Sunday, August 29, 2010

The Alaska Senate Race

Joe Miller's surprise victory in the Alaska Republican Senate primary has given Democrats at least a marginal opportunity for a pick up this fall, although that will fade if Lisa Murkowski stays in the race for the general as the Libertarian candidate.

Miller leads Scott McAdams 47-39. McAdams is counteracting several of the trends causing Democrats trouble across the country this year. He's running even with independents at 42% and he's benefiting from a more unified party, getting 81% of the Democratic vote while just 73% of Republicans are committed to Miller. In most states that equation would be enough for the lead but in Alaska, where there's an 18 point Republican party identification advantage, it leaves McAdams running behind.

The reason for the closeness of the race is Miller's unpopularity. 52% of voters in the state have an unfavorable opinion of while only 36% see him positively. Democrats (84%) are almost universal in their dislike of him and independents array strongly against him as well by a 54/32 margin. His poll numbers within his own party are positive but somewhat tepid at 57/32. Miller is the latest in a long line of candidates unpopular with the general electorate that Republican primary voters have nominated this year joining Sharron Angle, Rand Paul, and Ken Buck. But this political climate may just be GOP friendly enough that all those folks get through in the fall anyway.

In a three way contest with Murkowski running as the Libertarian candidate Miller leads but with only 38% to 34% for Murkowski and 22% for McAdams. Democrats have been hopeful that Murkowski will run on the assumption she would split the Republican vote but her supporters actually go 47-23 for McAdams in a two way contest with Miller.

Murkowski's support is pretty evenly distributed across the board- she gets 38% of independents, 32% of Republicans, and 28% of Democrats. Miller gets 59% of the Republican vote and McAdams gets 57% of the Democratic vote.

Despite her shocking loss on Tuesday Murkowski remains one of the most popular Senators in the country in PPP's polling with her 50% approval rating placing her in the 90th percentile of the 52 Senators we've looked at this year. But her 52% approval rating with Democrats and her 51% with independents exceeds her 47% standing with Republicans and that speaks to the trouble she had with the primary electorate this week.

Only 60% of the voters who approve of Murkowski's job performance say they'd vote for her in a three way general election contest and the key to her chances would be getting more of the Democrats and Republicans who like her to vote for person over party and abandon their nominees.

If by some chance Murkowski did come back and win the Senate nomination it would be no contest in the general- she leads McAdams 60-28 in a hypothetical contest.

There's plenty more content from our Alaska poll that we'll release over the next few days including an 'exit poll' looking at the factors that drove Miller's surprise victory, how Alaskans feel about Sarah Palin and a potential White House bid, and numbers on the state's races for Governor and House.

Full results here


Anonymous said...

Who says that it's better to nominate a moderate and win with a landslide? In my opinion sweating to a 10-point victory for an ultra-conservative is better.

Anonymous said...

Miller's high "unpopularity" shows just how deranged democrats are. Oh nooooos! He was endorsed by Palin! Now I gots too hate him.

SenatorMark4 said...

I'm a Texas so I can't really speak to an Alaskan politician's standing with the people up there but I do believe that anybody that takes money from a political party and then decides to run under another banner should be roundly defeated as a noodle-spined, liar, and cry baby. If you can't win on your policies, you should accept losing gracefully. You can be given another chance maybe. Taking donors money and then acting like it is yours to run with as you see fit is what is wrong with the national parties. GIVE TO THE POLICITIAN ONLY. The PARTIES inhale your donations for their own benefit.

Anonymous said...

Mama Grizzly I could give you a big bear hug right now. On a serious note did you happen to poll potential 2014 numbers for Begich? I know a lot of people have written him off but his approval is not horrible considering the political environment and the D by his name. Thanks!

Anonymous said...

I guess when you are a liberal biased polling firm and down the list on accuracy, it helps to try to lean the results towards the Democrats... sort of a soft bull crap spreading or in more polite terms, propaganda.

Anonymous said...

I find it interesting that this is perhaps the first state poll in which the sample is more Democratic than in 2008 (R+16 vs R+21.) But that's perhaps because of the strange Alaskan vote that year (polls uniformly overestimated the Democratic vote in Senate/House/Presidential race by ~10-20 points), perhaps because the overall presidential race was already clear by the time Alaskans voted.

I do have a question: Would it be possible for you to break down the Miller-McAdams race among those who do and don't have an opinion of the latter (whether it's positive/negative.)

DBL said...

It's unfair to look at favorables immediately after a primary. Independents will be turned off by the negative advertising the candidates did and supporters of one candidate can be so angry that they dislike the other. That will likely go down after a month.

Anonymous said...

Anon, PPP isn't a liberal pollster. They were the only ones to get Florida's races (on the Republican side) correct. They constantly present non-biased results, that are if anything, a little too Republican.

DBL said...

Can we please leave the idea that pollsters "cook the books" to sites like Swingstate and dailykos? Polls have little effect on the narrative, since very few voters pay attention to them. Take the Alaska vote here. There was no poll that had Miller within spitting distance.

Pollsters live and die by the accuracy of their polls. Candidates want the real story. Tell a campaign they are 5 points up instead of 5 points down and no one will hire them.

Look at Research2000. The "professional left" on dailykos loved their polls. Then they found out they might be fiction and fired them. Even they didn't want fake numbers.

PPP is transparent with their questions and order, leaving it to all of us to judge if they are biased. We saw their party breakdown and could adjust it based on what really seems likely to happen. On dailykos, they're even putting out their raw data.

I'm guessing that PPP has one reason for doing this blog and that's to get more business. They'll get that by being right, as they were with FL-Gov, FL-Sen, Ari-Sen, and Ari-Gov last Tuesday.

Their interpretation of the numbers is pretty much in wacky lefty land, but then that's what the comment section is for.

NRH said...

It makes sense that the 2010 Alaskan electorate wouldn't be as heavily Republican as the 2008 turnout. There's no Presidential ticket on the ballot with an Alaskan governor involved. Palin certainly got the vote out in Alaska among her supporters, the same way John McCain got an unusually high Republican turnout in Arizona and Barack Obama won Illinois by an unusually large margin. All those states went the way they normally would have anyhow, but by larger margins than usual. Without the presidential race this year, that extra boost is gone and the turnout sample returns to something more similar to the Alaskan registration numbers.

Anonymous said...

They weren't the only one to get Florida correct.

In fact, they weren't the closest.

Susquehana nailed it and was closer than PPP.

They were off by around a decent amount in the Republican primaries Colorado, Texas, and Kentucky.

Anonymous said...

PPP is absolutely a liberal pollster, which they don't often advertise but do admit to.

That's aside from the question of whether PPP polls are slanted to one side. Their choice of questions is absolutely chosen to shine a negative light on conservatives - I don't think there's much dispute on that - and one could argue that some of the choices of RV vs LV, how stringent to make the LV screen, when to include minor party candidates, and so forth, is intended to tilt the scale toward liberals for PR purposes.

But their topline results are usually pretty good and tend to track closely with Rasmussen, the industry standard.

The Interesting Times said...

Did you poll the 2012 GOP Presidential primary in Alaska?

I would be shocked if Palin didn't win...but then I wonder if Palin's abrupt resignation didn't hurt her favorables with Alaskans...

The Interesting Times said...

"Anon, PPP isn't a liberal pollster. They were the only ones to get Florida's races (on the Republican side) correct. They constantly present non-biased results, that are if anything, a little too Republican."

I'd say the polling itself is unbiased, but the analysis can have a moderately liberal pro-Democrat lean.

But, since PPP is an open-source pollster that publishes all its crosstabs, you're free to substitute your own analysis if you prefer.

Anonymous said...

It doesn't make sense, a 44% approval for Obama in Alaska! Higher than FL, OH, PA, and MO.

NRH said...

"Their choice of questions is absolutely chosen to shine a negative light on conservatives - I don't think there's much dispute on that"

That's right-wing projectionism right there. Rasmussen has been explicitly caught using slanted questions to favor right-wing outcomes (for an easy example, see here and note that the Republican, Pawlenty, gets the 'friendly' options of 'strongly/somewhat approve' and 'strongly/somewhat disapprove' while all the Democrats get the 'ambivalent counts as negative' options of 'excellent/good/fair/poor'). Now please take a look at the PPP poll this blog entry is written about and point to the 'questions ... chosen to shine a negative light on conservatives'. It might not be disputed among those with the bumper-sticker mentality of conservatives, but when you stick your head out of your comfortable little Beckistan bubble, you have to actually provide some proof.

Or, to put it more succinctly, {citation needed}.

Anonymous said...

Rasmussen is not the gold standard; the fact that you say that shows that you have no idea about pollsters.

Do you really think a pollster that was a paid consultant to George W. Bush's reelection in 2004 in the industry standard!?!?

Do you think a pollster that spews out polls post-primary showing candidates (mainly Republican) up more than usual is the industry standard?

Here's a Rasmussen question: " it’s always better to cut taxes than to increase government spending because taxpayers, not bureaucrats, are the best judges of how to spend their money?" Did you catch that? Not bureaucrats? That's a right wing talking point.

Do you think a pollster that asks if America is the last greatest hope is the industry standard?

Did you know Rasmussen signed on for free to talk on a right wing celebrity cruise sponsored by the national review? The industry standard, really?

Do you think a pollster that polls races early in the cycle to set the "narrative" that Democrats are doomed and then never polls them again is the industry standard? They manage to give us results in Idaho Senate, and Hawaii Senate where there are not pressing elections.

Rasmussen doesn't even release his internals to the general public. We have no idea if his polls are 60% Republican, 35% democrat, and 5% Independent or what.

Rasmussen asks questions calculated to demonstrate support for conservative questions.

Rasmussen says party identification among adults in D +3.7. No! Every other pollster (10 or so) has it between 5.2 and 13.

Nate Silver of the nonpartisan 538: "the sample included in Rasmussen's polling is increasingly out of balance with that observed by almost all other pollsters."

So next time don't say Rasmussen is the industry standard without finding the facts for yourself.

Anonymous said...

For those of us Alaskans who are moderate conservatives- the Miller-McAdams ballot leaves us wondering if it is time to leave the State? The options are pretty grim.

LALaw said...

I was laughing so hard when I read "Rasmussen is the industry standard". Only if by "industry", you mean the conservative media industry. Rasmussen is to polls what Fox News is to news, which is an ironic analogy because Rasmussen's polls seem to be almost the only polls ever cited on Fox News. His health care repeal polls are hilarious - anytime any other organization does a poll on this, they find 60% of the public opposed, but Rasmussen always finds near 60% in support of it. A 20% difference is not just a coincidence.

Anonymous said...

If there is a gold standard, it is SurveyUSA or Quinnipiac.

wt said...

Thank you to PPP for giving us these numbers. The very first polling organization to take on the effort of letting us know where the Alaska Senate race stands.

As for the debate on who's a better pollster, I think we'll find out a lot more in November. My feeling is that any pollster motivated by politics, rather than accuracy, will live a short life, full of recriminations and failure.

If I'm winning, I want to know accurately by how much. If I'm losing, it's important to know that fact. Campaigns establish narratives, set expectations, and adjust their messages on the basis of polling data. Politically skewed data doesn't help anyone.

Besides that, what's the point of skewing data that will be debunked in 2 months anyway? If the GOP wins the House with 60 pickups, what good does it do any pollster (PPP or otherwise) to have predicted that Dems will keep the House? PPP is a polling organiation, not an arm of a Democratic committee.

The same goes for Rasmussen. It's just methods and weighting. There's no polling organization out there (who wants to remain in business) that will sacrifice its hard-earned reputation for a short-term political advantage.

Dustin Ingalls said...

"On a serious note did you happen to poll potential 2014 numbers for Begich?"

No, just his job approval. We had too many questions to ask as it is.

"I guess when you are a liberal biased polling firm and down the list on accuracy"

Third most accurate firm of the most active, according to Nate Silver, and we bested Silver's most accurate, Mason-Dixon, in Florida.

"Pollsters live and die by the accuracy of their polls. Candidates want the real story."

Exactly. Our revenue depends on our reputation for accuracy.

"They were off by around a decent amount in the Republican primaries Colorado, Texas, and Kentucky."

Actually, we only underestimated Paul's margin by 5 points. We did have Mongiardo beating Conway, but that was 17 days before the primary, not 3 like in the GOP race.

"Did you poll the 2012 GOP Presidential primary in Alaska?"

How could we not?

"If there is a gold standard, it is SurveyUSA or Quinnipiac."

Definitely not the latter.

Anonymous said...

SurveyUSA is not the gold standard. Putting out polls with younger voters voting Republicans by 25 points is not the "gold standard"

The Interesting Times said...

Rasmussen doesn't have a right-wing bias. They've been showing Obama with higher approval ratings than Gallup has for at least the past month or so.

The difference people are noting is mostly a result of the likely voter screen.

DBL said...

The NRSC is out with an Alaska Senate poll that has Miller leading by 16%. It's an incredibly convenient time for them to do it and the results say just what they want to hear. I take it with a grain of salt, but I like to give the benefit of the doubt.

PPP's poll was done Aug 27-28, while the NRSC poll was done Aug 28-29. One day shouldn't mean much, but we have seen some odd Rasmussen bounces immediately after a primary that have gone away the next week.

On the other hand, the NRSC polled only 400 people, while PPP polled 1,306. Thus the NRSC has an MOE of 4.9%, compared to 2.7% for PPP. The PPP poll may be off because it's so close to the election, but a poll of 400 people is dubious.

Anonymous said...

Did you see the list of complaints I lodged against Rasmussen up a couple comments, The Interesting Times?

DarfurWarrior said...

"Polls have little effect on the narrative, since very few voters pay attention to them."

This is not intended to be derogatory, but I'm a little curious how you would know this?

This form of data can only be obtained from a Polling survey, and you don't pay attention to those, Right?

I'm not sure about this Party thing, because I am an American whose allegiance is to my country,not to an elitist party or separatist ideology.

I prefer not to teach my children the difference between Left & Right, but the difference between Right & Wrong.

Some right or wrong questions I would love to see the poll results for:

Why is there a list of Nationality choices on the US Census, and American is not among the options?

Why would our government not give us the option to claim our nationality to be American?

Why do they say we are one people, and then ask us to define ourselves by divisive terms?

Like I said, nothing personal, just curious.

Anonymous said...

"Can we please leave the idea that pollsters "cook the books" to sites like Swingstate and dailykos?"

Don’t compare Kos and the Swing State Project in terms of horse race politics. SSP is the best political blog out there. It is very well respected. Kos is a good blog but not so much in terms of horse race stuff. SSP is the best horse race politics blog, hands down. Try going to it. It is VERY well respected among Dems and Repubs. Perhaps you should take a few days off from redstate. If a SSPer tried to promote cook the books theorys they would likely be banned. If a SSPer tried to promote cook the books theories they would likely be banned.

Jonny V said...

Hopefully some of these far right wing candidates that the Tea Party people have nominated for the Republicans will lose and keep the D losses to a min.

This Miller guy in Alasks, Paul in Arkansas, Rubio in Florida, and Angle in Arkansas all need to go down for the future of our country.

Anonymous said...

Rasmussen doesn't have a right-wing bias. They've been showing Obama with higher approval ratings than Gallup has for at least the past month or so.

False. And even if it was true, that's really what you're going to hang your hat on? Not the fact that Rassmussen works for Fox News, or the fact that his polls consistently show a house effect biased towards Republicans? You know jack about polling.

Web Statistics