The Unbearable Triteness of Polling: Monmouth Polls Find Clinton Supporters Support Clinton

Hillary Clinton, former Secretary of State and 2016 Democratic presidential candidate, right, and Senator Bernie Sanders, an
Hillary Clinton, former Secretary of State and 2016 Democratic presidential candidate, right, and Senator Bernie Sanders, an independent from Vermont and 2016 Democratic presidential candidate, applaud and stand on stage together during candidate introductions at the Jefferson-Jackson Dinner in Des Moines, Iowa, U.S., on Saturday, Oct. 24, 2015. With Vice President Joe Biden officially out of the presidential race, the nation's first nominating contest between front-runner Clinton and Sanders is gaining steam, according to a new Bloomberg Politics/Des Moines Register Iowa Poll. Photographer: Daniel Acker/Bloomberg via Getty Images

CLINTON: What do you read, Sanders?
SANDERS: Polls, Polls, Polls.

A Monmouth poll released this week showed Clinton statistically tied with Sanders in New Hampshire, 48 to 45 (The margin of error was 4.9 percent). But the criteria for the poll are odd: according to Politico (who didn't note that Clinton's lead was within the margin of error, because stats-yawn), the poll surveyed "voters who voted in a primary election in either of the previous two cycles or who voted in the past two general elections in 2012 and 2014 and said they are likely to vote in the Democratic primary in February." If that seems like a somewhat tortured standard that might exclude a lot of people who plan to vote, it's because it is. Monmouth used essentially the same criteria for a recent Iowa poll, which Huffington Post Polling Director Ariel Edwards-Levy wrote about last week:

Unlike some other survey houses, which have relied on voters' own assessments of whether they'll participate, Monmouth surveyed only registered Democrats who both voted in at least one of Iowa's last two state primaries (which in 2014 featured uncontested Senate and gubernatorial races) and said they were likely to attend the caucuses next February.

As The New York Times' Nate Cohn noted, those requirements could knock out many first-time caucus-goers, voters who aren't affiliated with a party, or those who vote in presidential caucuses but not most primary elections. Those groups helped President Barack Obama pull a win in 2008 and are more likely to lean toward Sanders in 2016.

Cohn should be quoted here himself, as the numbers involved are somewhat staggering:

The all but complete exclusion of young voters also works to Mrs. Clinton's advantage. The most recent Quinnipiac poll gave Mr. Sanders a 67-to-23-point edge among 18-to-34-year-old voters; he trailed by 73 to 20 among voters over age 65.

Sanders dominates with young voters, which makes their effective exclusion from Monmouth's Iowa poll dubious. Since Monmouth used the same method in New Hampshire, young voters would be just as effectively cut out of that poll as well. (Ironically, in New Hampshire that bias only resulted in a statistical tie; but still, it contributes to the fanciful yet prevalent narrative that Clinton is surging and Sanders is slipping.) It doesn't make sense; but then again, it wasn't sensible that every pundit called the first debate for Clinton, while every online poll and focus group that night called it for Sanders.

According to Cohn, Monmouth's polling director Patrick Murray said: "I don't feel this is 2008... I don't think we're going to get that big number who have never participated in a primary before." Edwards-Levy says Murray "noted in an email that Democratic caucus turnout in Iowa has varied substantially over the past four decades, and said that he based the poll's turnout model on the 2000 contest between Al Gore and Bill Bradley." (You'll recall that wasn't exactly an inspiring primary; and that cycle's general election gave us George W. Bush. Maybe Murray thinks we should be getting ready for President Trump.) In other words, they didn't think it was important to include, as Edwards-Levy put it, "first-time caucusgoers, voters who aren't affiliated with a party, or those who vote in presidential caucuses but not most primary elections" -- demographics that were instrumental in Obama's 2008 victory, and who are likely to support Sanders.

This is establishment bias in almost pure form: despite an immediate, recent example in which certain demographics were decisive, Monmouth arbitrarily decided Sanders' explosive grassroots support (which often outstrips Obama's pace) would not translate to a similar turnout at the polls, and so polled in a way that excluded those demographics. Put simply, establishment bias like Murray's determined Monmouth's polling methodology, blithely cut essential Sanders supporters out of the picture, and, surprise surprise, Sanders didn't do as well.

You don't need a tinfoil hat, here; Murray isn't necessarily a Clinton shill. He might quite simply be participating in the same pattern of denial that conventional political wisdom has displayed about Sanders since he announced. Even so, somewhere within all of that denialism, there must be those who know precisely what they are doing when they apply a finger to the scales in these and other contexts. Those people know that Sanders is doing far, far better than the mainstream narrative would have us believe; and if they are Clinton supporters (really, who else could they be), they are running scared.

Ponder this for a moment: it takes a flawed, deeply biased polling methodology to obscure Sanders' real status, and to make Clinton seem like a dominant front-runner. That kind of shenanigans and overwhelming pundit-bias may well be the only things standing between Clinton and complete rout. Sanders' success is, and continues to be, an untold wonder of the political world; as in, it makes you wonder how much longer the fiction of Clinton's pre-eminence can be maintained.

The Clinton campaign has been spending money like crazy. Sanders by contrast just made his first ad-buy -- his first! His numbers in the polls have thus far been almost solely due to his diligence on the trail and his dominance of social media (since conventional media tend to give him such short shrift).

This is the story of the election, and it is a huge one: Sanders' success, despite his eschewing almost every single "essential" ingredient of the typical political campaign, is so monumental that almost no one in the media even begins to know how to talk about it, or likely even think about it -- and so they ignore it.

Conventional political wisdom still clings fanatically to the recipe that Sanders studiously and deliberately avoids. If he clinches the nomination (and, to be honest, if you get outside the conventional wisdom enough to see what his campaign has accomplished, that seems quite plausible), history may record this election as the one where major-media political coverage was rendered obsolete; and perhaps even deem it more revolutionary than 2008. Sanders himself has made no bones about calling for a revolution. True to prediction, it has so far not been televised.

Apparently, it does not need to be.