House Prediction (Oct. 19 Edition): 223R-212D; Senate Prediction: 52D-48R

With fifteen days to go before Election Day, polls are rolling in from a wide variety of House and Senate races, some of which need re-rating.  I see a modest amount of movement in polls that could make Election Night slightly more blue in the Senate and House than appeared true last week, as this post explains.  This post supplements mine of September 12 predicting the composition of the House as 218D-217R, my similar October 2 post, and my October 23 post suggesting a GOP takeover, 225R-210D.  My Senate forecast moved from 53D-47R on September 22 to 51D-49R last week.  This post is shorter because it supplements and tweaks those earlier predictions based on current poll movement and campaign information.

House Supplement:  Democrats Projected To Be Six Seats From Holding the Majority

Last week, I showed the Democrats at 210 seats, eight below the total needed to stave off Speaker Boehner.  This week I have three re-rates toward the Democrats, and one toward the Republicans, netting +2 to Blue Person Group.  This leaves the Democrats six seats shy of the majority.

AZ-03.  Folks, I had written back on September 12 and again on October 15 about the possibility that Democrat Jon Hulburd could pull the upset in this redder-than-Bozo's-hair district over Republican Ben Quayle, who has faced charges that he blogged as Brock Landers, a fictional character in turn portrayed by Dirk Diggler, a porn-star character played by Mark Wahlberg in the movie Boogie Nights.  (Got it?)  Turns out, as I suggested a month ago, lots of conservative Phoenix-area Republicans don't want "Brock Landers" for their Congressman after all; a new poll by PPP shows Hulburd winning independents by 14 points, and shows Landers/Quayle with a 52% disapproval rating in this fire-engine-truck red district.  Rerating from jump ball leaning to Quayle (which we had here first, when NYT and Nate and others had Hulburd as an impossibility), to jump ball leaning to Hulburd.  Democratic pickup.

IL-17:  Phil Hare (D), former Chief of Staff and successor to Rep. Lane Evans in this western Illinois district, is at best roughly even with Bobby Schilling, the owe of St. Giuseppe's Heavenly Pizza in Moline.  Polls show Schilling moving slightly ahead, but this one remains within the margin for error.  Given the strong GOP tailwind in this cycle, Hare's incumbency, and his awkward moment suggesting that he was not concerned about the Constitutionality of HCR, a gift that keeps giving to Schilling on YouTube and across the net, I will re-rate to give a narrow edge to Schilling.  Republican pickup.

NH-02.  Here, veteran House member Charlie Bass tries to reclaim his seat (recently held by Paul Hodes, who vacated it to run for the Senate) in a race against Ann Kuster.  A poll by WMUR/University of New Hampshire shows Bass steadily losing independent support.  This makes sense, given that he is more an incumbent than Kuster.  Also, Bass now faces a financial scandal, calling into question whether he profited from his prior term in Congress.  (Oops, GOP primary voters in NH-02!)  WMUR has the Democrat losing traction in neighboring NH-01, and had Kuster down by a lot earlier this year before a slow trend up, disabusing me of the fear of a house lean.  Moreover, New Hampshire is going to send Kelly Ayotte to the Senate over Hodes; up in the Granite State they split logs and tickets with equal aplomb.  I rerate from likely Republican win to narrow Democratic advantage.  Kuster should pull this one out.  Democratic hold.

NY-23.  Here, incumbent Democrat Bill Owens won a three way race in a 2009 special election.  In my last post, I suggested that Owens would likely lose a squeaker because Doug Hoffman, the Conservative Party candidate who caused the three way split last time around, dropped out, thus consolidating Republican support.  Since then, I saw a poll indicating that Hoffman -- who is still appearing on the ballot -- polled at 15% most recently, and then 1% when respondents were told he dropped out.  The problem for Republican Matt Doheny is that the district is rural and sprawling, and that some folks won't know that Hoffman is off the ballot.  Because Owens appears headed for the upper 40s, Hoffman receiving 5% or more should be fatal to Doheny.  (So much is made of the Republican registration advantage in NY-23, but Obama won it 52-47.)  Based on the most recent polling, I re-rate to jump ball, leaning Owens.  Democratic hold.

And while I won't re-rate it now, keep your eye on MI-07, where JD Hayworth lookalike and former Rep. Tim Walberg (R) is slowly losing momentum to first-term incumbent Mark Schauer, who has a huge cash advantage.  I'd really like to see some more polling on this one, which, given Walberg's weaknesses, could still be a Democratic hold, though I've rated it otherwise for now.  Meanwhile, Titus-Heck in NV-03, Mitchell-Schweikert in AZ-05, and Pearce-Teague in NM-02 are all hot races that could flip blue in this column, putting the House almost back within Democratic reach… but not this week.

I suggested last week in response to a question from reader-blogger Obey (hey, how can you not respond to a dude whose handle is in the imperative voice?) that I would provide confidence levels for predictions in my next post.  Not enough time to do so.  Am planning on producing an omnibus rating of all House races, along the lines of my September 12 post narrating 100 of them, placing them in a rough hierarchy of likelihood.  If I actually do that, which could consume the entire coming weekend, I will also supply confidence levels.

Senate Supplement:  Democrats Lock Up Fiftieth Seat, Claw Way Toward Fifty-Two

I wrote last week of the six races that would decide the Senate, with Democrats controlling 48, and Republicans 46.  The update this week is that the Senate is locked up.  Done.  Book it.  The Democrats are in the fifties, and the question is where they land, with a miniboomlet in Senate polling in the past week.

Two Races That Are Essentially Done, Putting the Democrats To Fifty

California.  If you look at the last fifteen polls of Boxer-Fiorina, they have one thing in common.  Despite their relatively narrow margins, Boxer wins all fifteen.  Placing to one side all Scott Rasmussen's hyperventilating that the race is essentially tied, and within the margin of error, funny how Fiorina hasn't led a poll in six weeks and counting.  Congratulations, Senator Boxer.  We're done analyzing your race.  This is seat 49 for the Democrats.  RealClearPolitics, get a clue and remove it from the Tossup menu.  Size of margin and certainty of outcome aren't the same thing.  This is a highly stable three to four point race.  Deal with it.

Washington.  Patty Murray is very close to the same pattern as Barbara Boxer.  While Washington Senate polls show far greater variability, with Rasmussen suggesting a Rossi lead 1-3 percent in three polls falling between September 28 and October 9, Murray leads by 7.4% in the last five polls, which includes the newest Rasmussen, showing her up 49-46.  This one is very close to salted away.  Seat 50. 

Three Races Trending Toward the Democrats, Two of Which Look Like Pickups

West Virginia.  Our sometime guest readytoblowagasket might observe that Articleman doesn't know much about West Virginia.  Uncle, already.  I guess in the language Rumsfeld, West Virginia is simply a known that I don't know.  Having originally rated West Virginia solid for Governor Joe Manchin, the Democrat, and then predicted that Republican John Raese's September advantage would hold up, we now have two polls showing Manchin up since it emerged that Raese's admen recruited "hicky" Philadelphians to represent West Virginians in a campaign ad, a third showing him tied, and Rasmussen alone (pattern, anyone?) showing Raese up 3, but with rising negatives.  Manchin's favorables are perched at 70 percent, his unfavorables at 28.  With the electorate appearing increasingly Democratic in polls, with Manchin launching media, I admit I don't know anything about West Virginia, and re-rate this as a narrow Democratic advantage based on polling published since our last update.  Seat 51.

Illinois.  States tend to vote what they are.  Illinois is a blue state.  President Obama has a net positive favorability there, with 52% approval, per Rasmussen.  Giannoulias led in the last Rasmussen poll, as he is getting a percentage of votes that exceed his favorability rating, which is consistent with how blue Illinois is.  President Obama's last visit helped solidify the black vote in Illinois, and a further visit by the President could give Giannoulias the burst he needs, especially with the ugly capture of Republican Senate candidate Mark Kirk gearing up to dispatch poll watchers, out of alleged concern that black voting numbers could be fraudulently overstated.  Nothing like a good Republican voter suppression narrative to get out the urban Democratic vote.  The polls have been within the margin of error for so long, it's really up to intangibles.  I'm an Illinoisan by birth and early adulthood, I know what the Democratic turnout machine can do in Chicago, and I see the state giving the nod to Giannoulias, albeit barely and reluctantly.  Seat 52, by a margin requiring recount.

Colorado.  Michael Bennet (D) is peaking at the right time.  Rasmussen shows him coming within 2 at 47-45, but that was with a sample that was more blue than Colorado is.  Still, PPP showed him up by a point recently.  And last Sunday's appearance on Meet the Press generated some buzz with Buck comparing homosexuality to alcoholism, which helps equalize the local enthusiasm gap.  Buck wins on the economy, and in 2010 Colorado, loses by some margin on the culture issues.  I predicted several posts and weeks ago that Buck would win by 2.  I still hold to that, but given that the narrative was that Buck had close to put this race away, one would have to ascribe some momentum to Bennet and some hope, I suspect about a 15% chance, that he pulls the upset.

The Race With Anglementum

Nevada.  Harry Reid had a boomlet after his wished-for opponent, Far Right Angle, appeared.  And polls from mid-July to mid-September, typically showed him ahead, albeit narrowly, as his sustained barrage of negative media drove Angle's already high negatives to a height heretofore unattained outside laboratories, and then only for the briefest of periods.  Since then, Angle has hit back hard, with enormous advertising of her own, and on her behalf.  

It is hard based on recent surveys to see Harry hitting 48%, which is probably what it takes to win this race.  Two wild cards are these, and surveys do not typically account for them:  None of These Candidates, and Tea Party nominee Scott Ashjian.  Nevada lets voters vote for "None of These Candidates" and in a Suffolk University survey published on October 14, None polled at 3%, near the historic high-end of the range it attains in statewide races (no, it can't be elected, nor the seat vacant if None gets a plurality, Nevada elects the highest-tallying named candidate, even if None beats them like a drum.)  Ashjian polls at 2%.  Assuming None and Ashjian poll roughly consistently on November 2, then Reid still needs 47.5% to 48% to win.  And he's just not showing that strength in the recent polls that show Angle ahead by a bit.  The Suffolk survey shows Reid up, but its sample is D+5, and Reid only leads by 3, and there are too many undecideds for the poll to be accurate, as other polls uniformly suggest there are none left.  The Las Vegas Review-Journal puts Angle up 2 at 48-46.  That sounds about right, as poll internals show Nevadans having a very negative view of Angle, but with some minority of those who really dislike her willing to vote for her anyway.  I see her winning by 1-1.5%, at roughly 48-47.  When you hang your hat on driving your nut down to 48%, it's best if your polling suggests you can hit 48%.  

Please join me on November 2 for my liveblogging.  You can comment, send in questions, or throw tomatoes.  All good.  Hope to see you then.

Also, subscribe to my RSS feed here.

I'll also be on www.krxa540.com on Wednesday at 805 am Pacific/1105 Eastern to discuss politics.