Recent polling confirms Gary Peters (D-MI) is pulling away in his once-tight Senate race.
Memo to my pals in the U.S. political news media: look closely at the above caption. Now, I don't want to tell you how to do your jobs—but that caption has two things going for it.
For one thing, it is accurate. In the last six polls of the Michigan Senate race (released since October 10th), Democrat Gary Peters is leading Republican Terri Lynn Land by an average margin of 11.5 percentage points. In the October polls taken before that, Peter's average lead was 7.8 percentage points. September polls? 5.5 percentage points. The trend line is clear.
For another thing, these are actual polls asking the entire potential electorate who they are voting for. Not some subset, based on their particular view on an issue, or ... say ... whether or not they are following some "scandal" closely.
I bring this up because, quite frankly, some of our colleagues aren't quite "getting it." We are now rapidly approaching the finish line, and it's time to be on top of our game. And Thursday brought three glaring examples of total airballs thrown in the name of political analysis.
They were bad, they were avoidable, and we'll discuss them after the jump (along with the customary compendium of all the recent polling goodness, which numbers a fairly "holy crap" 88 polls today!).
CO-Sen (Ipsos-Reuters): Cory Gardner (R) 47, Sen. Mark Udall (D) 45
CO-Sen (Monmouth): Cory Gardner (R) 47, Sen. Mark Udall (D) 46
CO-Sen (PPP—D): Cory Gardner (R) 46, Sen. Mark Udall (D) 43
CO-Sen (Suffolk/USA Today): Cory Gardner (R) 46, Sen. Mark Udall (D) 39
GA-Sen (Insider Advantage): Michelle Nunn (D) 47, David Perdue (R) 45, Amanda Swafford (L) 4
GA-Sen (SurveyUSA): Michelle Nunn (D) 46, David Perdue (R) 44, Amanda Swafford (L) 4
IA-Sen (Monmouth): Joni Ernst (R) 47, Bruce Braley (D) 46
IA-Sen (Quinnipiac): Joni Ernst (R) 48, Bruce Braley (D) 46
IL-Sen (Univ. of Illinois-Springfield): Sen. Dick Durbin (D) 50, Jim Oberweis (R) 35
KS-Sen (Rasmussen): Greg Orman (I) 49, Sen. Pat Roberts (R) 44
KY-Sen (Voter/Consumer Research): Sen. Mitch McConnell (R) 49, Alison Lundergan Grimes (D) 41
MI-Sen (Clarity Campaign Labs—D): Gary Peters (D) 48, Terri Lynn Land (R) 33
MI-Sen (EPIC-MRA): Gary Peters (D) 45, Terri Lynn Land (R) 34
MI-Sen (Mitchell Research): Gary Peters (D) 51, Terri Lynn Land (R) 38
MI-Sen (PPP—D): Gary Peters (D) 53, Terri Lynn Land (R) 39
MN-Sen (SurveyUSA): Sen. Al Franken (D) 53, Mike McFadden (R) 38
NC-Sen (National Research—R): Thom Tillis (R) 42, Sen. Kay Hagan (D) 41, Sean Haugh (L) 6
NC-Sen (SurveyUSA): Sen. Kay Hagan (D) 46, Thom Tillis (R) 43, Sean Haugh (L) 7
NH-Sen (American Research Group): Sen. Jeanne Shaheen (D) 49, Scott Brown (R) 48
NH-Sen (CNN/ORC): Sen. Jeanne Shaheen (D) 49, Scott Brown (R) 47
NH-Sen (New England College): Scott Brown (R) 48, Sen. Jeanne Shaheen (D) 47
NH-Sen (PPP—D): Sen. Jeanne Shaheen (D) 49, Scott Brown (R) 45
NJ-Sen (Fairleigh Dickinson): Sen. Cory Booker (D) 56, Jeff Bell (R) 40
NJ-Sen (Stockton Polling Institute): Sen. Cory Booker (D) 57, Jeff Bell (R) 33
OR-Sen (SurveyUSA): Sen. Jeff Merkley (D) 53, Monica Wehby (R) 32
SD-Sen (Public Opinion Strategies—R [10/13]): Mike Rounds (R) 39, Rick Weiland (D) 31, Larry Pressler (I) 18
SD-Sen (Public Opinion Strategies—R [10/20]): Mike Rounds (R) 48, Rick Weiland (D) 24, Larry Pressler (I) 16
TN-Sen (PPP—D): Sen. Lamar Alexander (R) 45, Gordon Ball (D) 32, Tom Emerson (I) 13
TX-Sen (Univ. of Texas): Sen. John Cornyn (R) 57, David Alameel (D) 31
AR-Gov (Talk Business/Hendrix College): Asa Hutchinson (R) 49, Mike Ross (D) 41
AZ-Gov (McLaughlin and Associates—R): Doug Ducey (R) 37, Fred DuVal (D) 36
AZ-Gov (The Polling Company—R): Doug Ducey (R) 42, Fred DuVal (D) 35, Barry Hess (L) 7
AZ-Gov (Rasmussen): Doug Ducey (R) 47, Fred DuVal (D) 42
CA-Gov (PPIC): Gov. Jerry Brown (D) 52, Neel Kashkari (R) 36
CO-Gov (Ipsos-Reuters): Gov. John Hickenlooper (D) 45, Bob Beauprez (R) 45
CO-Gov (Monmouth): Gov. John Hickenlooper (D) 50, Bob Beauprez (R) 43
CO-Gov (PPP—D): Gov. John Hickenlooper (D) 45, Bob Beauprez (R) 44
CO-Gov (Quinnipiac): Gov. John Hickenlooper (D) 45, Bob Beauprez (R) 44
CO-Gov (Suffolk/USA Today): Gov. John Hickenlooper (D) 46, Bob Beauprez (R) 46
CT-Gov (Quinnipiac): Gov. Dan Malloy (D) 43, Tom Foley (R) 42, Joe Visconti (I) 9
FL-Gov (Quinnipiac): Gov. Rick Scott (R) 42, Charlie Crist (D) 42, Adrian Wyllie (L) 7
FL-Gov (St. Leo Univ.): Charlie Crist (D) 43, Gov. Rick Scott (R) 40, Adrian Wyllie (L) 8
GA-Gov (Insider Advantage): Gov. Nathan Deal (R) 44, Jason Carter (D) 44, Andrew Hunt (L) 5
GA-Gov (SurveyUSA): Gov. Nathan Deal (R) 45, Jason Carter (D) 43, Andrew Hunt (L) 4
HI-Gov (Ward Research): David Ige (D) 47, Duke Aiona (R) 35, Mufi Hannemann (I) 12
IA-Gov (Monmouth): Gov. Terry Branstad (R) 58, Jack Hatch (D) 37
IA-Gov (Quinnipiac): Gov. Terry Branstad (R) 55, Jack Hatch (D) 37
IL-Gov (APC Research/Chicago Tribune): Bruce Rauner (R) 45, Gov. Pat Quinn (D) 43, Chad Grimm (L) 4
IL-Gov (Rasmussen): Bruce Rauner (R) 48, Gov. Pat Quinn (D) 47
IL-Gov (Univ. of Illinois-Springfield): Bruce Rauner (R) 43, Gov. Pat Quinn (D) 41
MA-Gov (MassINC): Charlie Baker (R) 43, Martha Coakley (D) 42, Others 4
MD-Gov (WPA Opinion Research—R): Anthony Brown (D) 42, Larry Hogan (R) 41
MI-Gov (Clarity Campaign Labs—D): Mark Schauer (D) 44, Gov. Rick Snyder (R) 41
MI-Gov (EPIC-MRA): Gov. Rick Snyder (R) 47, Mark Schauer (D) 39
MI-Gov (Mitchell Research): Gov. Rick Snyder (R) 48, Mark Schauer (D) 46
MI-Gov (PPP—D): Gov. Rick Snyder (R) 48, Mark Schauer (D) 48
NH-Gov (American Research Group): Gov. Maggie Hassan (D) 53, Walt Havenstein (R) 43
NH-Gov (CNN/ORC): Gov. Maggie Hassan (D) 51, Walt Havenstein (R) 45
NH-Gov (New England College): Gov. Maggie Hassan (D) 51, Walt Havenstein (R) 43
NH-Gov (PPP—D): Gov. Maggie Hassan (D) 53, Walt Havenstein (R) 43
NM-Gov (BWD Global—R): Gov. Susana Martinez (R) 56, Gary King (D) 37
NV-Gov (Clarity Campaign Labs—D): Gov. Brian Sandoval (R) 68, Robert Goodman (D) 24
NY-Gov (Siena): Gov. Andrew Cuomo (D) 54, Rob Astorino (R) 33, Howie Hawkins (G) 9
OR-Gov (SurveyUSA—The Real One): Gov. John Kitzhaber (D) 51, Dennis Richardson (R) 38
RI-Gov (Brown Univ.): Gina Raimondo (D) 42, Allan Fung (R) 31, Robert Healey (Mod) 9
TX-Gov (Crosswind Media): Greg Abbott (R) 52, Wendy Davis (D) 31
TX-Gov (Univ. of Texas): Greg Abbott (R) 54, Wendy Davis (D) 38
WI-Gov (PPP—D): Gov. Scott Walker (R) 47, Mary Burke (D) 46
WI-Gov (Rasmussen): Mary Burke (D) 49, Gov. Scott Walker (R) 48
WI-Gov (St. Norbert's College/WPRI): Gov. Scott Walker (R) 47, Mary Burke (D) 46
AR-03 (Talk Business/Hendrix College): Rep. Steve Womack (R) 64, Grant Brand (L) 19
AR-04 (Diamond State Consulting—R): Bruce Westerman (R) 46, James Lee Witt (D) 39
AR-04 (Talk Business/Hendrix College): Bruce Westerman (R) 44, James Lee Witt (D) 42
CA-21 (SurveyUSA): Rep. David Valadao (R) 47, Amanda Renteria (D) 42
CA-31 (American Viewpoint—R): Pete Aguilar (D) 42, Paul Chabot (R) 38
CA-52 (SurveyUSA): Carl DeMaio (R) 46, Rep. Scott Peters (D) 45
IL-10 (We Ask America—R): Bob Dold (R) 47, Rep. Brad Schneider (D) 45
IL-11 (We Ask America—R): Rep. Bill Foster (D) 52, Darlene Senger (R) 40
IL-12 (We Ask America—R): Mike Bost (R) 44, Rep. Bill Enyart (D) 42, Paula Bradshaw (G) 6
IL-13 (We Ask America—R): Rep. Rodney Davis (R) 53, Ann Callis (D) 36
IL-17 (We Ask America—R): Rep. Cheri Bustos (D) 55, Bobby Schilling (R) 39
KS-02 (Tarrance Group—R): Rep. Lynn Jenkins (R) 49, Margie Wakefield (D) 37
MA-06 (Garin Hart Yang—D): Seth Moulton (D) 47, Richard Tisei (D) 36, Chris Stockwell (I) 9
ME-02 (Normington Petts—D): Emily Cain (D) 42, Bruce Poliquin (R) 34, Blaine Richardson (I) 8
MI-11 (Mitchell Research): Dave Trott (R) 47, Bobby McKenzie (D) 35, Rep. Kerry Bentivolio (W/I) 7
ND-AL (DFM Research—D): Rep. Kevin Cramer (R) 46, George Sinner (D) 39
NH-01 (New England College): Frank Guinta (R) 47, Rep. Carol Shea-Porter (D) 46
NY-04 (Siena): Kathleen Rice (D) 52, Bruce Blakeman (R) 42
Wow. Thursday was just a festival of electoral/polling derp, the likes of which I am not sure I have seen in the five years I have worked here at Daily Kos as a contributing editor. Three different examples, ranging from mildly annoying to absolutely awe-inspiring batshit stupidity, all contained within the same 24-hour span, flashed before our eyes right at the time that even the casual voter is starting to consume political coverage.
Two of them, while on the minor scale of cringe-worthy, were somewhat forgivable.
First, there was the Twitter maelstrom over the CNN poll out of New Hampshire. The number of reputable folks who referred to the poll as evidence of the "race tightening" was quite high. Which was a little perplexing, of course, to those of us who know that Republican Scott Brown actually, on the margin, performed two points worse in this CNN poll than he had managed in the previous CNN poll, which was conducted in mid-September. This was not a sign of a Democratic collapse, it was freaking float within the margin! And damned near everyone, including some names the casual political junkie would recognize, missed it.
It's a subtle point, but not an unimportant one. Not all pollsters are created equal. If a Sheehan fan wanted to be totally without scruples (and a poor poll analyst), they could say "Sheehan's improved five points in a week! She has the momentum!" Which they could claim, given that a poll last week had Brown +1 (New England College), and one this week showed Shaheen up 4 (PPP). So, there are two ways to semi-credibly gauge momentum. One way to do so is to compare polls by the same firm. Since the methodology is (presumably) the same, it makes for the fairest comparison.
So why is the Twitter freakout over the CNN poll forgivable? Because even if this poll (when compared apples-to-apples with the same pollster) showed no movement, the race on the whole has clearly moved. Which brings us to the second credible method for assessing movement and "momentum", which is to look at polling averages. And on that score, without a doubt, Shaheen has seen slippage. What was a 6-8 point Shaheen lead in the summer became 4-6 points last month, and is now 1-3 points. So, perhaps, the Twitter commentariat, while errant on the particular, was correct in the general sense that the overall race is tightening, even as they ignored the fact that this poll, when compared to its immediate predecessor, showed nothing of the sort.
The second leg in the journey that was "WTF?" Thursday was courtesy of a headline writer for the New Republic, who should otherwise be credited for giving a forum to an excellent piece by Princeton Election Consortium head Sam Wang. The piece is good, and should be read in full. But the header that TNR's person-in-charge-of-giving-shit-names decided to run with was: "2014 Midterm Elections: Republican Wave Won't Include Governorships."
Uh...what?!
Maybe I'm old school, or maybe just one of those people who annoys folks at parties by insisting on people explicitly defining their terms, but if it doesn't include governors races (which are a big part of the midterm election cycle), how the hell can you call it a "wave" of any kind? By definition, I hope we can all agree, a "wave" means a party has a distinct edge across the board. In 2010, you might recall (though Democrats would presumably prefer not to), the casualties were across the board, from House to Senate to Governorships to State Legislatures. As is the case in all wave years. You can't have a kinda, sorta wave.
To say nothing of the fact that incumbent Republican governors being in deep trouble in many of these races (which, polls show, is certainly a possibility) demonstrates that this cycle may be more about the perils of incumbency when everyone is ticked off, as opposed to a clear partisan tsunami like we've seen in the not-too-distant past.
But again, forgivable. Like I said, they ran with a good article, and one could assume that Assistant Editor in Charge of Pithy Headlines may not have perfect knowledge of defining "wave elections." On this one, I am willing to acquit.
Considerably less easy to condone, however, was the way far too many people in the press, people who really should've known better, gave light to that "game-changing" Oregon governor's poll, that wasn't really a poll (in the electoral/horserace sense), nor was it in all likelihood game-changing.
This may well have been the most aggravating unforced error, however brief, that I have seen from the political media in recent history. It started on Wednesday, when SurveyUSA released a survey, contracted by Oregon-based media outlet KATU-TV, which sought to look at the impact of recent revelations involving Democratic Gov. John Kitzhaber's fiancé, Cylvia Hayes.
The very premise, quite frankly, was idiotic. Above, in the week's polling links, you will find a poll ... by SurveyUSA, no less! ... which showed that Gov. Kitzhaber was leading his Republican opponent, state legislator Dennis Richardson, by 13 points. This margin was essentially unchanged from last month's SUSA poll, which was conducted before the recent media storm over Kitzhaber's fiancé.
The stories, which have been flogged fairly relentlessly, were known prior to this weekend's SUSA poll. A quick look at the Google Trends shows that the peak interest in either Kitzhaber or Hayes came in the week before the most recent SUSA poll was in the field.
Ergo, there was no need to publish this follow-up poll, because the question had already been answered. No, the Hayes story had not impacted voter perceptions of the election.
So, why do the second poll? Not to interpret the election, quite clearly. To influence it? Perhaps. After all, when you limit your polling universe to only people who are closely following the story (read: people obsessed with "scandal" or folks who ain't all that crazy about John Kitzhaber), and you prime them by implying that the allegations are something worth changing their vote over, can you really expect anything other than the result that they received, which was a somewhat sizable mass of scandal junkies saying that they were planning on switching their vote? The Richardson campaign, which has struggled for relevance pretty much since day one, and has trailed by an average of 11.5 points in the polls conducted since September, couldn't have received a better in-kind contribution than the one that they got from KATU-TV and SurveyUSA.
And lo and behold: a cursory glance of the sponsor for the poll (a representative article is right here) makes it look like a Richardson oppo researcher's haven. And, if you look at the very bottom of the website, you can see who owns the TV station.
The Sinclair Broadcast Group. As in "Stolen Honor" in 2004. That Sinclair Broadcast Group.
What was really galling, as I mentioned casually (or, perhaps, seethed) to Mediaite's Evan McMurry (who wrote an excellent read on the whole matter) was the number of top-flight political writers who got sucked into this non-story. Early Thursday morning, journalists all the way up to the national level were breathlessly tweeting out the poll as if, as the KATU article somewhat deceptively implied, the election had really shifted to the tune of 30-ish points overnight.
While that is on KATU (and, by extension, SurveyUSA), these guys honestly should've known better. The original SurveyUSA polling memo was not hard to find (I found it, after all), and was in circulation long before the "ZOMG!" tweets hit the Twittersphere. It would have taken less than a minute to look at the sample composition of the poll and realize something was hinky. Hell, to their credit, in SurveyUSA's poll memo, they even make clear that this sample does not necessarily sync up with the Oregon electorate. In itself, that was a little bit of CYA—they had just polled there two days prior, they knew it did not sync up with the Oregon electorate.
To believe the premise of the poll, in the way that these high priests of the political media did, you'd have to believe that (a) the story had been out a week; (b) SurveyUSA had polled October 16-19, and found Kitzhaber up 13; (c) SurveyUSA polled again from October 20-21 and suddenly found Kitzhaber down 16.
Unless Kitzhaber and Hayes had gone out on the night of October 19th and spent the evening insulting the mothers of roughly 29 percent of Oregon's electorate, this seemed a tad unlikely. But, even if for an hour or so, way too many reporters bought in. A failure to read the polling information carefully, and once again failing to make an apples-to-apples comparison (the full electorate versus a group interested in a particular event, like a glorified debate poll), led to some real shoddy work in the final analysis.
It's less than two weeks until Election Day, and the slowly building interest among the general electorate (read: not the crazy-ass political junkies like you and I) means that the volume of electoral coverage in our media will necessarily tick up. Which is fine. In fact, it is good. As long as it is done well. Thursday was not a good day for those demanding halfway decent electoral analysis. Perhaps our political press needs to take a quick look in the mirror, steel themselves for the final week ahead, and mutter something along the lines of "I can do better. I will do better."
Though, admittedly, the odds of that happening are pretty damned slim.