Marilinda Garcia (R-NH) has led in two of 13 polls. Which makes her the favorite, according to one pollster.
On Friday evening, I was in line at Costco purchasing a pizza for a little family dinner (because, when you're married with two kids, that qualifies as a wild Friday night). I was scanning my phone for new polling after a fairly soft day of volume (which, regrettably, continued into the weekend) when I saw a series of tweets from northeastern-based folks live-tweeting an event sponsored by the New England chapter of the AAPOR (American Association of Public Opinion Research). One caught my attention and slightly amused me:
Democratic Party not a great brand name right now, per Andy Smith, which colors the entire race. #neaapor
— @skoczela
Of course, this statement had the unspoken implication that the Republican Party brand name was somehow better than the Democratic one, which has been disproven by virtually every poll taken in this cycle to date (including the most recent national poll: one released Monday morning by
Politico.) However, perhaps what Smith (who runs the polling shop for the University of New Hampshire) meant was that the gap between the favorabilities of the two parties has narrowed since 2012 (which it has), or perhaps he was referring specifically to New Hampshire.
This statement, on the other hand, seems tougher to defend, and shows that Smith has learned very little from his colleagues:
Both NH CDs likely GOP pickups, says Andy Smith #neaapor
— @NEAAPOR
Why that statement is particularly outrageous, and why Smith should have known better, awaits you past the jump, along with the surprisingly sparse
46 polls conducted since last Friday (Oct 17-20).
AR-Sen (Talk Business/Hendrix College): Tom Cotton (R) 49, Sen. Mark Pryor (D) 41
CO-Sen (Benenson Strategies—D): Sen. Mark Udall (D) 47, Cory Gardner (R) 44
CO-Sen (Gravis—R): Cory Gardner (R) 48, Sen. Mark Udall (D) 43
CO-Sen (Mellman Group—D): Sen. Mark Udall (D) 39, Cory Gardner (R) 37, Others 5
IA-Sen (PPP—D): Bruce Braley (D) 48, Joni Ernst (R) 47
IL-Sen (Southern Illinois Univ.): Sen. Dick Durbin (D) 47, Jim Oberweis (R) 37
KS-Sen (Monmouth): Sen. Pat Roberts (R) 46, Greg Orman (I) 46
KY-Sen (Rasmussen): Sen. Mitch McConnell (R) 52, Alison Lundergan Grimes (D) 44
KY-Sen (SurveyUSA): Sen. Mitch McConnell (R) 44, Alison Lundergan Grimes (D) 43, David Patterson (L) 5
KY-Sen (Western Kentucky Univ.): Sen. Mitch McConnell (R) 45, Alison Lundergan Grimes (D) 42
LA-Sen General (Multiquest Int'l): Sen. Mary Landrieu (D) 36, Bill Cassidy (R) 32, Rob Maness (R) 6
LA-Sen Runoff (Multiquest Int'l): Bill Cassidy (R) 43, Sen. Mary Landrieu (D) 40
MT-Sen (MSU Billings): Steve Daines (R) 47, Amanda Curtis (D) 31
NC-Sen (Gravis—R): Thom Tillis (R) 48, Sen. Kay Hagan (D) 43
NC-Sen (PPP—D): Sen. Kay Hagan (D) 46, Thom Tillis (R) 43, Sean Haugh (L) 5
NH-Sen (Suffolk): Sen. Jeanne Shaheen (D) 49, Scott Brown (R) 46
NH-Sen (Univ. of Mass/YouGov): Sen. Jeanne Shaheen (D) 48, Scott Brown (R) 45
AL-Gov (Cygnal—R): Gov. Robert Bentley (R) 57, Parker Griffith (D) 36
AR-Gov (Rasmussen): Asa Hutchinson (R) 49, Mike Ross (D) 47
AZ-Gov (Tarrance Group—R): Doug Ducey (R) 43, Fred DuVal (D) 36, Barry Hess (L) 5
CO-Gov (Gravis—R): Bob Beauprez (R) 48, Gov. John Hickenlooper (D) 44
CT-Gov (Rasmussen): Tom Foley (R) 50, Gov. Dan Malloy (D) 43
FL-Gov (Rasmussen): Gov. Rick Scott (R) 47, Charlie Crist (D) 47
FL-Gov (St. Pete Polls): Charlie Crist (D) 45, Gov. Rick Scott (R) 44, Adrian Wyllie (L) 8
IL-Gov (Southern Illinois Univ.): Bruce Rauner (R) 42, Gov. Pat Quinn (D) 41, Chad Grimm (L) 3
KS-Gov (Monmouth): Gov. Paul Davis (D) 50, Gov. Sam Brownback (R) 45
MA-Gov (SocialSphere/Boston Globe): Charlie Baker (R) 41, Martha Coakley (D) 41, Others 8
ME-Gov (Critical Insights): Gov. Paul LePage (R) 39, Michael Michaud (D) 36, Eliot Cutler (I) 21
MN-Gov (SurveyUSA): Gov. Mark Dayton (D) 50, Jeff Johnson (R) 40
NH-Gov (Suffolk): Gov. Maggie Hassan (D) 49, Walt Havenstein (R) 39
PA-Gov (Magellan—R): Tom Wolf (D) 49, Gov. Tom Corbett (R) 42
SC-Gov (Susquehanna Research—R): Gov. Nikki Haley (R) 51, Vincent Sheheen (D) 31
TX-Gov (Univ. of Houston): Greg Abbott (R) 47, Wendy Davis (D) 32
AR-01 (Talk Business/Hendrix College): Rep. Rick Crawford (R) 52, Jackie McPherson (D) 30
AR-02 (Talk Business/Hendrix College): Patrick Hays (D) 46, French Hill (R) 42
CA-17 (SurveyUSA): Rep. Mike Honda (D) 37, Ro Khanna (D) 35
MA-06 (Emerson Polling Society): Richard Tisei (R) 43, Seth Moulton (D) 40
MA-06 (Øptimus—R): Richard Tisei (R) 47, Seth Moulton (D) 45
ME-01 (Critical Insights): Rep. Chellie Pingree (D) 53, Isaac Misiuk (R) 19, Richard Murphy (I) 10
ME-02 (Critical Insights): Bruce Poliquin (R) 41, Emily Cain (D) 36, Blaine Richardson (I) 6
MN-08 (SurveyUSA): Stewart Mills (R) 47, Rep. Rick Nolan (D) 39, Skip Sandman (L) 4
MT-AL (MSU Billings): Ryan Zinke (R) 40, John Lewis (D) 33
NH-01 (Univ. of Massachusetts/YouGov): Rep. Carol Shea-Porter (D) 54, Frank Guinta (R) 37
NH-02 (New England College): Rep. Anne Kuster (D) 49, Marilinda Garcia (R) 43
NH-02 (Univ. of Massachusetts/YouGov): Marilinda Garcia (R) 48, Rep. Anne Kuster (D) 43
VA-10 (The Polling Company—R): Barbara Comstock (R) 51, John Foust (D) 35, Others 4
Andy Smith, meet David Paleologos.
(Actually, I am pretty confident that they've met.)
Even if neither name rings a bell immediately, their work should be familiar to the political junkie. Smith is the head of the polling crew at the Survey Center at the University of New Hampshire. Paleologos is the Director of the Political Research Center at Suffolk University.
Last week, as alluded to in the introduction, Smith gave a talk to the New England chapter of the AAPOR. In the midst of it, he declared that both of New Hampshire's Democratic incumbents, Carol Shea-Porter (NH-01) and Anne Kuster (NH-02) were "likely" to fall to their Republican challengers.
Now, let's be clear—it is entirely possible that both incumbents could fall on Election Day to their Republican challengers. New Hampshire has shown an odd tendency, over the last several cycles, to be very sensitive to national shifts in mood. What's more: they've tended to move in unison:
2006: NH-01—D+2; NH-02—D+7
2008: NH-01—D+6; NH-02—D+15
2010: NH-01—R+12; NH-02—R+1
2012: NH-01—D+4; NH-02—D+5
So, if 2014 really is a "Republican year", as many in the media have insisted, it sure seems to open up the possibility that Smith's bold declaration will come to pass.
That having been said, there is something inherently wrong with Smith's bluster on "calling" these races with more than two weeks to go. The issue here is that Smith isn't some ill-informed pundit—his survey center has polled NH-01 a half-dozen times. Plus, they've polled the race in NH-02 five times.
And, quite frankly, it is pretty tough to see in the data, even Smith's own data, where he gets the high degree of confidence in declaring these races likely GOP pickups.
Let's start in NH-01, which is yet another encounter between Democrat Carol Shea-Porter and Republican Frank Guinta. This is the third consecutive showdown between the two, with Guinta winning in 2010 and Shea-Porter seizing the seat back in 2012. In the fourteen polls of the cycle to date, Carol Shea-Porter has led in eight of those polls, and Guinta has led in five polls, with one tie. In the five polls released this month, Shea-Porter led Guinta by an average of 2.2 points (though there were two pretty clear outlier polls in there, including today's super-comical YouGov offering). Our own Daily Kos Elections race ratings have NH-01 as a Tossup.
NH-02, meanwhile, has typically been a little more remote as a prospect for the GOP. They have a young contender that they're high on (GOP state legislator Marilinda Garcia), but the data hasn't been quite as enticing for the Republicans. In the 13 polls in that race, Garcia has led in just two of them, and the average lead for the Democratic incumbent has been 3.8 percent (and that includes an equally eyebrow-raising Garcia +5 from that tiny-sampled YouGov poll). We here at DK Elections have that race as Leans Democratic.
So, right away, it is a bit tricky to make a clear, data-driven case for Smith's confidence. Neither Guinta nor Garcia have led in the majority of polls throughout the cycle, and neither of them hold a lead in the October polling averages.
Now, perhaps Smith is relying on the "50 percent rule" for incumbents, which is predicated on the theory that undecided voters will tend to break for the challenger, meaning the incumbent will score a percentage of the vote fairly close to the poll numbers. He hinted about concern for that when discussing the Shaheen-Brown race, after all.
The problem there is that it is rebutted by his own data. Look, if you will, at their final polls in the six races they polled in 2012. Sure, in a couple of them, the undecideds broke more to the challenger (NH-01, specifically, though it'd be a little bit of a stretch in that case to call Shea-Porter a true "challenger"). But, in others, the incumbent took the bulk of the undecideds (MA-Pres and NH-02, in particular).
Of course, if Smith needs more evidence of the folly of mid-October projections, he needs to hang out with Suffolk's David Paleologos. It was Paleologos, after all, who probably wishes he could have back this moment back in 2012 when he offered up this famous moment of pure derp:
“I think in places like North Carolina, Virginia and Florida, we’ve already painted those red," David Paleologos, the president of Suffolk University Political Research Center told Fox host Bill O'Reilly on Tuesday. "We’re not polling any of those states again. We're focusing on the remaining states.”
Of course, Obama went on to win two of those three states.
What made that call by Paleologos even more bizarre was the fact that Obama led in Suffolk's own polling in two of the states that Paleologos so memorably chalked off on Fox News. But he was below 50 percent, which presumably factored into the calculus for Paleologos when he made his bold call, one that ultimately blew up in his face.
One has to assume that Smith knows that, even if he truly thinks the two GOP challengers have the edge today, data close to an election tends to bounce around. Actually, as our David Nir so clearly demonstrated in a piece in August, Smith should know that better than anyone.
Hell, Smith may be right, when all is said and done. Both Shea-Porter and Kuster could be dispatched on November 4th. But there is simply no way he knows that right now, not at the level of confidence he projected last week. And for someone who is limned as an expert in political polling, he probably should know better. Clearly, he didn't learn from the embarrassment of Paleologos (or a similar face-palm moment for Mason Dixon's head, Brad Coker, who famously called Florida for Romney the week before the election).
But maybe, in the final analysis, Smith felt comfortable with his pro-GOP prediction because he knew that, if past is prologue, there really aren't any consequences for pollsters making errant election predictions on thin data. After all, Paleologos' outfit snared a new deal for the 2014 cycle as the exclusive pollster for media titan USA Today. Meanwhile, Mason Dixon, despite the Florida gaffe and an even uglier unforced error (where they had to retract a media poll for UT-04 because they had weighted using statewide numbers, instead of the data for the district), is still getting work. So maybe Smith knows that he can make such sweeping statements beforehand, and won't feel any pain even if his pre-election bluster manages to go awry.