Trump Says Polls Showing Clinton Winning Are ‘Phony.’ His Argument Is Ludicrous.

By
Trump loves polls if they favor him; otherwise, they are “phony.”Photo: Mary Schwalm/AFP/Getty Images

At a time when Donald Trump and his supporters are trying everything imaginable to undermine the credibility of public-opinion polls showing his candidacy sinking slowly into the pages of American history, they will clearly pick up any rock they can and hurl it at the hated messengers of bad news. But Team Trump’s latest effort to turn an eight-year-old email about a totally noncontroversial polling practice into proof of an ongoing Democratic conspiracy is so blatantly cynical that it bespeaks contempt for the mogul’s followers.

Here’s the backstory: One of the stolen Podesta emails released by WikiLeaks was from a well-known progressive activist named Tom Matzzie, who in early 2008 recommended that the Clinton campaign do some polling that “oversampled” for this or that demographic in this or that place. That kind of thing is totally standard when a campaign wants to get a read on a specific demographic and has nothing to do with the kind of public polling that now shows Hillary Clinton winning the presidential election handily.

Dug up and misinterpreted by a pro-Trump website called Zero Hedge, this less-than-shocking and ancient news was blasted into the right-wing media world by Drudge, and then promoted by Trump himself on Twitter (“Major story that the Dems are making up phony polls in order to suppress the the Trump . We are going to WIN!”). Oversampling is the standard practice of creating a subsample large enough to make findings about a given group’s views statistically significant. It has been turned into a sinister-sounding method of “skewing” polling results. It is anything but that, as the Washington Post’s Philip Bump patiently explains: “Oversampling” has zero impact on a poll’s topline results (the horse-race Smith-versus-Jones numbers the public generally sees); it is only used for subgroup analysis via cross tabs.

Even if oversampling somehow did affect topline results, it is unfathomable that their use in an internal Clinton poll in 2008 has anything whatsoever to do with the various public polls coming out right now showing Trump slipping toward Palookaville.

In fact, the whole supposition of this line of attack is so mind-numbingly ludicrous that both Drudge and the Trump campaign (I don’t know enough about Zero Hedge to say the same of its writers) must know it is total BS and are nonetheless trotting it out there to inflame Trump’s supporters. You wonder why they even bother to link such claims to vaguely empirical data, however irrelevant and unverified, as the WikiLeaks material. They clearly think their audience will believe anything evil of the hated opponent and the hated media. And it is particularly rich coming from a campaign allegedly being run by a professional pollster and a candidate who touts favorable polls as though they have come down from Mount Sinai on stone tablets.