Just a week ago, the big story from Gallup was that the Republicans had hit an all-time high lead in their poll. I covered it despite questioning the poll in the past. Everyone covered it.

But now, suddenly, in the Gallup poll the race is even? How can that be, and what does it mean?

Of course, last week saw more than the Republicans taking a 51-41 lead in the Gallup generic ballot. The party also took the largest lead of the year in voter enthusiasm: 50-25, with independents at 28. In depth: Republicans favored their own 96-3, Democrats favored their own 88-9, and Independents favored Republicans 48-31.

What’s changed? Democrats have swung nine points to favor their own party 93-5. Independents have swing a point to favor Republicans 49-33. Republicans swing 5 points to favor their own 93-5. Somehow, all three of these add up to a ten point total swing. Further, voter enthusiasm remained unchanged among all three groups, but decreased overall by one point.

There’s only one conclusion to draw here: this week’s Gallup captured a lot more Democrats than last week’s did. That’s the only way these numbers can add up. So unless we actually think that the electorate is actually beginning to identify more with the Democrats than with the Republicans, this is what we call an outlier: one of those one-in-twenty of worse events where the poll just isn’t even close to the actual results, and isn’t reflective of reality.

How they managed to get an outlier rolling up a week’s worth of daily tracking though is beyond me.

Possibly Related Posts

• August 30, 2010 -- Gallup: Generic ballot shows record Republican lead
• August 24, 2010 -- Swingometer Update
• August 3, 2010 -- Gallup retreats and I claim victory

Yes, it’s the tiniest of all possible leads, but Republican Carly Fiorina has taken another polling lead over Democrat Barbara Boxer. Up until now we’ve had a discrepancy in the polling, in which SurveyUSA had Fiorina ahead but everyone else had Boxer ahead. I wondered how long that gap would last.

Apparently the answer was “one week.”

Of course nobody’s going to be able to claim that 48-47 (MoE 4) is a result marking a runaway freight train with unstoppable momentum creating a sure win. I show a 46% chance that Boxer still leads, according to this poll. It could be random noise, but it is something Fiorina had not yet accomplished: a lead in Rasmussen Reports to go with her leads in SurveyUSA.

If Boxer wasn’t worried before, she won’t be worried after this poll. She was well ahead in Public Policy Polling and very close in the other two. But at the same time, if Fiorina was concerned that SurveyUSA had a partisan slant and nobody else was capable of showing her ahead, that fear is now proven unfounded.

This is the most competitive I have ever seen a California Senate race, given that I started watching politics at all back in 1992 when Dianne Feinstein and Barbara Boxer first won office to begin with. Boxer’s not in the spot that Blanche Lincoln is in, but she’s no longer safe like Barbara Mikulski, either.

Possibly Related Posts

• June 9, 2010 -- Grading the Pollsters: June 8 edition
• September 3, 2010 -- Carly Fiorina maintains SurveyUSA lead
• July 14, 2010 -- Is Rasmussen biased toward the Republicans? Not in California.

I’m definitely a latecomer to following the Delaware Senate primary between Republicans Mike Castle and Christine O’Donnell, but now that I’m aware of it, it’s striking to me just how differently the two candidates perform in the new Rasmussen poll featuring each candidate against Democrat Chris Coons.

This is a case where the top lines truly speak for themselves. Castle leads Coons 48-37 (MoE 4.5) for an 88% chance of being ahead after this poll. However O’Donnell trails 36-47 for an 11% chance of being ahead. The results basically flip from Republican to Democrat depending on which candidate the Republicans nominate, with Coons going from down 11 to up 11 if O’Donnell manages to knock off Castle.

As I said before, O’Donnell’s challenge in doing that is to find a way to get Delaware Republicans to vote against the very popular Castle, and Castle still shows as popular in this poll. His favorability ratings are lopsided in his favor at 67% favorable/30% unfavorable/3% not sure. I question how Coons at 49/34/15 can come back to win in the general against Castle, but I also question how O’Donnell at 39/44/17 can beat either Coons or Castle.

In particular, Liberty.com’s attacks on Mike Castle, questioning his loyalty both to his party and to his wife, seem not to have moved the needle in Christine O’Donnell’s favor.

If Castle weren’t quite as popular among Delaware Republicans, maybe O’Donnell would have a shot. But I just don’t see it happening for her. Delaware is a small state and Castle has done so much retail campaigning over the years during his runs for the House that it’s all added up in his favor.

Possibly Related Posts

• September 1, 2010 -- Extrapolating about the Delaware Senate primary
• April 30, 2010 -- Obama and Biden to be succeeded by Republicans?
• September 8, 2010 -- Carly Fiorina takes another lead

My plan had been to push out projections every Monday until election day but Labor Day brought out the laziness in me, so we do this week’s on Tuesday. Last time, my third projection showed the second straight gradual gain for Republicans. Let’s see if the trend continues.

Situation

Senate right now: 59 D-41 R.

Seats up right now: 19 D-18 R.

Seats therefore not up: 40 D-23 R.

Minimum result needed for majority: -9 for Democrats, +10 for Republicans.

Method

I’m going to work off the polls when they’re available, working from Real Clear Politics and make guesses otherwise. That’s usually easy because the seats that aren’t polled are typically safe. Changes from last time will be in boldface.

Seats

AL: Richard Shelby: Primaries done. Lead has grown to +32 in Rasmussen. 99.9% R victory.

AK: Lisa Murkowski (primaried): Primaries done. Joe Miller leads the early polling but not quite overwhelmingly. 80% R victory (from 99%).

AZ: John McCain: Primaries done. McCain maintaining 20 point leads. 99% R victory

AR: Blanche Lincoln: Primaries done. John Boozman up 20+ in polling. 99% R victory.

CA: Barbara Boxer: Primaries done. Polling mixed but favoring Boxer. 33% R victory (from 45%).

CO: Michael Bennet: Primaries done. Only PPP has Bennet ahead. 75% R victory.

CT: Chris Dodd (retiring): Primaries done. McMahon making steady gains on Blumenthal. 15% R victory.

DE: Joe Biden (resigned): Primaries in progress. Mike Castle’s lead down to the teens. 95% R victory.

FL: Mel Martinez (resigned): Primaries done. The wrong Democrat won for Charlie Crist. 85% R victory.

GA: Johnny Isakson: Primaries done. Isakson lead in the 20s save for one Insider Advantage poll. If it’s confirmed by others this rating will move further. 99% R victory.

HI: Daniel Inouye: Primaries in progress. Inouye nearing 70 in poll. 0.1% R victory.

ID: Mike Crapo: Primaries done. Crapo lead hovering around 40. 99.9% R victory.

IL: Barack Obama (resigned): Primaries done. Last two polls both tied. 50% R victory (from 55).

IN: Evan Bayh (retiring): Primaries done. Dan Coats lead growing. 99% R victory.

IA: Chuck Grassley: Primaries done. Grassley rebounds from one bad poll. 99% R victory.

KS: Sam Brownback (retiring): Primaries done. Republican Jerry Moran leading all polls. 99.9% R victory.

KY: Jim Bunning (retiring). Primaries done. Randal Paul’s lead seems to be growing. 85% R victory (from 75%).

LA: David Vitter: Primaries done. Vitter gets a bounce in Rasmussen. 95% R victory (from 90%).

MD: Barbara Mikulski: Primaries in progress. Mikulski’s primary opponent’s internal even has her ahead. 0.1% R victory.

MO: Kit Bond (retiring): Primaries done. Blunt marching on from one bad poll I didn’t trust anyway. 85% R victory.

NV: Harry Reid: Primaries done. Reid takes another tiny lead in a back and forth race. 45% R victory (from 50%).

NC: Richard Burr: Primaries done. Burr keeping lead but yet unable to pull back ahead by double figures. 70% R victory.

ND: Byron Dorgan (retiring): Primaries are done. John Hoeven leads by so much he breaks the graphics in my analysis tool. 99.9% R victory.

NH: Judd Gregg (retiring): Primaries in progress. Kelly Ayotte bouncing back. 85% R victory.

NY: Chuck Schumer: Primaries in progress. Schumer’s lowest lead in any poll was 19. 0.1% R victory.

NY: Kirsten Gillibrand: Primaries in progress. Gillibrand leads every matchup. 1% R victory.

OH: George Voinovich (retiring): Primaries done. Portman extends his lead. 85% R victory (from 70%).

OK: Tom Coburn: Primaries done. Coburn lead approaching 40. 99.9% R victory.

OR: Ron Wyden: Primaries done. Recent polling very good to Wyden. My thoughts of this being a sleeper GOP victory seem to have been premature. 1% R victory.

PA: Arlen Specter (primaried): Primaries done (especially for Specter). Pat Toomey finally hits double figures in one poll. 85% R victory (from 80%).

SC: Jim DeMint: Primaries done. No polling I can find. Alvin Greene’s indictment seems to have diminished interest in this race. 99.9% R victory.

SD: John Thune: Primaries done. Unopposed. 100% R victory.

UT: Bob Bennett (primaried): Primaries done. Lee lead down to 25 in the sole poll. 99.9% R victory.

VT: Patrick Leahy: Primaries in progress. Leahy up 35. 0.1% R victory.

WA: Patty Murray: Primary done. Rossi takes another poll. 60% R victory (from 45%).

WV: Robert Byrd (died): Special election in progress. Manchin lead cut from 16 to 6 in sole poll. 25% R victory (from 1%).

WI: Russ Feingold: Primaries in progress. Still no new polls other than Rasmussen’s with Johnson’s small lead. 45% R victory.

The simulation

100,000 trials using the probabilities above.

Results

As with last time, a 7 seat Republican gain is most likely according to these estimates. Though there is a tiny Republican gain from last week as 8, not 6, is the second most likely outcome now. Those three outcomes of R+6 to R+8 account for 62% (61,798/100,000) of the curve. 6% (6,386/100,000) of the trials gave Republicans a Senate majority, up from last week’s 4%. Only 3 of 100,000 trials gave Democrats a gain.

So we have another slight gain from last time for Republicans, making it three in a row that the opposition has gained, though this time the gain was much less than a whole seat. A majority change is still unlikely according to the polls, but 6% starts to become worth talking about.

Possibly Related Posts

• September 8, 2010 -- Carly Fiorina takes another lead
• September 7, 2010 -- Rasmussen: Castle greatly outperforms O’Donnell
• September 5, 2010 -- SurveyUSA: Paul well ahead in Kentucky

As I hinted yesterday, it is now time to update my House projection. My last comprehensive review gave Republicans 52 seats over 2008, but let’s see how far the Swingometer needles move this time, as even I have been surprised by how far some of these new Generic Ballots have shifted toward Republicans.

Old reliable Real Clear Politics again has seven generics for us to look at. Two of them polled likely voters and the remaining five polled registered voters. When I average the Swingometer projections from all seven polls, it will be a weighted average giving the likely voter polls double weight.

To get a “swing” to put into the Swingometer for each poll, what I do is subtract out all of the voters in each poll that don’t choose Republicans or Democrats in order to get a bare two party vote. I then compare that with the actual, total two party vote in 2008’s House elections. This compares apples with apples.

First thoughts: Gee, that Newsweek looks pretty far off, doesn’t it? And I find it interesting that the Gallup poll done for USA Today has a sample size of 928 (a typical size) while Gallup’s own has a sample size of 1540 (second largest, behind only Rasmussen’s 3500). Who paid for a poll matters.

But moving on, the weighted average of these swings is R+22. Which means even with Newsweek included, Swingometer says Republicans gain 60 seats over 2008 and take a 238-197 House majority. If all of this is accurate then we’re in territory never seen since the Truman administration in terms of a single year swing. Back in the day, Presidential coattails would move the House far his way, but the mid-term would react to that, causing the House to be remarkably swingy. That’s not entirely the case anymore, as two of the last three midterms have even favored the party of the President, so a 60 seat win just isn’t done anymore.

But the generics suggest the Republicans just might.

Possibly Related Posts

• August 24, 2010 -- Swingometer Update
• September 6, 2010 -- Swingometer does not rest on Labor Day
• August 30, 2010 -- Gallup: Generic ballot shows record Republican lead

I didn’t know if Real Clear Politics would give me any new polls to discuss on Labor Day, but we actually have two interesting ones today: two new generic ballots to feed into the Swingometer, holding us over to my next big Generic Ballot roundup.

The first new generic we have is the ever-busy Rasmussen Reports at 48-36 for a whopping R+12 in the raw figures. If we boil it down to the two party split, it’s 57-43 for R+14.

The other generic is from Opinion Research for CNN. This one will get some attention as in the raw figures it has the GOP over 50, ahead 52-45. Again paring that down to the pure two party vote it becomes 54-46 for R+8.

Rasmussen swings us from 2008’s D+12 to R+12 for a 24 point swing. Swingometer says a 24 point swing from 2008 means a massive 64 seat Republican gain from 2008, resulting in a 242 R-193 D House majority.

CNN/OR’s result isn’t quite as drastic. Going from D+12 to R+8 is “only” a 20 point swing, and “only” a 57 seat gain over 2008 in the form of a 235 R-200 D House.

News like this can only continue to depress Democrats, and the CNN poll isn’t even using a likely voter model, instead just polling all registered voters.

Possibly Related Posts

• August 24, 2010 -- Swingometer Update
• September 7, 2010 -- House Projection for September 7
• August 30, 2010 -- Gallup: Generic ballot shows record Republican lead

SurveyUSA is no fly-by-night operation in polling. They’ve been around a while, they have a reputation, and a great many newspapers seem use them to poll local House races.

So we can’t dismiss their continuing series of polls which look very good for Republicans, including this new Kentucky Senate poll for the Louisville Courier-Journal with Republican Randal Paul seeming to overwhelm Democrat Jack Conway.

The top line of this poll is just brutal: Paul 55, Conway 40 (MoE 4.2). With this poll we’re at the range of a 96/4 split of lead probabilities. Paul wins 82% of Republicans, 32% of Democrats, and 56% of Independents, and I think it’s that number for Democrats (double what I would expect) that is the difference between Paul a close race and this large margin.

I’d be curious to see that split for Democrats without the Likely Voter filter put in. I suspect among all registered Democrats it’s closer to 15% for Paul and not this massive 32%. Barack Obama’s “professional left” just isn’t showing up this year, according to SurveyUSA’s projection..

It also doesn’t help Conway that the TEA party activists are so important this year. Those people who favor it (40%) back Paul 87-10. Neutrals (26%) also favor Paul 60-35. Those people with no opinion of the movement (9%) also support Paul 45-37. Conway has to find a way to win when he only leads among the 25% who oppose the TEA party, because those he wins 90-8.

This is a TEA party year, and Randal Paul is (a particular kind of) TEA party candidate, since some supporters of his father claim that they were the TEA party before anyone paid attention to the TEA party. So while some may question SurveyUSA’s likely voter screening, I think this poll reading is definitely plausible.

Possibly Related Posts

• July 6, 2010 -- What is going on in Kentucky?
• May 31, 2010 -- Conway closing in Kentucky
• August 20, 2010 -- A Kentucky poll I just don’t believe

New Hampshire apparently tries to hard in Presidential years to have its primaries first, that it tires and has to have its Senate primaries last. So we’re still on primary watch for that state, and it looks like the Republican race has shifted again.

Kelly Ayotte still leads the primary race to decide Democrat Paul Hodes’s opponent, but it appears the race for second is wide open now.

Before the big news was that Bill Binnie, from the left wing of the Republican Party, had surged and become the clear second place opponent to Ayotte. But according to Magellan Strategies, he’s back in third, with Ovide Lamontagne, self-described true conservative of the race, back in second: Ayotte 34, Lamontagne 21, Binnie 17, and Jim Bender brings up the rear at 13 (MoE 3.3).

Binnie’s unfavorables are way up now, and he’s at -24 net with a 30 favorable/54 unfavorable rating now. Meanwhile Ayotte rides high with +20 for 54/34, and the lesser known Lamontagne also sits at +20 38/18.

What I find most interesting is that Republicans and Independents support Bender and Lamontagne by the same amounts, but Binnie takes 10 points from Ayotte when looking at the likely Independent voters in the primary, nearly making the race a tie in that subset.

It’s clear, then, that independents may still approve of Bill Binnie, but Republicans increasingly do not, and Republicans seeking an alternative to Ayotte are turning to Ovide Lamontagne instead. That said, it’s still Kelly Ayotte’s show right now.

Possibly Related Posts

• May 31, 2010 -- Binnie closing on Ayotte in New Hampshire
• May 14, 2010 -- Ayotte cruises; Hodes passes Lamontagne
• April 21, 2010 -- Quick Hits for Wednesday

We’re seeing a real discrepancy in the California Senate polling. SurveyUSA’s last two results were Republican Carly Fiorina ahead 5 and and now 2, while Rasmussen Reports has had Democrat Barbara Boxer up 5.

Rasmussen has Fiorina behind with leaners 44-49 (MoE 4) for a 26% chance Fiorina leads, while SUSA has Fiorina up 48-46 (MoE 4.2) for a 59% chance, over double Rasmussen’s. I would love to know why this is, but Rasmussen doesn’t give us any crosstabs, only favorabilities, but SUSA doesn’t give us favorabilities. So we’re stuck.

All we know is that Rasmussen, of the alleged Republican House Effect, is still showing Fiorina behind 5, but SurveyUSA is consistently now showing Fiorina ahead. That showing of Boxer at 42 was clearly anomalously low, but the Democrat still trails. It will be interesting to see how long this gap between the pollsters persists.

In other results in this new SurveyUSA poll, Meg Whitman has a clear lead over Jerry Brown, 47-40. Whitman wins women 45-44, which I find interesting. If even a strong position in favor of abortion only brings the Republican into a virtual tie with the Democrat among women, then abortion is not the only issue driving the gender gap at all.

Possibly Related Posts

• June 9, 2010 -- Grading the Pollsters: June 8 edition
• August 26, 2010 -- Rasmussen on California: Whitman and Boxer up
• August 13, 2010 -- Boxer falls behind Fiorina, Whitman even with Brown

When the first major post-primary poll came out for the Washington Senate race, some questioned whether it was skewed toward Republican Dino Rossi over Democrat Patty Murray, or if it was simply an outlier.

But now we have confirmation that Rossi does appear to have a lead at this point, though perhaps not as large as SurveyUSA showed.

Not that Rossi and Washington Republicans are going to be disappointed with a 50-47 lead (MoE 4). Even if Murray shows a 36% chance of being ahead in this poll, as I calculate, a 64% chance to unseat the incumbent is something Dino Rossi should be pleased with at this point in time.

Seats like this I believe are the kind that turn a modestly good year for Republicans into a rout. Public Policy Polling’s Twitter feed suggested that this year could turn so bad for Democrats that Washington, Wisconsin, and California become the battleground. If that’s the case then a Republican Senate majority might not be probable yet, but it will at least be possible.

Possibly Related Posts

• June 26, 2010 -- Rossi Ties Murray
• May 6, 2010 -- A pair of updates from Washington
• April 12, 2010 -- Republicans need Dino Rossi for a chance at the Senate