North Carolina Senate Moves From “Toss-Up” to “Leans Dem”

Published by:

A big day for Democrats as, for now, I’m taking North Carolina off the Toss-Up list. While I moved this race to “Leans Dem” for a brief period in the middle of the summer, Tillis had a decently strong month of August. Yet, that’s quickly faded again as we’re now halfway into September. With a month and a half until election day, all the data points would indicate that Kay Hagan has likely pulled into a very real, though potentially small, lead.

For the last several model updates, I’ve consistently had Hagan up in the 3-4% range. The problem for her is that she still remains below 48%. The advantage for her is that she’s been slowly climbing toward that mark as Democrats and moderates have been coming home to her. She ended strong against Elizabeth Dole to win six years ago, and she may be doing so again here. Hagan is a moderate Democrat from Greensboro, a bellwether part of the state that’s not known for being the liberal hubs of Charlotte or the Research Triangle of Raleigh-Durham-Chapel Hill. Instead, Hagan, like Mark Pryor, plays well in middle of the road cities like Greensboro, Winston-Salem and Wilmington. The difference between Pryor and Hagan is that NC is simply much, much bluer than Arkansas. The state is now anchored by the 1-40 corridor which is filled with young professionals and educated, white collar transplants from the North and West. (Raleigh and Durham were just ranked as the #2 and #3 most educated cities in the country, only eclipsed by Ann Arbor, if that gives you a perspective on the make-up of these cities. Throw in Chapel Hill/Orange County in, which is the most liberal county in the entire South, and Democrats have a decent base in the state.) To win, Democrats need big margins in these parts of the state and then must carry Winston-Salem, Greensboro, and Wilmington by at least a few points. With her base in the Winston-Salem/Greensboro area, Hagan gets an advantage.

Tillis is simply stuck in the low to mid 40s. He’s yet to show that he can convert those remaining voters in the state. He’s tied to a deeply unpopular state legislature that has dragged him down, much like the unpopular Democrat-controlled legislature in Colorado has hurt Udall and Hickenlooper.

Overall, I have Tillis tracking in the 40-44% range with Hagan in the 45-48% range. As such, I’m moving this race to “Leans Dem” today.

North Carolina Senate Rating: Moves from “Toss-Up” to “Leans Dem”

Model: Hagan(D) 47.5% – Tillis(R) 43.7%

Alaska Senate Moves From “Leans Dem” to “Toss-Up”

Published by:

The Alaska Senate race is anyone’s guess. I really mean that- anyone who tells you they know who is ahead is being dishonest.

First of all, there’s very little data from the race. What does exist is additionally varied and neither candidate is nearing 50%. (That’s not uncommon in Alaska where 3rd Party candidates can get decent numbers, but it does make predicting the race very difficult when the numbers are so close.) Additionally, I’d be very leery, no matter what the data points show, of predicting a Begich(D) victory with numbers this close. In August, the race moved to “Leans Dem” because Begich(D) was up 4-6% in the model. Even then, I wrote about my leeriness to move the race that way, but I was relying on the data. Today, Begich’s lead is down to a tiny and almost meaningless 0.6% in the race. For now, that means I’m still at Dems 51 in the ‘if you forced me predictions,’ but I may qualitatively bet against Begich if this were the day before the election. Why?

Alaska polling has traditionally been quite terrible. I sometimes wonder if Alaska is the place social science goes to die. Begich was supposed to beat Ted Stevens by a pretty decent margin but ended up squeaking it out by a hair. In the Lisa Murkowski write-in drama, the Democratic candidate way underperformed come election day. Remote voters? A vast land of a state with drastically different demographics? Strong third party and independent numbers? A small population where small swings can make big differences? Thousands of voters who live out-of-state and keep their residency for the tax benefits, resulting in a wave of absentee ballotts? (Many of these voters are conservative military members who may have been stationed in Alaska at some point in their career and maintain residency years later). All of these small differences make for more unpredictable results than in most states. With that, a 3% candidate lead in the model in Alaska does not instill in me the same type of statistical confidence that I would have in any other state.

For as little polling as there is in the state, the data is actually not ‘that’ noisy. PPP has Begich up 5-6% in their two most recent surveys but has him stuck at 43%. Back in late June, Basswood, a R-leaning pollster, had Sullivan up 5%. Meanwhile, the Senate Majority PAC’s poll by Harstad recently showed Begich up 5% as well. That is actually not a great number for him considering the internal nature of the poll. Thus, today’s model showing the race stuck around 46% for both candidates illustrates the unknown nature of the state’s victor.

As further indication of the closeness of the race, Sullivan is tracking in the 41-46% range overall. Begich is in the 42-47% range. Thus, this race could easily end up being a 4-6% victory either way. So, hold onto your hats. At the moment, Senate control could VERY well come down to this state. At the earliest, if a candidate is a clear victor, we’ll know in the wee hours of the morning on election night, East Coast Time. If this race is close, we may not know for weeks who controls the Senate. Keep in mind that early returns from East Coast states will be coming in while Alaska is still voting. How might that affect last minute voting on election day? An interesting election year it is…

Alaska Senate Model: Moves from “Leans Dem” to “Toss-Up”

Alaska Model: Begich(D) 46.5% – Sullivan(R) 45.9%

Michigan Senate Moves to “Safe Dem”

Published by:

This race hasn’t been all that competitive all year, but Land has been ‘close enough’ for the race to stay potentially interesting. As election day draws closer and closer, the race has remained quite stable. If anything, Peters(D) has seemed to be shoring up the Dem leaning bent of the state as his model numbers have slowly ticked up from the 5-6% range to the 7-8% range over the last several weeks.

Today, I have Peters(D) up 49.8% to Land(R) 42.1%. While Peters is still shy of that magic 50% mark and a decent amount of voters are hanging out there in this race, the money for Land is going to dry up in October if there are not some dramatically different numbers in the next two weeks. Land’s numbers are simply stuck in the low-40s with no signs of improving. At this point, it appears she’d have to win every undecided voter or convert Peters voters. With data like this, it normally would take a collapse of the leading candidate in the remaining weeks for that to happen. Peters may not be the most enthralling candidate in the world, but he’s certainly shown he can run a competent campaign.

Land’s model lines range from 38-43% while Peters’s range from 46-50%. The trend is clear. While Land could perhaps close this race back to the mid-single digits, there are simply no signs she is on track to win this race. In fact, it would be far more plausible at the moment that she’d lose by double digits than actually win. As such, I’m moving this race to “Safe Dem” today.

Michigan Senate Rating: Moves from “Likely Dem” to “Safe Dem”

Model: Peters(D) 49.8% – Land(R) 42.1%. 

Kentucky Senate Moves to “Leans GOP”

Published by:

Mitch McConnell may be shoring up his very red state as the last days of summer tick down. This race has been a Toss-Up all summer long, but I’m shifting it to “Leans GOP” today as GOP and conservative Democrat voters in the state of Kentucky are “coming home” to the candidate that is McConnell. The problem for Grimes is not her but Barack Obama. This was always going to be an uphill climb for her to win in a state where the President has a low 30s approval rating. Voters in the state simply don’t appear to want another vote for the Obama administration in the Senate, no matter how much they don’t like Mitch McConnell. The same type of ad campaign is hitting Mark Begich, Mark Pryor, Mary Landrieu and Kay Hagan. As I reside in NC, I can tell you I don’t think I’ve gone a day without seeing an ad connecting Kay Hagan to Barack Obama since last winter. While Obama is more popular in NC than in KY, his numbers are pretty sour here as well. Keep in mind, KY, WV, and TN are some of the only places that Obama did worse than John Kerry, so there’s been a challenge there for Grimes that is quite daunting. That reality is finally being shown in the numbers.

Is Grimes out of it at this point? Of course, anything is possible. The best news for her is that McConnell is still below 50%, albeit narrowly so. No pollster has had Grimes better than 4% behind since July, and the momentum is clearly against her. While McConnell is not popular, all of these ads have taken a toll on Grimes whose own popularity has been on the decline, as is often the case for any challenger faced with such a negative barrage. And so, the GOP may be finally shoring up this seat.

Here’s an example of the McConnell Ads:

KY Rating: Moves from “Toss-Up” to “Leans GOP”

Model: McConnell(R) 49.4% – Grimes (D) 44.4%

Kansas Senate Gets Shaken Up; Gets Moved to Toss-Up

Published by:

Well, this is interesting. If you haven’t heard, the Democratic nominee in the Kansas Senate race has just dropped out of the race. So why is that potentially bad news for…Republicans? Because there’s an Independent candidate named Greg Orman who has plenty of money, a moderate agenda, and even more endorsements. This race just got interesting, folks.

Kansas politics are a mess for the GOP right now. The state is heavily red, but Sam Brownback is highly unpopular. He may still be able to pull out a victory this fall if disgruntled Republicans come home to him and just can’t force themselves to pick a Democrat, no matter how moderate. After all, few governors have been more unpopular than Pat Quinn in Illinois, yet he has been able to survive in that state because he has a (D) next to his name. Thus, whether it’s Senate or Governor, non-GOP candidates still have a tough climb. Yet, moderate Paul Davis(D) has so far proven to be a very strong candidate with a slight edge in the Governor’s race. All of that is the backdrop for why Kansans are frustrated, which gets us to the Senate race…

While Sam Brownback is unpopular because of his governing agenda, as well as relationship with GOP folks around Topeka, Pat Roberts has a different problem. He is not a Kansan anymore, at least from the point of view of many Kansans. A few months back, the NYT reported what had long been known: that Pat Roberts is a creature of Washington, to the point of no longer even having a residence in the state. The paper found that Roberts’s listed address was that of some campaign contributors. The result? Not good for Roberts. It has epitomized him as that aforementioned creature of Washington, something that is loathed something extra special this year in particular. The nation is frankly tired of both parties. Enter Greg Orman.

While many polls showed Democrat Chad Taylor just a few points back, I had no indication that he was actually in a position to win the race, especially with a viable Indpt. running. Now that Taylor is out, it comes down to an unpopular GOP Senator and a well financed Indpt. moderate alternative. Of extra note? Kansas is not the most expensive state to run in, making a non-major party victory all the more possible. Throw in a local and national media that would LOVE the story line of a Washington insider getting booted by an Indpt. candidate, and you’ve got the makings of trouble for Roberts. A PPP poll about two weeks ago showed that, in a head to head, Roberts would be down 10 points to Orman. I am sure Taylor and his influential backers have been looking at these numbers for the last few weeks themselves. His campaign surely realized that his own chances of winning were close to none, but some wealthy folks, some party insiders, and others suggested he make way for the Independent in the race.

With all of this, we’ll have to see more data from the state, now that it’s one on one. But, for now, I’m moving this race from “Likely GOP” to a “Toss-Up.” We now have our most interesting race of 2014.

Arkansas Senate Moves Back to Toss-Up

Published by:

Mark Pryor(D) is keeping things alive in the Arkansas Senate race. I tweeted a few days ago that, to hold the Senate, Democrats would need to hold Alaska, Colorado, and Iowa, then pick up North Carolina, or somehow scoop up Louisiana or maybe even Arkansas. Well, Iowa seems to be getting trickier by the day, but Arkansas has been clawing back over the last few weeks to become a stronger hold potential for Democrats.

At the start of the month, Cotton(R) seemed to be narrowly on his way to picking up the seat as the race had slowly moved his way throughout the summer. All signs pointed toward a 4-5% lead in his favor. Then, August happened. Mark Pryor(D) has been spending heavily this past month, and he’s climbed his way back up into a Toss-Up position. Now, the overall dynamics of the race still narrowly favor Cotton, and most of the recent movement has so far shown a Pryor climb as opposed to a Cotton fall. The importance of that is that Cotton is still sitting at 49.3% in the model, awfully close to that 50% mark. Additionally, that’s the exact same data point for him as the last model. What’s changed is that Pryor has climbed from data points in the mid-40s back up to 47.9% today. The indications are that some Arkansas residents who may be quite sour on the Democratic party are still okay with Mark Pryor, a Democrat with strong Arkansas ties and a sturdy local brand. The question in the remaining two months of the race is whether there are enough of them out there for him to hold onto, at what the moment, would be a squeaker of a victory, if he were somehow to pull it off by scooping up a very narrow pool of remaining voters. Still, Pryor’s been making gains for the last several weeks, and as I always say, the trends are what often matter. Cotton’s 4-5% margins shifted to a 2.7% margin about ten days ago. Today, they’re down to a 1.3% margin. So far then, Pryor’s still on the climb and has yet to plateau on his current upward trend in the model. Hence, this race officially moves back to being a Toss-Up. A recent data point from Rasmussen today showed Pryor up, 44-43%. Rasmussen had Cotton up 4% a few months ago, showing consistent movement for Pryor as of late.

The question now becomes: will the trend continue? Will Pryor actually be able to recapture a lead in the model, something he hasn’t been able to do since spring?

Arkansas Senate Rating: Moves from “Leans GOP” to “Toss-Up”

Arkansas Senate Model: Cotton(R) 49.3% – Pryor(D) 47.9% 

Kansas Senate Moves From “Safe GOP” to “Likely GOP”

Published by:

I’m making a small downgrade in the Kansas Senate race today as I’m moving the race from “Safe” GOP to “Likely” GOP. The numbers for Roberts continue to look decent overall as there are certainly no indications that he’s behind, but his margins are slipping and he’s not over 50% in the model. Thus, the downgrade. For the first time, I have him under a 10% lead in the model. Today, he comes in at +8.6% and he’s peaking at 47.4%. He may well win this race without hitting 50% in a reliably red state. Roberts has been pummeled all year for being a Senator that’s been MIA within his own state. The real wildcard here is Greg Orman, a self-financed independent running in the state. Kansas is not a large state population wise, so it does not take as much money for an independent to make some noise than in other locals. If Orman continues to run and ends up in a debate, we’ll see if his numbers continue to hold. Right now, he’s getting about 10-12% in the model, a very real number for a third party candidate. It’ll likely take more than that if Taylor(D) is to catch up with Roberts as Taylor is stuck below 40% himself. The problem for Taylor is that Orman is taking all of those moderates and independents with him, voters that may have been willing to pick Taylor over Roberts if that was the only viable option. Third party candidates are thus major wild cards. Normally, they dissipate quickly in October as voters decide to vote for the “lesser of two evils” instead. But, that’s not always the case, especially if the third party candidate proves to be viable ’til the end, as shown by Cutler(I) surging in the Maine Gov race at the end of 2010. Orman’s got a long way to go for that to happen, but this race is, potentially, getting interesting…

Kansas Senate Model: Roberts(R) 47.4% – Taylor(D) 38.8% – Orman(I) 11.0%

State Rating: Moves from “Safe” GOP to “Likely” GOP

Arizona Gov Model Starts as a Toss-Up Post-Primary

Published by:

The primary season saw its last big hurrah last night as Charlie Crist secured his status as the Democratic nominee in the FL Gov race and Doug Ducey made it through a tough AZ GOP primary for Governor. The businessman turned State Treasurer is of the Cold Stone Creamery fame. He built the company up into a national brand before selling it and surely taking some major profits along the way. Now, he faces Fred DuVal, a moderate policy and business person.

The dynamics in Arizona still favor the GOP at the state level. The party has a registration advantage and holds most of the power in the state. That said, that’s not always a good thing for a party as voters grow tired of one party control. This year, there are plenty of examples: Dems Hickenlooper and Udall struggling with the unpopularity of Dem control in CO, Tillis in NC and Brownback in KS with the unpopularity of GOP control.

The polling data out of Arizona remains pretty sparse, so we’ll have to see where this race goes. Both candidates are likely to be well funded in the final few weeks, and DuVal’s moderate background may help propel him with Independents. Additionally, if Democrats are worked up enough in the state to switch gears post-Jan Brewer, then DuVal may hold close. At the moment, the race is still a Toss-Up, though there’s a lot of play here as, even with implementing an election day forecast model, neither candidate is above 43%. That makes this race worthy watching.

Arizona Governor Model: Ducey(R) 42.5% – DuVal(D) 40.4%

Rating: Toss-Up

Iowa Senate Model As Close As It’s Been

Published by:

Pretty much every poll in Iowa will tell you the same thing: The Iowa Senate race is a virtual tie. (Check out PPP’s new numbers here.) Unlike in other states, the polling data in Iowa is not only aplenty, but it’s also quite consistent and much of it done by reputable and long standing pollsters. The reality in the state seems to be this: die hard Democrats are in Braley’s camp, die hard Republicans are in Ernst’s camp, and no one else wants to vote. Thus, while the state has a slightly blue tinge, those moderate voters have yet to decide if they just want to stay home or vote for Braley, who they don’t like. While Ernst has proven to be a good campaigner with her base, her rightward-politics are a turnoff for most voters outside of her base. Meanwhile, Braley has proven himself to be a weak campaigner statewide, getting plenty of flack for his own comments throughout the race.

Two of the four model lines have the race at an exact tie while Braley is up by 1% or less in the other two. The good news for him is that he’s never officially been behind in the model, but that’s really just a silver lining when the numbers have been consistently so close. Do I still have a hunch that a few stragglers are going to break his way with the 6% or so of voters who I still can’t assign in the model? Yes, though, coupled with Ernst’s arguably better job of firing up her base, Braley is certainly in trouble if those two factors negate or those moderate voters just stay home. For those expecting him to break out at some point, it has yet to happen. He’s got about two months to make it happen but he’s so far not been too convincing. As such, the race is as close as it’s ever been in today’s model:

Iowa Senate Model: Braley(D) 47.0% – Ernst(R) 46.7%

Rating: Toss-Up

Alaska Senate Moves to “Leans Dem”

Published by:

The Alaska Senate race moves to “Leans Dem” from “Toss-Up” today. The central measures show a different dynamic to that race now that more data has been coming in. That said, there are some major caveats here that should prevent Democrats from getting too confident.

Today, Mark Begich(D) hits 50% for the first time, hitting 50.2% overall. Dan Sullivan(R) is still stuck down at 44.3%. That 5.9% margin, coupled with Begich leading in every model line by a few points, makes me shift this race today. Make no bones about it, overall, Begich is doing well in many polls. The PPP polls consistently show him up by the mid-single digits while the NYT ‘model’ has Begich up by double digits. The Rasmussen poll out today shows Sullivan up by a narrow 2%. In some ways, that’s actually not a bad number for Begich as most candidates, in this case Sullivan, get a decent bounce post-primary. So, there is a chance that number may reflect Sullivan’s high end scenario.

Now, let’s get to those caveats. First, I could use A LOT more data out of Alaska. There has not been a lot of polling there, especially from non-partisan sources. Second, those PPP numbers still leave the race wide open because, even though Begich has a consistently nice lead, they have a lot of undecideds left. Third, and perhaps the biggest X-Factor of all: this is Alaska we’re talking about. That means a few things. First, Sullivan may end pulling most of those undecideds. Thus, it’s important that Begich is at 50% as, even if that ends up being the case, he still looks likely to pull out a narrow victory. But the bigger factor here is the history of bad polling data from Alaska. The state is frankly hard to poll, is my take from the historical data I’ve seen. There’s a huge diversity of residents, some who live off the grid. Additionally, there are thousands of Alaska residents who vote absentee and may never get picked up by pollsters. In a small state, those thousands of voters matter a lot. Many such residents are military members who were once stationed in Alaska and have since moved yet keep residency. (After all, Alaskans still get major income tax perks, so why switch residency?) Many of these voting ‘non-residents’ are very conservative leaning. Thus, no matter what the polls in Alaska say, this is the one state where, even as a data-person, the social science has some difficulties. After all, if such research is often accurate at a 95% level, than that one out of 20 times it misses may be Alaska.

That said, the other reality is, it’s always smart to rely on the data. Many didn’t believe the numbers in the 2012 North Dakota Senate race and Heitkamp(D) bested everyone’s predictions. Thus, for today, Begich(D) seems to be in a pretty decent position. Democrats should probably still hold their breath with data like this, but there are at least okay signs that he may indeed pull out this race.

Alaska Senate Model: Begich(D) 50.2% – Sullivan(R) 44.3%

State Rating: Moves from “Toss-Up” to “Leans Dem”