Combined 2014 Senate Rankings and Guide
This project took far longer than I expected, culminating in over 10,000 words. I did however feel the need to be comprehensive. Predictions and listings are a dime a dozen these days, and one of things I wanted to make sure of is that I was not solely basing my predictions on polling averages or fundamentals. While I appreciate the greater interest in the nuts and bolts of politics that has come with the quantitative revolution, I feel that it is ultimately limited by the lack of any sort of consistent quality control for inputs. Left to myself, I will always prefer actual numbers, like early voting figures and registration changes, over pollster speculation, and if the results of a poll fly against consistency, common sense, and history, it should not be averaged but set aside.
I may have come to similar conclusions in the end as most other ranking sites, but the order in which I see seats falling and how I expect that to proceed is very different. I tend to think that the electoral legal environment, national factors, and momentum, all currently favor Scott Brown in New Hampshire, while I also suspect that Republican Cory Gardner in Colorado peaked too soon. Yet momentum is different from winning, and I have identified in my predictions of both races that polling and the overwhelming consensus of prognosticators believe that Democrat Jeanne Shaheen will be reelected in New Hampshire, and Democrat Mark Udall will be comfortably defeated in Colorado. The seats I feel more comfortable on are the ones with the greatest early voting data; early results from North Carolina make me more confident in my call for Hagan, while those in Georgia and Louisiana confirm in my mind that the moment of crisis when Democrat Mary Landrieu of Louisiana might have fallen without a runoff, and Republican Perdue forced into one in Georgia, are likely past. My views have also changed over time. The last few days have made me more optimistic about Mark Begich's odds of survival, and I now can see both North Carolina and Iowa falling before Alaska.Things can move, and I reserve the right to change my guesses as more data comes in, though I promise to highlight where and when I have done so.
In any event, here is the summary, because for some people, all that matters is the numbers.
Republican Gains: Arkansas, Alaska, Colorado, Iowa, Montana, South Dakota, West Virginia
Democratic Gains: None
Upset Picks: Scott Brown(R) wins New Hampshire, Mark Udall(D) wins Colorado. Bonus: Charlie Crist(D) loses FL Gov.
Projection for New Senate: 54 Republicans, 44 Democrats, 2 Independents
I should note that I feel fairly good on the 52 Republican scenario, which occurs if they win 2 of New Hampshire, Kansas, and Iowa. As I feel in all probability they will lose at least one, and believe they will also lose North Carolina, I tend to think 51 is as likely a number as 53 if not more so.
Incumbent: Mark Pryor(D),Previous Margins: 82%, 54%Obama %: 37%, 39%RCP Average: Cotton(R) + 5.5%Outcome: CottonCertainty: High
Probably few candidates will have as great a contrast between their 2008 and 2014 reelection challenges than Mark Pryor. Scion of a political dynasty, Pryor was unopposed by a Republican in 2008, facing only a Green opponent. But few states have undergone as great a transformation as Arkansas. Uniformly Democratic as late as 2006, Obama underpeformed even the Massachusetts liberal John Kerry substantially, managing only 41% of the vote in 2008, down from Kerry's 44%. In 2010 incumbent Democratic Senator Blanche Lincoln went down to a 58-37 landslide defeat, while Republicans gained two out of three Democratic congressional seats. In 2012 they took control of both Houses of the state legislature for the first time since reconstruction along with the remaining congressional seat, and this year Mark Pryor was challenged by Tom Cotton, a two-term congressman.
Cotton is something of a favorite of the national conservative media because he is an oddity in the south. Rather than a good old boy, Cotton is a graduate of Harvard and Harvard Law School and a military veteran. Strong academic credentials do not guarantee moderation in the South, witness Ted Cruz of Texas, veteran of Princeton and Harvard Law School, but the combination has led to a lot of gushing over Cotton, gushing that evidently persuaded himself to jump into the race ahead of several other Republicans who had arguably been waiting in line.
Despite Pryor's status as an incumbent, the transformation of the state probably made Cotton at least a slight favorite from the start. Despite those odds, Pryor has made much more than a pro forma effort, convincing former Congressman Mike Ross to run for Governor, and hitting Cotton for being young, untested, and out-of-touch, lusting after national fame. Its hard to measure the state of this race, or the changes it has undergone given the poor quality of polling in Arkansas. The Real Clear Politics average listed above is heavily skewed by the results of a recent Talk Business poll showing Cotton with an 8 point lead, a shift both from their previous results and other polling generally showing Cotton ahead by 3-4. Democratic internal polling here, which the party has not been shy about releasing, has on the contrary shown a very different picture with Pryor generally ahead by a single point.
If one is to buy the Democratic numbers, that has been the state of the race at least since June. By contrast, the public polls have portrayed something of a roller-coaster, with Cotton surging into a substantial lead in the late Spring only to fall behind in late summer, and then surging into a mid-single digit lead again in the last month. So much of the narrative of the race is entwined with which picture you buy; if its the public polling then you assume that Pryor made progress when Cotton looked to be taking the race for granted and focusing on national issues, only to fall behind again as Cotton retooled his campaign and reengaged.
I tend to buy the stability argument presented by the Democratic pollsters, albeit with the caveat that I believe Pryor has been consistently down by a point or two in the low to mid forties rather than up by that margin. In either case I don't think it matters. While the undecided voters distrust Cotton for his national ambitions, their hostility to Obama and inclination to vote Republican for everything else makes me think that Pryor will have a hard time winning a large portion of them, a necessity if he hopes to actually prevail. Hence I think Cotton will almost certainly be the next Senator from the state, albeit by a disappointing and underwhelming margin. That probably won't halt his national momentum, but I suspect that when a young Tom Cotton is tapped as VP for a flailing Presidential campaign and billed as the new "wunderkind" of the party, people will look back at his underwhelming performance here and be able to say I told you so.
Incumbent: Mark BegichPrevious Margins: 49%Obama %: 41%, 39%RCP Average: Sullivan(R) + 4.3%Outcome: Sullivan(R)Certainty: Moderate
The reelection bid of Democratic Mark Begich of Alaska always promised to present an interesting experiment. Democrats, while more competitive than one might think given Alaska's Republican history in Presidential races, have struggled to win statewide for the last forty years. The only two exceptions were former Governor Tony Knowles, who on a three-way race in 1994, and barely scraped by with 51% against a write-in opponent in 1998, along with Begich himself, who managed a razor-thin 2% win against indicted incumbent Ted Stevens in 2008. Otherwise Democrats have been often the bridesmaid, never the bride. In 2002 Knowles' lieutenant governor, Fran Ulmer, was seen as running a strong race against then-Senator Frank Murkowski, only to lose 56-42. Two years later it was Knowles' own turn. Considered the favorite against appointed Senator Lisa Murkowski, whose selection by her own father was seen as symbolic of nepotism, he failed to close the deal 49-46. Two years later his political campaign came to an end with a 49-45 defeat against Wasilla Mayor Sarah Palin.
Begich therefore always had the example of history arrayed against him in his effort to secure a second term. Some Democratic analysts comforted themselves that the supposed Democratic bias in Alaskan polling was in fact a pro-incumbent bias, a claim that might be plausible in a state where every citizen receives a check for oil royalties from the government if it was not for Knowles' under-performance in his 1998 reelection bid. Others pointed to Begich's strong record of constituent service, as well as Alaska's hostility to the Tea Party; Lisa Murkowski won reelection as a write-in after losing the Republican primary in 2010.
In the end Democrats didn't get their wish of a Tea Party opponent, even if it appeared at the time that they dodged a bullet when Governor Sean Parnell decided to forgo a bid in favor of reelection. With the benefit of hindsight, Begich might have been better off with Parnell, who now appears poised to lose his own reelection campaign against a Sarah Palin-backed, Tea Party supported Democrat-Independent Fusion candidate, a set of adjectives that should serve to illustrate exactly how odd Alaskan politics can be.
Probably not odd enough for Begich to prevail however. His opponent, former State Attorney General Dan Sullivan is about as bland as one can get, and Begich, who had hitherto gone almost five and a half years without a mistake, blundered when he ran an ad accusing his opponent of failing to prosecute a criminal only to have the victim's family demand he pull the offending piece. Since then Begich has generally polled behind, with the singular exception of a bizarre late October poll purporting to show him up 10.
Democrats speak of Begich's unprecedented field operation, the underpolling of native communities, and previously mentioned affection for incumbents Alaskans purportedly hold. The actions and spending of both parties tell a somewhat different story.While a degree of uncertainty hangs over this race, as it does over any Alaskan contest, especially one where votes will be cast for hours after the first results in the battle for the Senate will be released on the East Coast, the general consensus seems to be that Sullivan likely has this. I put this race nonetheless in the "moderate" category due to the uniquely Alaskan factors, and uncertainty I have heard from Republicans who have visited the state, who have expressed worries about the race despite the overwhelmingly favorable metrics.
Incumbent: Mark UdallPrevious Margins: 55%Obama %: 52%, 54%RCP Average: Gardner(R) + 3.8%Outcome: Gardner(R)Certainty: Low
Most years there is always an incumbent who looks safe only to run into trouble down the stretch. In this case it is Mark Udall of Colorado.
Democrats began the year unconcerned about Colorado. Once a solidly Republican state, George W. Bush carried it in both 2000 and 2004, both Senate seats were held by Republicans between 1995 and 2005, it had turned so dramatically to the Democrats in recent years. Democrats won a Senate seat in 2004, both houses of the legislature and the Governorship by landslide margins in 2006, and Republicans struggled to put up even a halfway decent challenger to Udall who slipped into the Senate by double-digits in 2008. Even in 2010, as a Republican wave was sweeping the nation, Democrats John Hickenlooper and Michael Bennet defied the polls to win reelection, and Republicans were left to console themselves with a one-seat majority in the State House which was wiped away in 2012 when Obama easily carried the state.
Democrats, facing the daunting prospect of a map in which no less than six Democratic seats in Romney-won states were up therefore had plenty of reason to write Colorado off. Initially the Republican party seemed to do so too, with the most prominent candidate in the race on their side being Ken Buck who had managed to lose the 2010 race against Bennet. It was only the last minute entry of Congressman Cory Gardner that transformed the situation.
At first Gardner's entry worked less magic than was hoped. First elected in 2010, Gardner represented a rural one-seventh of the state, and had a solidly conservative record. He struggled to keep up with the incumbent in the fundraising race, all important when he was much less well known than his opponent. But then Udall offered Gardner an assist.
Ever since Senate Majority Leader Harry Reid returned from the political dead in 2010 by using negative adds to first select his Republican opponent by nuking her competitors, and then to defeat her by destroying her own numbers with attacks, it has been part and parcel of Democratic senate strategy to go hard negative against undefined Republican opponents. The 2012 elections added an extra-layer of this by providing an issue, namely abortion/contraception, on which numerous lesser known Republicans had taken relatively extreme positions on in order to appeal to GOP primary voters. Yet what worked in the low-profile of Republican primary for state legislature or congress went over much less well in statewide general elections.
Like Todd Akin and Richard Mourdock, Cory Gardner had a liability on "women's issues", namely his support of a so-called "Personhood" Amendment to the Colorado state constitution which would define life as beginning at conception. Udall's campaign, having witnessed the success achieved by Claire McCaskill against Akin, and of Joe Donnelly against Mourdock, apparently concluded that the issue could be used to eliminate Gardner before he could become a threat. And so the Udall campaign went up with a massive barrage of negative ads assaulting Gardner's position on abortion.
But something unexpected happened on Udall's road to inevitable victory. Gardner, rather than maintaining his position and in the process committing a fatal or near fatal gaffe, chose to evade the issue by taking a letter from the Supreme Court's recent ruling that Hobby Lobby, a Christian-owned company, did not have to compromise its principles by paying for its employee's birth control, Justice Kennedy, in writing for the majority, suggested that the burden placed on Hobby Lobby was excessive because alternative means existed of ensuring that the employees in question had access birth control. Gardner chose to take the matter a step further; while opposing the government provision of birth control, he came out in favor of making it legal over the counter at pharmacys.
This did not deter Udall, but it did blunt the efficacy of the attacks. Furthermore, it created a situation in which Udall's attacks began to backfire. Despite prominent examples of negative campaigns that have proved highly successful, negative campaigning is almost as high-risk as it is high-reward. Voters tend to dislike the attacker regardless of whether they buy the attacks themselves, and as a consequence failed attack ads can put a campaign out of its misery, as Jack Conway, Rand Paul's opponent in the 2010 Kentucky Senate race demonstrated when he repeated allegations that Paul at college had kidnapped a girl and first her to pray to "Aqua Buddha".
That attack demonstrates another issue with negative ads; they have to seem plausible to viewers, and hence have to somehow fit into a preexisting narrative about a candidate. The attacks by the Swift Vote veterans were effective not so much because people believed that John Kerry threw someone else's medals over the White House Mall at a protest in order to save his for a future campaign, but because a narrative had been constructed about John Kerry such that people believed that it was something John Kerry could have done. The gaffes that destroyed Todd Akin and Mitt Romney in 2012 were so deadly not because they were important on their own, but because they added credibility to a host of other attacks by making them seem plausible.
By contrast, one of the least successful negative campaigns I have ever witnessed was that run against Scott Brown in the Massachusetts special election in 2010. With the race not emerging onto the national radar until its final two weeks, commercials were filled with nothing but campaign adds, as both parties tried to see what would happen if they spent $30 million in the space of ten days. On an average commercial break from the local news I witnessed four political ads back to back, one from Scott Brown and three from Democratic groups attacking him.
Despite being outnumbered, Brown emerged much stronger from that barrage than he entered. The attack ads were overly dramatic, using the same black and white coloring, and the issues they covered were to broad ranging from abortion, to global warming, to labor issues, to taxes. By contrast the single Brown ad featured the candidate in his kitchen with his daughter smiling and mocking the negativity of his opponent. The over-the-top imagery and music of the attacks when contrasted with the good-humor of Brown's spot not only wrecked their effectiveness, but added credibility to Brown's ad. If Brown was telling the truth about the absurdity of the attacks, then might he not be telling the truth about other things?
Udall seems to have fallen into the same trap. Gardner, despite Democrats' best efforts, has steadfastly refused to commit suicide in an interview, and in the absence of a gaffe that can be used to establish his "extremism", Udall has been forced to run ever more negative ads. These in turn have legitimized Gardner's argument that Udall has nothing to show for six years in the Senate; afterall, if all an incumbent Senator can do is say bad things about his opponent then clearly has nothing to brag about regarding his own record.
Worse than that, Udall has managed to appear negative, mean, and most importantly, part of the problem. Voters currently dislike both parties, and rightly or wrongly they view extreme partisanship and the negativity of politics as being one of the key reasons for dysfunction. The example of an incumbent US Senator constantly trashing his opponent with cookie cutter ads makes Udall appear as about as generic a politician as is possible.
For evidence of this all one has to do is look at recent polls. Quinnipiac has Udall at 42-49 favorables compared to Gardner's 47-41. PPP found him in even worse shape, with 37% viewing him favorably to 52% viewing him unfavorably even as he only trails by 3. The latter numbers are not something that happens naturally, but are generally only seen in terms of blowback from a negative campaign. Udall is not being dragged down by his party, but is rather being propped up by it despite being widely loathed.
It is hard to separate the spin from the reality in the case of Colorado. Polling in the state has a history of understating Democratic candidates. Polls showed Democrats losing both the Governor and Senator races in 2010, and since then Colorado has switched to an entirely vote-by-mail system in which every voter in the state receives a ballot in the mail rather than having to vote. Democrats claim that this increases turnout among less engaged groups such as minorities and young voters, rendering traditional likely voter models, and the pollls that use them, inaccurate. As a result the national party claims confidence, though whether that is the cause of their lack of investment, rather than resignation, or the belief that there is little to be done given how badly the existing negative barrage against Gardner has fared is an open question.
By contrast conservative commentators have been gleeful, chalking Colorado up as a win, and all but taking it off the table, National Republicans have been more cautious; while outside groups have poured in millions, the NRSC has continued to display caution, much greater caution than they have displayed in Iowa.
As Colorado's vote by mail system transforms election day into a two-week process, the early voting statistics can provide a clue, albeit a limited one. As of Thursday, a little over half a million voters had cast ballots compared with a total turnout of 1.8 million in 2010. Republicans lead returned ballots 43.8% to 31.7%. Yet caution is required. Updates are on a county-by-county basis with no obligation that counties update on a daily or even regular basis, and larger counties like Denver both sent out ballots later and have generally not updated as often. Furthermore, as encouraging as those results may look, so did the 2010 numbers which showed Republicans leading 40.7% to 34.6%.
This is one race where I reserve the right to change my call later in the week as more numbers are available. More than two thirds of the Colorado vote was cast early in 2010, and a similar percentage should be available for party registration analysis before election day. Udall will lose anything resembling the electorate shown in the polls, not least because of its Republican nature, but also because the poor reviews he gets from the electorate as a whole make his supporters less reliable and undecideds sour on him. But the fact he is trailing at all is a product of the likely voter models used by those polls. The recent Quinnipiac poll showed him tied with Independents and doing better with Democrats than Gardner is with Republicans. It is almost unheard of for a Democrat to lose with that breakdown in a purple state. Hence despite Udall's poor campaign, I cannot have more than a low degree of confidence here.
Incumbent: Open Seat, Republican, Candidate David Perdue
Previous Margins: 49%(57% runoff), 53%, 49%
Obama %: 45%, 47%
RCP Average: Nunn(D) + .4%
Democrats have been obsessed with demographics since John Judis and Guy Teixeira published his 2002 book, the Emerging Democratic Majority, which argued that the growth of non-white and educated white voters would eventually create a permanent national Democratic majority. The timing was unfortunate; published after the razor-thin 2000 election, it had the misfortune of preceding the 2002 and 2004 elections when it appeared that it would be a new Republican national majority based in the upper Midwest, rather than any sort of Democratic one, which would control the nation's destiny.
Yet even in 2004 there were signs of movement. National turnout jumped nearly 20%, and it was the Democrat's misfortune that increased turnout by non-white voters was more than matched by the Bush campaigns successful mobilization of Evangelicals and wooing of Southern and blue-collar whites. In the end, the Democratic pool proved to be deeper. The Democrats further increased their vote totals while the GOP numbers stalled in 2008, and Mitt Romney failed to match even Bush's 2004 numbers in 2012.
While the focus of the 2002 book was on Texas, a prize that long tempted liberal Democrats due to its connection with the Bush family, the prominence of its liberal tradition, and the pleasure that would ensure from winning the Conservative counterpart to California, Democratic efforts came ups short. No Democrat has won a statewide race there since 1994, an exceedingly poor record in a state which has 32 such offices, and there is little prospect of that changing this year.
But Democratic hopes in Texas were always built upon feet of clay. The demographic change their was based upon Hispanic growth, and Hispanic voter turnout rates are abysmal, even at higher educational and income levels, and both levels are exceedingly poor among Texas Hispanics. This was demonstrated in 2012, when Obama fell back in Texas, both in absolute terms and relative to the nation at large. By contrast, two of the states he improved in, Mississippi and Louisiana, were uncontested Deep South states with a large African American population.
This year that experience has led Democrats, after a disappointing flirtation with Wendy Davis' ill-conceived and even worse executed campaign for Texas Governor, to focus their hopes and efforts on a different state, Georgia. No state has seen faster change in its electorate as opposed to its total population than Georgia, which as of 2000 still had an electorate which was 77% white, only to see it fall to 60% in 2012. At the same time, the rapid growth of Atlanta has provided a source of white voters who are at least open to considering backing a Democrat.
Georgia Democrats, unlike their unlucky compatriots in Texas, have been gifted with a much stronger statewide slate than their Texas counterparts. Headed by two legacy candidates in State Senator Jason Carter, grandson of the President, and Michelle Nunn, daughter for the former Senator, it also includes Valerie Wilson, their candidate for Superintendent of Education, who has been endorsed by the Republican incumbent, and Greg Hecht for Attorney General, who lost that office in 2010 by 9%, the closest statewide margin in that awful year.
The Republicans in turn have been saddled with mediocre nominees. Incumbent governor Nathan Deal, a former Democratic congressman from Rural Georgia who switched parties in 1995 and limped into office in 2010 amidst allegations of ethics violations widely blamed for mishandling a major snowstorm last year, while Senate nominee David Perdue has been an unspiring gaffe-machine.
This contrast allowed both Carter and Nunn to take the lead last week, and if Georgia had normal election laws, I might conceivably have placed their races as toss-ups. But Georgia does not. Georgian law requires a runoff if no candidate in a statewide race achieves 50%, and unlike in Louisiana where these have generally been a godsend for Democrats, in Georgia they have been the bane of their existence, costing them a Senate seat as far back as 1992, and a number of offices since. To confuse matters more, assuming both the Senate and Governor's races go to runoffs, there will be two separate runoffs, one for Governor in December, and a Senate one on January 6th.
This places the odds facing both Democrats on another level, and has led many prognosticators to write them off even before election day. This may well have been premature. The same factors that have made Deal and Perdue bad candidates in the general would also presumably apply in a runoff; in contrast with 2008, when a clear majority of Georgians wanted a Republican in the Senate, the operative dynamic would instead be a clear majority not wanting Nathan Deal to be their governor. Nonetheless, other factors, such as the awkward January date and the fact that Nunn and Carter seem to have fallen in the polls and now seem likely to finish second leads me to conclude that their efforts are likely to be a bridge too far, at least for this year.
The degree to which Nunn is being written off in a runoff is excessive. Democrats have a decent record in federal races elsewhere in the South, including in areas where they have far less organization. Travis Childers improved his margin in the 2008 Special Election runoff in Mississippi's first district, while Democrats only narrowly lost heavily Republican Louisiana districts in late 2008. The major contrary examples, such as the 2008 Senate runoff where a 49-46 Democratic loss turned into a 57-43 rout, took place in situations where the Democrat had come second in the first round by a sizable margin and clearly had no chance of coming ahead of the Republican regardless of the electoral system utilized. If Nunn were to beat Perdue 48-45 or 49-45 in November the situation would be very different.
Different but different enough? A lot would depend on what happens elsewhere and who controls the Senate. Democrats will also face turnout and motivation issues especially given the timing. But in past years the national political climate has turned against the Party doing well in November, and right now that looks like the Republicans. It is plausible that the Republican party will be in far worse shape in January than it is now. All in all, I rate Perdue a favorite, but a narrow one, to be Georgia's next Senator.
Incumbent: Tom Harkin (D) - Retiring
Previous Margins: 63%, 54%(2008), 52%, 51%, 51%Obama %: 52%, 54%RCP Average: Ernst(R) + 2.1%Outcome: ErnstCertainty: Medium
Here is the second place where I part ways with the conventional wisdom on a race, though which conventional wisdom I am bucking has become a bit ambiguous given the many phases this race has gone through. For most of the year, both parties ignored this race. Democrats, complacent in their success in winning federal office in Iowa, which has neither matched that of their Colorado compatriots in the last decade, nor extended regularly into congressional races, believed that Iowa was a slighly left-leaning state and they would win with a generic candidate. Republicans failed to get their top pick in retiring Congressman Tom Latham and were forced to fall back on what appeared to be a pool of B-Tier candidates.
Unfortunately for Democrats, this sense of complacency appeared to extend to their candidate, first district Representative Bruce Braley, who bragged about being a proud trial lawyer in a rural state, and whose campaign was wracked both by his own unpreparedness, and incompetence. By contrast, the national media fell in love with Sarah Palin endorsed State Senator Joni Ernst. Always on the lookout for minority and female Republicans, and increasingly reliant on Republican insider leaks as Democrats have generally stopped leaking to them in favor of direct engagement overt social media, the major prognosticators dubbed her a "rising star."
Democrats were incredulous. Enrst had the backing of Sarah Palin, had a record of hard-core conservatism including support for fetal personhood and the suggestion that defense against the government was a legitimate ground for firearm ownership. But nonetheless she was surging and then leading.
About two weeks ago that lead reached a point at which most prognosticators begin to write the race off. Unfortunately for Ernst, so did the NRSCC and evidently her campaign, yet the election was not last Tuesday but instead next Tuesday.
The result was a predictable outcome; a candidate who has peaked too soon. Supremely confident, Ernst began to aggressively bite the hand that fed her, walking out of meetings with two editorial boards, including the Des Moines Register which had endorsed Mitt Romney in 2012. Not only did this guarantee a Braley endorsement, but it generated several days of coverage, which in the absence of any degree of activity on either her or Braley's part, the latter for once showing a proper appreciation of his limitations, became the narrative for the week. Suddenly, outside liberal groups, ranging from the ant-gun organization chaired by Ex-Rep Gaffy Giffords, to the DSCC and Environmental groups unleashed a barrage on her. Democratic early voting figures, trailing GOP returns, tracked upwards this week, passing 2010s total vote today with several more days left.
Ernst has been lucky in that the polling picture over the last few days has been mixed enough to prevent the obvious sequel to the events of the last week, "Ernst Campaign in trouble?" from taking over the airwaves. But her coverage has been almost uniformly negative in the local press.
Is it too late to matter? Maybe, but Democratic organization in the Hawkeye state is not to be underestimated, and contrary to appearing a rising star, Ernst has been pricklish and defensive in response to criticism. As in New Hampshire current polling points to a narrow Ernst lead. But coverage, early voting figures, and the behavior of outside groups creates a more complicated picture. Ernst is probably still marginally up right now, likely by 4 points or so, but that likely down from a 8-9 point lead ten days ago, and the nature of Iowa's voting system; early voting last three weeks, means that she seems to have peaked a few weeks too early. Whether that will be enough for an uninspiring Democrat is unclear
Incumbent: Pat Roberts
Previous Margins: 61%, 81%
Obama %: 38%, 42%
RCP Average: Orman(I) + .4%
Kansas is a strange place for Democrats to play offense. They have won exactly one Senate race this century, in the Roosevelt sweep of 1932(though they came within 2% of knocking off Bob Dole in 1974), and the Democratic party was decimated in 2010. It has no Congressman, and no Statewide officeholders. Even in the strong Obama year of 2008 when they ran former Congressman Jim Slattery he managed only 36%. While they managed to recruit a candidate, Shawnee County DA Chad Taylor, it seemed highly unlikely they would make a serious play for Kansas.
Yet forces are in motion within Kansas this year that have upset everyone's calculations. The 2010 sweep brought Senator Sam Brownback, a fiery conservative to the governors office, with an agenda to bend every state institution from the schools to the courts to his will. When the Republican leadership in the State Senate resisted his agenda, he recruited primary challengers and ousted eight State Senators in 2012. With full-control. he sought to eliminate almost all of Kansas' merit-based and nonpartisan institutions. Merit selection for judges, a bane of conservative activists who believe that it produces "liberal judges" was done away with for all but the Kansas Supreme Court in favor of political appointments, while extensive tax cuts combined with substantial budget cuts were instituted. Had this policy produced results, Brownback might have been vindicated. Instead it produced record deficits, and the denial of the Brownback administration threatened to see the state downgraded to junk bond status. In the Republican primary Brownback faced the unknown Jennifer Winn who raised only $13,000 but nevertheless won 37% of the vote. For the last few months polls have shown him trailing his Democratic challenger Paul Davis.
Pat Roberts who has never had a close race also seemed to fall victim to the headwind, not least as questions were raised as to whether or not he was even a resident of the state he claimed to represent. Facing a challenge from activist Milton Wolf, who he narrowly beat with anemic performance under 50%. The picture for the general election was complicated by the presence on the ballot of independent businessman Greg Orman who promised to spend millions of his own money. That image however was transformed when Democrat Chad Taylor withdrew from the race, endorsing Orman. Orman rapidly surged to a lead, aided by the decision of the Roberts camp to spend several weeks fighting in the courts to keep the Democrat on the ballot against his will rather than actually campaigning against their independent opponent.
Roberts fell behind by mid-September, but over the following weeks he began a slow recovery, helped by attacks on Orman for being a Democrat in disguise. These charges were challenging for Orman to rebut, given that he had been endorsed by the Democratic candidate, and was receiving logistical support from the Kansas Democratic Party. Slowly Roberts, along with Brownback recovered into small leads.
In the last ten days that process seems to have reversed as Orman has regained a slight lead, and Davis a larger one. I now feel that Davis is likely to win - Brownback is too disliked, partisan politics matters too little in a state race, and even the polls that showed Brownback close also showed him hated. Orman on the other hand has a tougher road. He is currently being overwhelmed 5-1 on the airwaves by Roberts and Republican groups, a testament less to the total resources available to the sides than to the advantage that experience can provide with securing the best rates for ad placement. As problematic, Orman as an independent lacks organizational muscle. Roberts can ride Brownback's mobilization of his evangelical base, but Orman is dependent on the Democratic turnout effort whilst being unable to coordinate openly with it for fear of providing credence for Roberts' attacks. This leaves him at a disadvantage.
It is not an overwhelming disadvantage, especially if the turnout is particularly adverse for Brownback. Given the intensity of the Governor's detractors, that is a real possibility. But for now I am betting on a Davis/Roberts split decision.
Incumbent: Mitch McConnell(R)
Previous Margins: 53%, 64%, 56%, 52%
Obama %: 37%, 41%
RCP Average: McConnell + 4.4%
A good target makes for good fundraising, especially if the target in question carries emotional appeal for a party's base voters. In 2006, that target was Rick Santorum, the socially conservative Pennsylvania Senator whose race continued to draw substantial financial investments from activists on both sides even though for all practical purposes the race was over nine months before election day. For Democrats in what is otherwise a defensive year, Senator Majority Leader Mitch McConnell is that target. Not only did McConnell play a decisive role in frustrating Obama's legislative agenda in 2009-2010, but targeting him allowed Democrats to seek payback for Republican efforts to target their leaders, defeating Tom Daschle in 2004, and failing to beat Harry Reid in 2010.
This is not to say that there was no strategic calculation behind the decision to go after McConnell. Democrats had few viable targets this year, Kentucky remains relatively Democratic downballot, and McConnell underpeformed against a C-Tier opponent in 2008. Democrats got their choice in Secretary of State Alison Grimes, and polling showed, and continues to show McConnell unpopular. Had this been 2008, the candidate and the resources invested might well have been enough to knock off the Republican Senate Leader.
Alas for Democrats, they are trying six years too late, after an Obama term and a half of erosion of their political base in Appalachia. Obama's unpopularity generally adds to their challenges, as does in the increasing prospect that Democrats will lose the Senate anyway. That the nature of the year means that rather than taking seats safely off the table, with the prominent exception of Michigan the reverse has happened and their exposure has increased as the year has gone on. That has left less and less resources available for an emotionally satisfying long-shot bid in Kentucky.
This is not to say Grimes has done much wrong. On the contrary she has acquitted herself well in what is a difficult year, and resisted the urge to self-destruct in a surge of desperation as State Attorney General Jack Conway did in 2010. But running a good campaign only goes so far when McConnell has also run a competent if somewhat uninspired campaign, and Barack Obama is sitting at 31% or lower in the state. Grimes will perform credibly, likely in the 44-46% range, and McConnell will win another clear but not overwhelmingly victory with 52-54% of the vote.
Incumbent: Mary LandrieuPrevious Margins: 52%, 52%Obama %: 41%, 40%RCP Average: Cassidy(R) + 4.8%(Runoff)Outcome: Cassidy(R)Certainty: Moderate
Louisiana, like Georgia has a runoff if no candidate achieves a majority in the general election. Unlike in Georgia however, the runoff has not been the consistent bane of Democratic candidates. May Landrieu has won two of her three victories in overtime, triumphing by 52-48% in both 1996 and 2002, before managing that same 52% in 2008's general.
This is important when considering her prospects for a fourth term, since Louisiana seems almost certain to go to a December runoff. Depending on the pollster, Landrieu seems to be sitting anywhere from 39%-46% in the general, with a series of Republican challengers led by Congressman Bill Cassidy leading the pack.
Most prognosticators have generally either written Landrieu off, or consider her prospect poor. Part of this is because of the likelihood that the majority of voters who cast ballots for a Republican in the first round can be expected to rally around the single GOP candidate in the runoff, even though this did not happen in 2002, when a 54% GOP vote in the general fell to 48% in the runoff. Others have suggested that either control of the Senate will be at stake, in which case voters who like Landrieu will cast votes for Republican control(plausible) or for the majority party, though the latter did not stop Landrieu from winning in 2002 after the Democrats lost control. A third reason is the collapse in the state Democratic party, Obama with his 40% in 2008 and 41% in 2012 did worse than even John Kerry who managed 43%.
If so, Louisiana whites would only be following the example of their compatriots in Alabama, Mississippi, Arkansas and Texas. Yet unlike those states, the rebirth of New Orleans as a hipster center has drawn in a substantial number of more liberal whites along with an African American population that now exceeds pre-Katrina levels, one reason why the state was one of the few where Obama improved on his 2008 performance. A big question is however whether they will both to vote.
My personal viewpoint, backed by provisional early voting data that the answer for African Americans will generally be a YES. Their turnout has been on an upward trajectory since 2004 in any case, and they are currently making up about 31% of early votes, close to their population level. Liberal whites, especially younger ones, seem by and large to be disengaged.
That will likely pose a serious problem for the Landrieu campaign, albeit one they may get two chances to solve. One of the most important dynamics of the runoff, which is a well-known institution in the state as opposed to the situation in Georgia, is that it gives the electorate a chance for reflection and an opportunity to reconsider their choice. This separates the election from the national climate. Landrieu almost certainly would have lost the 2002 election in the Republican climate of November 2002; by the time the runoff rolled around Trent Lott was deeply involved in the Strom Thurmond scandal that would end his leadership, and Landreiu was able to triumph. By contrast, David Vitter may well have lost a runoff in December of 2004 had one been held then; by that point George W. Bush was already running into trouble, and there was no John Kerry to tie Vitter's foe to.
For 2014, a lot will depend on what happens between November and December. If Republicans see a victory as an excuse to discuss impeachment or another government shutdown that could aid Landrieu by mobilizing Democrats; if by contrast the situation worsens for her party, she will have serious problems.
Given the importance of the national environment in December, and the fact that so much of the nature of that environment will depend on the outcome of the November midterms, it is impossible to project Louisiana with a significant degree of confidence. What I can say is that the incumbent should come first in two weeks; if she does not, she will face serious problems six weeks later. I can also say with a degree of confidence that she would very likely lose a two-way race if one were to be held on November 4th. For that reason I give an edge to Cassidy, even as I consider the possibility that while the bland and uninteresting congressman may be the perfect candidate for a general election in a Republican wave year, he may be a very bad fit for the high-intensity nature of a runoff where the eyes of the nation are solely focused upon the contest.
Incumbent: Jeanne Shaheen (D),
Previous Margins: 51%Obama %: 52%, 54%RCP Average: Shaheen(D) + 2.2%Outcome: ShaheenCertainty: Low
While Colorado may be getting the most credit for being a surprise this year, the idea that Jeanne Shaheen could face a tougher road to reelection than Kay Hagan would have been met with the same sort derision and mockery which marked former Massachusetts Senator Scott Brown's decision to run in New Hampshire two years after losing to Elizabeth Warren in a neighboring state. On the heels of his decision to forgo a special election run which the underwhelming Ed Markey won by a mere 54-44, and a gubernatorial bid that was almost a guaranteed success, Brown's decision to switch states was seen as the likely end of his political career. Nine months later Brown seems to have had a better idea of what he was doing than his critics.
I have to confess to a degree of personal bias here, having been involved with Senator Brown's office in 2012, and to sharing that skepticism. But with the benefit of hindsight, Shaheen's position was always built on what amounted to base of sand. Despite her long political career, Shaheen has never been popular, at least since her 1999 veto of a state income tax which badly split her party. She limped to her third term as Governor in 2000 by an underwhelming 49-44 margin, and then managed to lose a senate race she was favored for in 2002 by a surprisingly large 51-46 margin. Even her 2008 victory was disappointing. Her opponent, incumbent Republican John Sununu, had never fit into the Senate, and seemed to show little interest or faith in his own reelection bid. The NRSCC appeared to share his doubts, all but pulling out in the spring. Nonetheless, while Shaheen eventually reversed her 51-46 defeat, this looked far less impressive when compared with Obama's 13 point victory in the state.
Nor should Democratic confidence have been inspired by the supposed "left-turn" of the state. While it underwent a political revolution in 2006, turning over both houses of the legislature along with both congressional seats to the Democrats, the counterrevolution in 2010 more than matched it, providing the GOP with more than 75% of the seats in the legislature, both congressional seats, and a 60-36 win for Senate over an incumbent Democratic congressman. While Democrats bounced back in 2012, they appeared to see the experience as an aberration rather than a lesson.
This appears to have been a serious mistake. New Hampshire has quite a bit in common with Colorado, another high-income, highly-educated white state that shifted towards the Democrats over the last decade, yet in which the Democrats find themselves in serious trouble in 2014. One commonality is that both states started moving to the Democrats not in 2008 or 2006, but in 2004. New Hampshire was the only state John Kerry won which Al Gore had lost, while he improved on Gore's performance in Colorado. This provides evidence that rather than being a result of "demographic changes" which could serve as an excuse for Democratic complacency, both states actually moved against the Republicans because of the Iraq War. I warned this summer that Obama's actions regarding ISIS had likely inflicted untold damage on the Democratic party, not because they were ineffective per se, but because by buying into the narrative that ISIS was a threat and that it was America's "responsibility" to somehow do something about it, he had thrown away the singular advantage almost every Democrat had held over their Republican opponents on foreign policy for the last decade.
With foreign policy a non-issue, a high-income, white, elderly population, and no early voting of any sort, New Hampshire is a bad state for the Democrats in 2014, arguably almost as bad a state as it was in 2010. In a sense it is more surprising that Brown is not running away with this race. The latter is a testament not to the Democratic sympathies of the electorate, polls show Obama well under his national average at 39% or so in the state, but to all those weaknesses that Democrats convinced themselves made Scott Brown's bid futile. There has not been a single poll taken this year where Scott Brown is viewed favorably by the likely electorate, much less the registered voter pool, while by contrast, Shaheen is viewed by favorably by 8-10 points in most of them. I cannot think of a single major statewide race in history where an incumbent lost or even had a close race with those sorts of numbers,
It was this dynamic that made me certain for months that regardless of what polls showed, Shaheen could not lose to Brown, as his personal negatives imposed a ceiling, albeit perhaps a high one, on his performance that was well below 50%. And if this were a Presidential election year, with another race driving an electorate that somehow resembled New Hampshire's as a whole to the polls that might still be right. But with generic non-entities holding down the Democratic fort in the congressional delegation, a legislative system typified by gridlock, and Obama having done absolutely nothing to motivate them, they can simply drop out of the electorate as they have likely voter samples. And that has been the trend. Rather than Brown increasing his position among registered voters, the Likely Voter samples have become more favorable until one of the most recent CNN polls having him down 49-48 had his ratings among likely voters at 48-50, still below Shaheen's at 51-44, but getting close to the finish line.
Most New England Republicans I have spoken with remain skeptical, but that is motivated by their exposure to a much less well-managed Brown effort in 2012. Shaheen may be the narrowest of favorites but if I was going to pick a place where the averages are going to be wrong next week, it would be here and Colorado. I can actually see a scenario where Udall wins and Shaheen loses, which would have been unimaginable even a week ago.
Incumbent: Kay Hagan(D),Previous Margins: 53%Obama %: 48%, 50%RCP Average: Hagan(D) + 1.6%Outcome: HaganCertainty: Medium
North Carolina is simultaneously one of the most predictable states, and the least predictable Senate races of 2014. Predictable because it is incredibly polarized; not only do racial and ethnic sub-groups vote in a polarized manner, but the white community is also polarized along economic and geographic lines. Combined with perhaps one of the best early voting reporting systems of any state in the country, and the tendency of half or more of the electorate to vote early, and it is quite easy to make an educated guess about the outcome almost without reference to polls. That is a good thing, because using early voting statistics only in 2012, I was able to come up with a guess, Romney +2.6, which was far closer to the actual result than almost any of the polls.
The decision of the Republican state legislature to cut the early voting period in half does not mean that we will lack this data by election day. On the contrary, all signs point to a higher early voting total than in 2010. It does however limit the amount of information available to me right now in order to make a guess, given that I only have the first 330K votes to analyze with. Those can be unrepresentative, with different partisans apt to vote at different times, as has been evident with the eroding Republican advantages in Florida and Colorado.Therefore the current strong Democratic performance could be unrepresentative.
That is only half the problem with trying to make predictions about this year's North Carolina Senate race; the other half is that it features Kay Hagan, who my gut, which is generally right about almost every electoral outcome around the world, has been repeatedly befuddled by. I'm at least not alone here. Hagan, a State Senator in 2008, was the DSCC's first choice, or even in their second tier. Their first choices, all congressman, declined to challenge popular moderate incumbent Elizabeth Dole, while their second-tier recruit Randy Cunningham, was beaten by Hagan in the primary. Hagan then went on not only to beat Dole, but to beat her conclusively by a 53-44 margin. Why almost every prognosticator entered this year asserting that Hagan had somehow ridden Obama's "coattails" when she had beaten him by 8 points is inexplicable, but the only thing that compares with Hagan's extraordinary political skills is the hostility evidently felt against her by the center-right establishment of political journalists and analysts.
The result was that Hagan was written off early this year even before she underwent a nearly $50 million assault by SuperPacs, which sent her personal numbers into freefall. When her leading opponent, State House Speaker Thomas Tillis won the GOP primary against a Tea Party foe without a runoff those commentators rushed to declare that the GOP had learned from its 2010 mistakes and nominating "solid" "moderate" candidates. Ignored were the claims of the Hagan campaign for the past several months that Tillis was their preferred opponent, and played down was the prospect that highly unpopular Republican state legislature would play against him. It was really not until August when Hagan built up what seemed to be a consistent, if narrow lead, that the conventional wisdom truly began to turn.
With the background out of the way, the basics of this race are simple, and remarkably similar to the Florida Governor's race. As with Rick Scott and Charlie Crist, both candidates in North Carolina are hated. They are both disliked by sufficient majorities, their numbers usually featuring 20+ point gaps between those holding favorable and unfavorable views, that the relative extent of that dislike is largely irrelevant. The winner will win not because they are liked more, but because they manage to convince more people who dislike them to nevertheless cast ballots for them, or as important in a midterm year, not sit the election out.
Normally this would favor the candidate closest to partisan balance of the state, which in North Carolina is marginally Republican. That Republican demographics historically have turned out better should also be added to the scales here. Yet things are not so simple. The Republican State Legislature has given a huge assist to Hagan. Unrepresentative of the state, Democrats won a majority of the votes cast for chambers with Republican veto-proof majorities, the legislature is proud of it, and has not limited itself to merely imposing unpopular policies at the state level. It has exercised its power, used against Republicans local governments after reconstruction in the late 19th century, to redraw and re-organize the local municipal governments and even local school boards. The result has been at least on one hand to mobilize Democratic voters. While occasional Democrats, especially younger voters and single women, appear to have been happy to sit out the election out of frustration with Obama's ineffectiveness, the deliberate efforts of the legislature to poke them with a sharp stick seems to have them turning out. The same is true of African Americans whose turnout has been on an upward trajectory for a decade, and have generally seemed to have seen less of a drop-off than white liberals nationwide. The result, is that the very early figures show an electorate that looks far more like 2012 than 2010. That electorate is if anything more Democratic than the one the pollsters showing a marginal Hagan lead have been finding.
There is another similarity with Florida's Governor's race. One of the two hated candidates is running a better campaign. Hagan has been relentlessly on-message, and raised an unprecedented amount of money. Tillis by contrast has been listless, recently flip-flopping even over the Medicare expansion provision for Obamacare. He has not been helped by the actions of Senate President Phil Berger, his counterpart in the upper house who has been a bane of both his and Governor Pat McCory's existence. Berger not only held up a teacher pay rise for months until mid-August, keeping the legislature in session, but forced Tillis into a series of embarrassing compromises. Banned from raising money while the legislature was in session, and looking powerless over his own chamber, Tillis had circles run around him.
Is Tillis doomed? Well there are reasons to believe not. There seems to have been marginal movement in the polls, but whether that represents the entry of a series of poor pollsters working off bad likely voter models into the field or actual movement is unclear. More concerning for Hagan is where she leads. She has generally been stuck in the low-40s with a 2-3 point lead over Tillis. North Carolina has a habit of sharply turning against Democrats in the final weeks of campaigns, and there is little cause for comfort for Democrats until Hagan has won on election night.
That said this is probably the originally competitive seat from the start of the cycle that Democrats can feel most confident about, even if the concept of Mark Udall losing and Hagan winning would have been inconceivable twelve months ago. It is also one where I reserve the right to revisit my thoughts later this week when more numbers are available. But given present dynamics I think Hagan has a decided advantage.
Incumbent: Jeff Merkley(D),Previous Margins: 49%Obama %: 54%, 56%RCP Average: Merkely(D) + 13.5%Outcome: MerkleyCertainty: High
There was a time when certain publications felt that first-term Senator Jeff Merkley faced a serious threat from surgeon Monica Wehby. That time is long past. Republicans have had bad luck in Oregon even in good years, coming up short in the 2010 statewide races, and failing to make a dent in the Democrats 4-1 advantage in the congressional delegation. Wehby has been revealed to have far too serious problems to offer the prospect of making a serious dent Merkley's positions even if this were a comparable year, and without a viable statewide slate Democrats have not even had to hit her particularly hard. This race is included only as a demonstration as how utterly absurd DC/NYC candidate hype can be.
Incumbent: Mark Warner(D),Previous Margins: 63%Obama %: 52%, 53%RCP Average: Warner(D) + 11%Outcome: WarnerCertainty: High
If any illusions have been dispelled this year, it is that Politico, whatever its power on the Washington gossip circuit, has the power to determine Senate races. The entry of former RNC Chair Ed Gillespie into the contest led to a series of articles, echoed by the Wall Street Journal Editorial page and National Journal, that Virginia was on the verge of becoming a top-tier race, a claim Fox News maintained as recently as this weekend. It never quite happened. Few knew who Gillespie was, and those who did enough to attach the terms "Former RNC Chair" and "Washington Power Broker" were unlikely to be attracted in a year when Ed Cantor went down in flames against a largely unknown opponent. Even recent efforts to play up Mark Warner's non-role in one of the most absurd non-scandals of recent memory, former Democratic State Senator Philip Puckett's efforts to find someone to pay him to do something he was already going to do anyway, had little impact on public opinion.(Philip Puckett resigned giving control of the State Senate to the GOP in exchange for a job for his daughter that public opinion forced him to turn down. Before doing so his son had apparently sounded Warner out on the prospect)
To the extent the race has closed, it has been from 20-25 points to 10-15. There may be a separate discussion as to why that happened, or more accurately, why competitiveness has not happened, but its pretty clear this seat is not in much danger in the near future.