those states, which is, not coincidentally, about the same chance they currently have at keeping the Senate in our overall forecast.  

Either way, we won't know for sure who will control the Senate likely until January.  It's going to be a fun and crazy night.

Election Projection - The Governors

It's finally here!  Tomorrow, the nation will vote in the 2014 midterms, and as such, Battleground270 is revealing our complete election projection for both the Senate (later tonight) and Governors races.  This will include not just the chances of each candidate winning, but what we project to be the final vote share.  Let's get started! 

If you are a frequent reader, you'll notice a similar-looking image above.  This represents the likelihood that the Republicans will have control of a certain number of Governor's mansions after the election.  It's remained remarkable consistent throughout our forecast.  At the beginning of our forecast, Democrats had a slightly better chance to have 23 seats, picking up two seats.  Now, the model forecasts that the most likely outcome is that Democrats will pick up only one seat.  The likelihood of this happening is 21.39%.

But like any model, it represents the most likely chance, not the final result.  That's where our vote share projection comes in.  Like in previous years, we use data from the model to help determine who will actually win.  In that, we factor in state fundamentals, polling recency, and some other more subjective factors to come up with a final vote tally.  

Below, you will see the official projection.  You will likely note that not all of the Gubernatorial races are represented on this chart.  I have only calculated the chances for races that are not in the 'safe Dem' or 'safe GOP' column.  We'll begin at the top and work our way down.

Georgia's race looked to be a nail biter, with a likelihood that it would go to a runoff.  That conventional wisdom no longer seems to hold, as a string of polls in recent weeks has shown Governor Nathan Deal pulling ahead.  We currently project that he will get greater than 50% of the vote and therefore avoid a runoff.

Massachusetts is another case in which our original assumptions were way off.  The state is deep blue, and it was expected that the Democratic candidate would have a relative cake-walk to winning the election.  We were probably wrong.  The Republican, Charlie Baker, has pulled ahead in recent polls and taken a small but stubborn lead against Martha Coakley.  It's likely that Massachusetts' liberal base will end up helping Coakley a bit, but in the end, we still project Baker to win by about 1.2%.

Michigan Governor Rick Snyder has presided over a relatively liberal state for the past four years, and despite having decent approval ratings, he is barely ahead of his challenger.  Again, this is because Michigan's electorate is much more favorable to Democrats.  Additionally, voters have a relatively negative association with almost any incumbent this year, so that is also likely helping his challenger.  Still, we project Snyder to win 49% to 48%.

Wisconsin's Scott Walker is in much the same boat as Rick Snyder.  He is facing a credible challenger and the polls are neck and neck.  The most recent polling has shown a slight one to two point advantage for the incumbent, which translates to about a 67% chance of him pulling off a victory tomorrow.  But it will be close.  We expect him to win by only 1.3%.

Now we are getting to the races we expect the Democrat to win.  Despite being the closest incumbent to having a 50% chance, Maine's Governor's race is not expected to be the closest race of the night.  This is due to the fact that the forecasting model does not take into account subjective factors when determining a percentage likelihood of winning, but we do take that into account when predicting the final vote share.  

In Maine's case, there is an independent candidate on the ballot who recently made news by essentially saying that his voters shouldn't vote for him if they don't believe he will win (and he won't).  Therefore, we've concluded that about 3% of Cutler's voters will end up voting for another candidate, and the majority of those voters will likely vote for the Democrat.  That means we expect Michaud to beat the incumbent governor by about 1.4%.

Next up is Colorado.  This state is probably the 'swingiest' of all the battleground states this year.  While the Senate race currently favors the Republican candidate, the Gubernatorial race has seen the lead switch to a different party almost every week.  This week, Governor Hickenlooper has a slight edge, and we believe that will hold through tomorrow.  However, his win is going to be by the slimmest of margins, about 0.5%.

Alaska is yet another curious case in which we originally thought last December that Governor Parnell would skate to victory.  However, after the Independent and the Democrat joined forces on a unity ticket, the race became very competitive.  Most polls in September and early October showed the unity ticket leading by a healthy margin.  But recent polls from the state show a tighter race.  It is also worth noting (as we have in our Senate forecasts) that Alaska is an incredibly tough state to poll.  We are basing our prediction off the polling, but we wouldn't be surprised if Parnell pulled off a win.

On the complete opposite side of the country, Florida Governor Rick Scott is in the political fight of his life to retain his position.  He and former Governor Charlie Crist have been running neck and neck all fall, and this race is going to be decided on who shows up to vote.  Crist has had the polling edge over the past couple weeks, and we have him winning by 1%.

One incumbent who seems to be staying afloat is Illinois Governor Pat Quinn.  Though his Republican opponent, Bruce Rauner, was leading him in almost every poll throughout the summer, Quinn has finally broken through by defining Rauner as a one-percenter who doesn't care for the working class.  The line of attack seems to be working, as we expect Quinn to end up winning by 3.9% in President Obama's home state.

Kansas is, for once, getting a huge amount of attention this election cycle due to competitive Senate and Governor's races there.  Governor Sam Brownback is in trouble with voters due to his aggressively conservative moves in the state, and it looks as though Democrats may strike back this year in the deepest of red states.  Brownback has only a 33% chance of pulling off an upset, and we believe the race, though close will be won by about 1.4%.

Rounding out the 'lean Democratic' states are Connecticut and New Hampshire.  Neither state was on our radar early on, but both have become more competitive.  Governor Malloy is now likely to win by only about 2.8%, while New Hampshire Governor Maggie Hassan will win by about 3.6%.

Republicans have an outside chance of grabbing Rhode Island and Maryland.  They've touted recent polls showing the races are closer than were expected.  Still, Republicans are not leading in these states, and the Democrats shouldn't have too much trouble defending them.  

If you're counting, by now you will realize that we are predicting Democrats will actually pick up three seats as opposed to the model's "most likely" projection of only one seat.  This is because, as we said before, the model gives someone who may end up winning by only half a point a 45% of still pulling off an upset.  When calculating total vote share, it became more obvious that Democrats could have a decent night, at least for Gubernatorial races.

It will certainly be a fascinating night for political junkies.  We predict there will be at least eight races in which the two major candidates will be within two points of one another.  That is relatively unheard of in an election year.  If we begin to see some of the close races that were predicted to be won by Democrats actually won by Republicans, it will likely be a terrible night for the Democratic party.  However, if states like Wisconsin and Michigan end up being won by the Democratic candidates, it could mean that polling was off nationwide, and Democrats could have a better-than-expected night.

Senate Update - Democrats Should Worry

Since the middle of September, the conventional wisdom here (and at other sites like 538) has been that, yes, Republicans have a slight edge when it comes to the possibility of control of the Senate, but they are not heavy favorites and the Democrats don't need to panic.  

That's probably changed now.  Republicans now have a 73.3% chance at winning an outright majority, up from 69.8% yesterday.  It's at a tipping point now where the GOP has about a 3 to 1 chance of winning Tuesday night.  

The balance of the Senate is almost assuredly going to rest on three states: Iowa, Kansas, and Georgia, with Alaska as a wild card.  Democrats would need to win three of those four contests, assuming they kept other states in which they are ahead (NC and NH) and lost the other states in which they have consistently trailed (CO, AR, LA, KY, WV, SD, and MT). That scenario would also include the Kansas Independent caucusing with the Democrats, as we assume.

Overall, the races in all of those four states are pretty close, with Alaska being a bit more of an unknown.  We had a flurry of polling out of Iowa yesterday showing a statistically tied race.  However, yesterday the Des Moines Register poll found Ernst leading by seven points.  The Register poll is typically an excellent bell-weather for where the state is at.  However, it's extremely unlikely that her lead is seven points, when every other poll has shown something between a two point lead for her to a one point lead for Braley.  Still, Ernst's chances of winning have risen to 63%.

Kansas is now ranked as the closest state in our forecast.  Each side has an exactly 50% chance of winning.  Orman has had a slight lead in most recent polls, but the conservative tilt of the state will likely help Roberts with undecided voters.  Democrats need Orman to win and then decide to caucus with them.

In Georgia, it now seems as though neither side will win on Tuesday.  Most polls show the race as tied or giving it slightly to Perdue.  But Perdue isn't getting to 50%, mainly due to the Libertarian on the ballot.  At this point, Democrats should be happy if the race goes to a runoff.  That won't necessarily give Democrats more of an advantage, but it gives them a longer time to get ahead again in the polls.  Currently, Perdue has a 60% chance of eventually winning.

Finally, Alaska is our wild card.  As we discussed this week, polling in Alaska is scarce and unreliable.  We have a couple polls now showing the Democratic incumbent with a large lead and multiple surveys showing the Republican, Dan Sullivan, with a marginal lead.  To make it even more suspenseful, we likely won't know the final tally of the votes in Alaska for a week or two after the election. 

It all ads up to the likelihood that we won't know the fate of the Senate on election night, unless Republicans pull off surprise wins in places like North Carolina and New Hampshire. If that happens, Democrats are going to have a long night, and it will be apparent that the polls were skewed towards the Democrats the entire time.  However, the opposite could happen; if Democrats win by larger than expected margins in those east coast states, it may be a sign that the polls were biased towards the Republicans.  

Senate Update - Iowa is Key

Lot of new polling from Iowa today, and basically all of it told us what we already knew: the race is a dead heat.  However, in every poll, Braley never led.  While all were within the margin of error, Ernst led by two in one poll, one point in two polls, and they were tied in the other.  This race will perhaps be the closest race on Tuesday night...

... Except perhaps Kansas.  A new poll showed Orman leading by only one point, and that matches up with his overall average against Roberts.  The polling in Kansas has jumped around a bit more, so there's the potential for either candidate to win by a larger-than-expected margin.  Still, this one could go into overtime.

In North Carolina, it seems as though Senator Kay Hagan may just hold on.  Two new polls out today had Hagan leading, though they were both within the margin of error.  The early vote numbers look good for her.  Still her chances haven't changed from 61%.

Finally, we have two other states of note.  Rasmussen polled the Alaska Senate race and found the Republican, Dan Sullivan, ahead by five points.  As we've noted before, Alaska is a tough state to poll, but nonetheless, Sullivan's chances increased to 58%.  And in New Hampshire, Senator Jeanne Shaheen got more good news with a poll showing her seven points against her challenger.  This is the second poll in as many days to show her lead in the high single digits, indicating that while the race may be close, she is clearly ahead. Shaheen now has an 80% chance of winning.

Overall, though, the Senate picture is completely unchanged.  Republicans still have a 69.8% chance of retaking the upper chamber.  At this point, no news is good news for Republicans.

Gubernatorial Update - Halloween Edition

It has been a few days since we last updated the Gubernatorial forecast here at Battleground270.  Overall, not much has changed as far as how many seats each party is expected to have after November 4th.  

Republicans still have a slightly better chance of only losing one seat as opposed to multiple or keeping or gaining on their majority.  There is about a 21% chance Democrats will pick up one seat, while there is an almost even chance, at 20%, that Democrats will pick up two seats.

So which seats are ripe for picking off?  Well, it seems as though Democrats will pick up seats in multiple states, but that will be offset by some gains by Republicans in currently Democratic-held states.

Democratic Pickups

Pennsylvania Governor Tom Corbett is pretty much dead in the water, and that seat is the only sure pickup Democrats have at this point. 

Beyond Pennsylvania, Democrats certainly have pickup opportunities, but they are less likely.  Kansas' Republican Governor is currently trailing in the polls, but because of the conservative tilt of the state, Governor Brownback still has a 43% chance of holding on.  

In Florida, we see a similar situation playing out.  The Democratic candidate, former Governor Charlie Crist, is leading Governor Scott in the polls.  This time, though, because Florida doesn't have a Democratic or Republican tilt in our model, Crist has a 62% chance of taking back the Governor's mansion.

Maine is another state that Democrats believe they will pick off on Tuesday.  The Governor there is deeply unpopular with a majority of the electorate, but he could still survive if the Independent candidate draws enough of the anti-LePage vote.  To perhaps combat that scenario, the Independent candidate said Wednesday that people should essentially vote for someone else (wink wink) if they don't believe he can win (which he can't).  It will be interesting to see the final numbers in this race Tuesday night.

Finally, Democrats are underdogs but still have decent chances against sitting Republican Governors in Georgia, Michigan, and Wisconsin.  Wisconsin is seen as the likeliest of these races going to the Democrats, but a poll out Wednesday showed Governor Scott Walker with a seven point lead.  While this poll is almost assuredly an outlier, it should be noted that there haven't been 'outliers' in Mary Burke's favor.  Still, before that poll, Walker's chances were only 51%.  They are now 64%.

Georgia and Michigan represent tougher opportunities for Democrats.  Of the two, Republicans have a better chance of holding onto Michigan, despite its Democratic tilt, because polls currently show Rick Snyder with a small but consistent lead.  In Georgia, however, if Governor Nathan Deal cannot secure 50% of the vote on Tuesday, the race will go into a runoff, where the Democrat will have another shot at taking him down.

Republican Pickups

Despite all these possible gains for Democrats, Republicans look like they will have some pickups as well.  Arkansas is the only state in which it looks like a sure thing for the GOP to win the seat.  A string of polls has shown the Democrat, Mike Ross, trailing by a large amount.  He now has only a 3% chance of winning.

Massachusetts is the next best pickup opportunity for the GOP.  The Democrat, Martha Coakley, was at one time the heavy favorite.  But her in a combination of her running a not so great campaign and the Republican running a near flawless campaign, the seat will now more than likely turn red.

In both Colorado and Connecticut, Republicans have recruited strong candidates to run against the incumbent Governors.  Colorado seems to be a dead tie at this point, with both candidates showing leads in recent polls.  In Connecticut, the Republican candidate has a slight lead in the most recent polls.  However, like with some other states, Connecticut has a deep blue tilt and that brings his chances of victory down to only 51%.

Finally, Republicans still have a slight chance to win in deep blue Illinois.  The Republican, Bruce Rauner, was leading in some polls earlier this month, but the most recent ones show Governor Quinn with a slight lead.  His chances are now 62% that he'll keep the seat.

Senate Update - Arkansas is Off the Map

Thursday brought a lot of unexpected (and contradictory) polling.  For starters, Elon University found North Carolina Senator Kay Hagan ahead by four points while Rasmussen showed her ahead by one.  Either way, Kagan's lead has probably stabilized to around one to two points.  She now has a 62% chance of winning.

A similar story played out in New Hampshire, where an ARG poll put the race at a tie, while UNH showed Senator Jeanne Shaheen leading by eight points.  Again, her lead is likely not that large, but the race is also probably not a tie at this point.  Brown has yet to lead in any poll by more than one point.  Brown's chances of victory have now dropped to 24%.

In Colorado, polls are continuing showing Cory Gardner with a lead, but they vary by how much.  Quinnipiac now has the Republican ahead by seven, while SurveyUSA has him ahead by only two, well within the margin of error.  SurveyUSA's result was unchanged from their previous survey, indicating the race likely hasn't moved in the past couple weeks to one candidate or another.  As I've noted before, especially in this race, it will be all about turnout.

Mitch McConnell is also resting easier tonight, with a new poll showing him ahead by five points over Grimes.  The SurveyUSA poll had previously had McConnell leading by only one point, so it seems the race has indeed shifted in the Republican's favor.  McConnell's chances of winning next Tuesday are now 86%.

Finally, we come to Arkansas.  We haven't had a whole lot of polling from the state, which has left the model somewhat uncertain.  However, in the past two weeks, we've gotten five

polls, each showing the Republican, Tom Cotton, ahead.  But what really hurts Democrats are the two polls out today showing Cotton with a seven and thirteen point lead.  While it's probably Cotton doesn't actually lead by thirteen, it would be highly unusual for a poll to be thirteen points off, meaning Pryor only has a 5% chance of still winning this race, down from 18% before those polls.

Mostly on the basis of the Arkansas numbers, Republican's chances overall rose from 65.7 to 69.8%.  That result now accounts for the highest chance the GOP has had since we launched this forecast in mid-September.  With only four more days until the election, Democrats are going to need all of them to be pretty good polling days in order to feel more comfortable going into Tuesday.

Senate Update - 29 October

On the whole, Democrats didn't get great news out of today's polls.  We saw two new polls out of Georgia, as well as a poll in Iowa and one in Kansas.  Republicans saw their chances of taking the majority increase to 65.7%, up from 64.4% yesterday.  

Let's begin with the good news for "Democrats".  That is, the poll from Kansas showed the Independent leading Senator Pat Roberts by two points.  Orman's average lead now is about one point, which would translate to about a 58% chance of victory on election day. However, that doesn't include state fundamentals, which bring his chances down to 52%. Still, that is higher than his chances yesterday, which were at 49%.  

Iowa, in contrast, gave Joni Ernst a slight advantage.  A new Quinnipiac poll put her up by four points over her challenger, Bruce Braley.  Still, because we have so many polls from this race, her average lead didn't change much.  She now has about a 61% chance of winning next week, up from 59% yesterday.

Finally, the two polls in Georgia showed a statistically tied race, which is probably best described as bad news for the Democrat, Michelle Nunn.  Her lead had been growing until a couple days ago, and it seems the race has reverted back to a tie.  The first poll, from SurveyUSA, had Perdue leading by three points.  The second poll, conducted by Rasmussen, showed a tie.  

Either way, this race looks headed for a runoff, as either of the candidates would need to win by about three or four points in order to get 50% of the vote.  And that is why it is probably bad news for Democrats.  The conventional wisdom is that in a runoff election, turnout will likely be lower, which usually helps the Republican.  Still, there is a week left, and it could be that the pollsters are missing something on either side.  But more likely, we won't know the fate of the Georgia race until January.

Gubernatorial Update - Still Many Tossups

Yesterday, YouGov blessed us with polling from all of the major gubernatorial races.  While some polls helped add some confidence to our model about certain races (such as Rhode Island, Arkansas, and Hawaii), others just added to the confusion.

There are still seven states which could be described as 'tossups' and many more which are on the edge.  The tossup states are Colorado, Connecticut, Florida, Illinois, Kansas, Maine, and Wisconsin.  Of those, the model is most sure about Colorado and Illinois, giving the Democratic incumbents a 58% and 57% chance of victory.  The others are all between five percent of one another.  

Democrats are now about equally likely to win 22 and 23 seats, with the latter being slightly more likely.  Still, their chances of getting to at least 25 seats are only about 10.7%, about the same chance Democrats have to come away with exactly 51 seats in the Senate.

Senate Update - What's Going on in Alaska?

Regular readers will know that there is always a lot of uncertainty in the polls in Alaska, which is why the model typically is less sure of the outcome.  Today, that is especially true.

This is because a new survey was released today showing Democratic Senator Mark Begich with a six point lead.  Another survey last week had shown him with a whopping 10 point lead.  Meanwhile, a CBS/NYT/YouGov poll over the weekend put Sullivan ahead by four points.  

Alaska has always been a very difficult state to poll, and that is certainly reflected in these wildly different results.  For much of the summer and fall, Sullivan has had leads of between two and six points.  But these new polls tell a wildly different story. 

Much hay has been made about Begich's incredible ground game, and as Harry Enten noted, a turnout of only a few thousand extra voters could swing the race by about 3%. Before today's poll, Begich's chance at winning were only 30%.  However, afterwards they stand at 45%.  This brings the GOP's overall chances at winning the majority in the Senate to 64.4%, down from 67.7% yesterday.  That amounts to the largest change in our forecast since two weeks ago, when Republicans on the whole had a really great day of polling.

Most of the other polls released had a minimal effect on the model.  Iowa's Bruce Braley's prospects increased to 41% from 37%, but Senator Mary Landrieu's fell from 14% to 11%. Meanwhile, both of the Republicans in Colorado and North Carolina increased their chances of victory to 72% and 41%, respectively.  

Overall, this means that the fate of the Senate is even more up-in-the-air than we previously thought.  Meanwhile, due to probable runoffs in both the Louisiana and Georgia races, there is now about a 70% chance we won't know the who hold the majority until much after November 4th.  That's enjoyable for political junkies like myself, but probably not so good for the candidates in those races.

Does Harry Reid Care More About Power or the Majority?

The race for the Senate is close.  Like, really close.  But there is an additional wild card that no one is talking about.  

That wild card is Greg Orman.  Orman hasn't decided yet who he will caucus with.  He has said he won't decide until he gets to Congress and will caucus with the majority party.  But it is very likely his vote will be the determining factor.  But wether it is the Democrats or Republicans, the Independent has also said he will not vote for either Majority Leader Harry Reid or Minority Leader Mitch McConnell.  

For Mitch McConnell, this scenario doesn't really matter.  If Democrats win 49 seats (not including Orman), and Republicans hold 50, Orman's vote would have to be for the Democrats for it to be a tie (which would be broken by Vice President Biden).  If Orman did not vote for Reid, however, McConnell would win with 50 votes to Reid's 49.  

So, this scenario would require Reid to willingly decide to give up his power as Majority Leader and allow another Democrat to step in.  

At first blush, it seems like the Majority Leader would not be willing to make this move, even if it means having a majority.  There would likely be a hard push from from Democrats to appease Orman in order to keep the majority, but the man in question would likely not relent.  

Also, who would be the next choice for Democrats?  Senator Chuck Schumer has been Reid's right hand man for some time now and would be an obvious replacement.  He is from a safely Democratic state, one which is the fundraising capital of the politicians he would be leading.  

More likely though, Harry Reid will play roulette and hope the Independent will eventually cave. 

Senate Update - 26 October

This year's midterms are shaping up to be among the most unpredictable and suspenseful in recent history.  This is due to the large number of races that neither party has been able to put away.  For the Democrats, those races include North Carolina and New Hampshire. For the GOP, the list includes Alaska, Arkansas, Louisiana, Georgia, Kansas, Kentucky, Colorado, and Iowa.  

As we discussed Friday, many of those races are likely to go to the GOP, not because there is a wave election, but because it is a year in which Democratic Senators must defend their seats when it isn't a wave year for their party, as it was in 2008.  

And despite a spate of new polling out today, we're still no closer to knowing who will likely control the Senate majority in January.  I won't go through every poll and race, but I will mention the races that moved to a larger degree and those that are still pretty much the same.

Senator Mark Udall got a slight bump today from two new polls showing the race a statistical tie (one showed him up by 1, the other down by 1).  This helps improve his odds of winning to 33%.

Both of those polls also showed David Perdue leading Michelle Nunn by a small margin. Still, Nunn has led in most of the recent polling, and the state is now only leaning towards Perdue, 51% to 49%. Chances rose today that the race will go to a runoff in January as neither candidate seems to be pulling ahead.

Senator Pat Robert's chances increased slightly today to 51% as one of the new polls showed his lead to be four points.  The other showed him trailing by one point.  Still, the race is as close as any, and it's very likely we'll be up late next Tuesday before we know who won the state.

Jeanne Shaheen's chances also rose today, on the heels of the CBS/NYT/YouGov poll that put her ahead of former Senator Scott Brown by five points.  This is her largest lead in the past couple weeks and is outside the margin of error.  With only one week left for Brown to hope for a game changer, Shaheen's chances are now 67%.

Finally, we at long last got another poll from South Dakota, and it appears the Republican, Mike Rounds, is in little danger of losing the race.  The poll showed him handily defeating both the Democrat and the Independent.  His win probability increased from 92% to 96%, as the three-way race can still make for a bit of unpredictability.

And despite all the new polls, each party's chances remained relatively the same. Republicans now have a 67.7% chance of winning control of the Senate, down about half a point from Friday.  Each party now has only nine more days to lock down the races they need, but realistically, Tuesday Nov. 4th's forecast is probably going to look very similar to tonight's.

Gubernatorial Update - 26 October

We haven't updated our gubernatorial forecast in a few days, but in reality, not much has changed.  There have been many polls out of Georgia, Colorado, Kansas, and Illinois, but most of them haven't moved the needles for their prospective candidates much.  Despite getting some good polling news on Friday, Georgia's Jason Carter received some more polls showing he was still trailing incumbent Governor Nathan Deal.  The race has tightened slightly, but Deal still has a 62% chance of winning.

In Colorado, it appears as if Governor John Hickenlooper is doing better than his Democratic counterpart, Mark Udall.  He consistently polls about three to four points better against his opponent than does Udall, and that will likely make the difference in the race for Governor.  Hickenlooper's chances have increased to 55% from 50%.

In Kansas, the race is still close, but two new polls from Rasmussen Reports and NBC/Marist show the Democrat, Paul Davis, with a seven and one point lead, respectively. The race appears to be leaning in Davis' direction, but Kansas is still a very red state, and that means Davis' probability of a win is still only 49%.

Finally, two fresh polls from Illinois show the Republican challenger, Bruce Rauner, leading Governor Pat Quinn by one and two points.  In a way, Illinois is the exact opposite as Kansas.  Pat Quinn is a Democratic Governor who is deeply unpopular in his state.  The challenger is leading slightly in the polls, but the outcome will be uncertain because of the deep tint of the state.  Still Quinn has a 47% chance of victory.  He faced longer odds in 2010 and still came out on top.  We'll see where the final week of the campaign takes this race.

Overall, the Democrats are still expected to have the best chance of gaining one seat from their current 21 seats.