After months and months of endless campaigning, the 2014 midterm elections are finally here. With the House firmly in Republican control, most of the attention this cycle is focused on the U.S. Senate.
In that spirit, I’ve gone through every U.S. Senate race in order of “Least Certainty who will Win” to “Most Certainty who will Win” and analyzed who is most likely to win each race, based primarily on raw and adjusted polling data compiled by Real Clear Politics and FiveThirtyEight, as well as what we can expect to happen during run-off elections based on less quantifiable voter behaviors.
State of the Races
Georgia
The Georgia Senate race almost certainly will not be decided on November 4th. Georgia is one of two states this cycle that mandates that the winner of its elections receives 50 percent of the vote. If no candidate receives 50 percent, then there is a run-off between the top two candidates. In a tight race between Republican businessman David Perdue and Democratic Points of Light CEO Michelle Nunn, neither of the candidates has attracted 50 percent of support in any major poll since the middle of September.
The race is already extremely close, with Nunn seemingly pulling ahead in mid-October, but Perdue reasserting a polling advantage within the last week. With an entire month of campaigning ahead, no outcome in this race would be surprising. But the structural factors are strongly in Perdue’s favor. Georgia has not elected a Democrat to the Senate since 1998, and President Obama’s approval ratings in the state are very low.
Unless the GOP pulls off an upset in a state where they are currently trailing, the Senate won’t be decided when Georgians go to cast their votes. That means that all of Georgia will see non-stop ads for a month about how their vote in December will single-handedly determine whether the President’s party will have control of the Senate, elevating the national implications of the race above the pros and cons of the individual candidates. With that in mind, I’d give a very slight edge to Perdue (R).
Kansas
The race for a once-safe Republican seat was turned upside down when Democratic nominee Chad Taylor dropped out of the race and endorsed Independent Greg Orman. Orman, a businessman and former Democrat, has supported a wide variety of Democratic, Republican, and Independent candidates over the years. He has attempted to form a broad coalition of Democrats, Independents, moderate Republicans, and anti-establishment conservatives, with seemingly some success.
The polls have been fairly consistent, with no major outliers in the past two weeks and both sides demonstrating occasional leads. But in FiveThirtyEight’s adjusted polling data, which takes into account historical biases of individual polling firms, Roberts has only been leading in one of the past eleven polls. Orman’s margin is still razor thin, and as with any hastily assembled and diverse coalition, it’s always possible that Orman’s support will dissolve come Election Day. But given the poling, the election-day favorite should be Orman (I).
Iowa
The same two men, Republican Chuck Grassley and Democrat Tom Harkin, have served as Iowa’s Senators since Harkin was first elected in 1984. Now, with Harkin’s decision not to seek reelection, Iowa will finally have to select a new Senator. Democratic Congressman Bill Bradley and Republican State Senator Joni Ernst won their parties nominations and have been neck and neck ever since. But the Des Moines Register, which has a strong track record of predicting Iowa elections correctly, just released a poll showing Ernst with a solid 7 point lead, seemingly propelled by the presence of the state’s extremely popular Republican Governor, Terry Branstad, on the ballot.
It’s never a good idea to base too much on a single poll, especially when it seems to be an outlier. But given the Register’s track record and Ernst’s overall slight lead in all the other polling (she’s led or tied in seven of the past eight major polls), Ernst (R) is clearly the favorite.
Alaska
Alaskan polling is notoriously sparse and inaccurate, making Alaskan elections extremely difficult to predict with any confidence. This Senate race is no exception. Incumbent Mark Begich is one of the many Democratic Senators from red-leaning states who benefited from the Obama-wave of 2008 to defeat a Republican incumbent. Given the anti-Obama sentiment in Alaska and a stronger Republican opponent than Begich’s 2008 foe, former Alaskan Attorney General Dan Sullivan, the GOP expected to pick up this seat with relative ease, and seemed to hold a firm lead since August.
The notion of a Begich upset gained steam when Alaskan pollster Ivan Moore released a late October poll showing Begich up six points, but the poll has yet to be substantiated by other polls with a Begich lead. Furthermore, the flawed Alaskan polling has tended to systematically overstate Democratic chances, and even still the clear majority of polls give Sullivan a lead. No one should ever feel too confident about Alaska because of its difficulty to poll, but all signs point to Sullivan (R) securing victory.
Colorado
(Disclosure: I worked on Cory Gardner’s campaign this summer, and I believe that I am still technically an employee of Gardner for Senate 2014 until after the election. The rationale for this prediction (and everything else written in this article) is entirely my own and is entirely based on publicly available information and polling data, and does not in any way reflect on Cory Gardner or his campaign.)
The once solidly red Colorado seems to have been rapidly shifting to the Democratic party in recent years, and the state hasn’t elected a Republican to the Senate since 2002. But in stark contrast to Colorado’s Republican most recent statewide Republican candidate, the Republican challenger, Representative Cory Gardner, has campaigned as a pragmatist who will work across the aisle, a message that resonates in a particularly pragmatic state. The Democratic incumbent, Mark Udall, has worked to combat this pragmatic image by focusing on Gardner’s record on reproductive issues, with the issue quickly becoming one of the defining topics of the campaign. For much of the summer, Udall’s strategy appeared to be working, but Gardner built up a lead in September and has been the favorite since then. The raw polling data has favored Gardner slightly more than FiveThirtyEight’s adjusted forecast, but both still give Gardner a noticeable lead.
At this point, Udall’s best hope is that the polling is fundamentally flawed. This cycle, Colorado is implementing a new system of all-mail voting, which has led to some speculation that Colorado will face increased voter participation among the state’s sizable youth and Latino population. If this speculation is accurate, then the polling data could be underestimating Democratic turnout, to the point that Udall will lead among actual voters. But given that thus far significantly more Republicans have voted early than Democrats, the speculation is probably overblown. If Udall does in fact end up winning the race, it will be in part because of the all-mail voting, and voting by mail could become a highly partisan issue nationwide. More likely however, is that the polls are accurate and Gardner (R) will break the Colorado GOP’s losing streak.
North Carolina
When the history books are written about the 2014 midterm elections, the dominant narrative will probably be that the Democratic Senators from red and purple states who had benefited from President Obama’s 2008 candidacy were unable to succeed when they were up for reelection in 2014. North Carolina’s Democratic Incumbent, Kay Hagan, seems to be the exception to this rule. While the GOP has fielded strong candidates in most other competitive races, Hagan’s opponent Thom Tillis was relatively unpopular as North Carolina’s Speaker of the House. This is still a Republican-leaning state in a Republican-leaning year, so it would be rash to totally discount the possibility of Tillis winning if there’s a nationwide Republican wave. But it appears that the polling simply isn’t there for Tillis. In the RCP average, the only poll since September with Tillis ahead was by a Republican firm and of registered rather than likely voters. And in FiveThirtyEight’s adjusted polling, Hagan has been either ahead or tied in the past 12 polls. The final result will likely be very close, but there’s enough evidence to think that Hagan (D) has an edge.
New Hampshire
Republicans became excited about their prospects in the Granite State when former Massachusetts Senator Scott Brown moved north to run against Democratic Incumbent Jeanne Shaheen. Brown was arguably the most moderate member of the Senate between 2010 and 2012, and New Hampshire is a fiercely independent state. But Brown’s recent Massachusetts residency made it easy for Democrats to paint him as a carpet-bagger, and he has never been able to gain the ground that the GOP hoped he would, with Shaheen consistently leading in the polls. Brown supporters site the occasional poll that demonstrates a Brown, but these polls have consistently been conducted by either the Republican Vox Populi or the non-partisan but regional New England College. During the entire cycle, there has not been a single poll conducted by a national non-partisan polling agency that has shown Brown ahead. The final number will probably be close, but expect a victory for Shaheen (D).
Louisiana
Like Georgia, Louisiana has a run-off should no one in its Senate race get above 50% of the vote. Unlike Georgia, the state parties do not have a primary before the November election, and consequently there are two different Republicans who will receive a significant portion of the vote on Tuesday. Rob Maness, the more conservative of the two choices, is polling between the high single digits and the low teens, but the other Republican, Bill Cassidy, will ultimately be in the runoff with Democratic incumbent Mary Landrieu.
Landrieu hasn’t led a non-partisan head-to-head poll against Cassidy since July, and odds are that between November and the December run-off that voices on the far-right will work diligently to unite conservatives around Cassidy. There’s still a month left before this election is over, and a month is an eternity when it comes to campaign dynamics. But it would be unprecedented for Landrieu to overcome these odds. In 2002, Landrieu was barely able to secure victory in the run-off, and only after unprecedented turnout among African Americans in the heavily Democratic New Orleans. That was when Louisiana had a Cook Partisan Voting Index of Republicans +1, and today, that number is +12. Louisiana is a Republican state, and barring something truly extraordinary, this election is for Cassidy (R) to lose.
Arkansas and Kentucky
These races deserve mentioning because they looked like they might have been competitive for quite some time, but simply never turned out to be. The stories behind the two states are remarkably similar. Mark Pryor may be an incumbent from Arkansas, and Alison Lundergan Grimes may be a challenger from Kentucky, but they are both Southern Democrats in a time when Southern Democrats are a dying breed. They both had a blip of hope due to favorable polling in the spring, but they were also both up against formidable opposition in the form of Tom Cotton and Minority Leader Mitch McConnell, respectively. Once Cotton and McConnell’s campaigns were in full stride, they obtained a solid lead in the polls and never looked back, with the most recent polling suggesting that the two Republicans may win by as many as eight or nine points. Every other race on this list has at least some uncertainty, but these races look like they’re already over, with Cotton (R) and McConnell (R) seizing the victory.
What it all means
When all of these Senators are added to those who are not up for reelection and those who are in uncompetitive races this cycle, the Republicans have 52 seats, the Democrats have 45 seats, and Independents have 3 seats.
That still begs the question of with whom the three Independents will caucus. Socialist Bernie Sanders of Vermont is guaranteed to go with the Democrats, but Maine’s Angus King and Kansas’s Greg Orman have both been open to the idea of caucusing with either party. In a vacuum, both King and Orman would probably prefer the Democrats, but the majority has substantial power in the Senate, and both Independents would be very tempted to jump on the Republican bandwagon.
It’s also entirely possible that the two Senators could team up with fellow moderates to form a true moderate caucus that leverages its power to fundamentally change the dynamics of the Senate. But if no moderate caucus forms and both candidates decide to simply leverage their caucus decision in a way that maximizes their political clout, a switch is unlikely. King has four more years left in his term, and if elected, Orman will have six.
The Senate map in 2016 is as unfavorable to Republicans as this one is to Democrats, so it’s possible that the Republican Senate majority would be very brief. And if King and Orman decide to be Republicans now, it seems unlikely that their party would welcome them back with open arms should the Democrats get the majority two years later. Both candidates would also inevitably have to win a three-way general election should they switch to the Republicans, but would likely face little to any challenge from the left should they stay with the Democrats. King and Orman will probably drive a hard bargain with Harry Reid, and like all negotiations it may fall through. But it ultimately seems likely that the prospect of lucrative long-term committee seats will keep King and Orman with the Democrats, making the final Senate numbers 52 seats for the Republicans and 48 seats for the Democrats.
