CNN‘s Brianna Keilar (8/17/16) lectures the Trump campaign’s Michael Cohen on how the “polls…all of them” show that “you guys are down.” (image: MMFA)
Liberals had a good laugh this summer when CNN’s Brianna Keilar (8/17/16) insisted to Donald Trump campaign lawyer Michael Cohen that his team was “down” in the race.
“Says who?” he asked.
Keilar snarked back: “Polls. Most of them. All of them?”
At the time, it was the sickest of all burns. As the polls kept forecasting a clear win for Hillary Clinton, Trump fans went back and forth between complaining about a rigged system and accusing the polls of being dishonest.
Again, it was a good laugh for liberals—until Election Day.
As the dust settles and America comes to terms with the election of Donald Trump, it’s time to take a look at the embarrassment of the media class, with the failure of polling only one part of the story. For those who have bemoaned the mediocrity of corporate media, this might appear as a well-deserved comeuppance, but it also brings the uncomfortable upending of decades of common wisdom about media and elections.
Let’s start with the polls. Polling is inherently a part of the media landscape, as many polls are conducted by news organizations directly and drive much of their coverage. They are meant not only to allow readers to know where an election is headed, but to inform partisans about how to adjust their positions as the race progresses.
538 sometimes came close to predicting the actual winner of the election.
In our internet age, the polling industry has its own media sphere. Nate Silver’s 538 was perhaps the closest to accurate, as Clinton’s chances on that site (mostly) fell as Election Day loomed, although its fluctuating probabilities gave some the impression that numbers were moving to keep the clicks coming—not to mention that it still predicted a Clinton victory until well after polls began closing. Sam Wang of Princeton University is perhaps the most embarrassed, with his 99 percent likelihood of a Clinton win and his projection of 323 electoral votes for her. And Real Clear Politics, which simply averages a collection of major polls without the sorcery of Silver’s weighting system, always had Clinton in the lead after Trump’s brief post-convention bounce subsided.
What happened? There are a few possibilities. One is that many Trump supporters lied to pollsters, too afraid or ashamed to admit how they would vote. But another, more interesting theory is that polls focus on “likely voters”—as in, voters who have a record of voting in the past. To the extent that Trump successfully brought in a demographic that was previously alienated from politics, they were left out of the polling.
The Trump/Clinton debates: In the end, did they matter? (image: CNN)
Another anomaly is the effect of the televised debates. Since the infamous Nixon/Kennedy debate, debate performances have been thought to potentially make or break elections. Highlighting personality, charisma and the ability to reason quickly and respond to arguments, these matchups were seen as giving Americans the clearest idea of who they should choose. Donald Trump, again, was widely mocked for his debate performances for everything from his sniffling to his call to throw Clinton in prison. It didn’t affect his chances in the end.
And then there’s the newspapers. Other than the Ku Klux Klan’s little rag, Trump’s only major-paper endorsement was from the Las Vegas Review-Journal, recently acquired by right-wing billionaire Sheldon Adelson, while most other leading newspapers, including historically Republican papers, backed either Clinton or Libertarian Gary Johnson, or went the full mile and called on voters to simply reject Trump.
Editorial boards were once taken to reflect the mood of the local elite and intelligentsia, or at least sectors of it, and had always been thought to influence the local electorate. This election has proven those endorsements meaningless.
It is quite easy to celebrate the failures here. Who are the old bats rotting at the Wall Street Journal’s editorial board to tell me how to vote, huh? Polls during elections, especially with the proliferation of places like 538, are thought only to be daily reminders of the tight race that are great at raising our blood pressure but do nothing to inform debate about actual policy. And the televised debates? These seem to be more important for drinking games and late-night writers than anything else.
But with that comes a great unease. How should the media be covering these races, and how should public opinion be measured and absorbed? What kind of debates between candidates would actually matter?
These are the valuable questions that media organizations should be asking themselves in the wreckage of this election. For the loser was not just Hillary Clinton. It’s also corporate media.
Ari Paul has reported for the Nation, the Guardian, the Forward, the Brooklyn Rail, Vice News, In These Times, Jacobin and many other outlets.
“How should the media be covering these races,”?
I don’t know, but I’m pretty sure how they should NOT be covering them: as cheerleaders for particular candidates (as most “news” organizations have been doing)
For the past year, NPR has been running around like chicken Little yelling “The sky is falling” to make sure Clinton got elected (first, Sanders and then Trump was the falling sky)
Of course, NPR probably now has a good reason to be afraid because Chicken Little’s head may soon be on the chopping block.
If it were not so sad, it would be funny.
Its true that Corporate Media is failed to track what is actual happening. Its big blender. All the time, Clinton was favorite but at the election day Trump was winner.
Why people are protesting? Any body can answer?
“Nate Silver’s 538 was perhaps the closest to accurate, ”
As they say, close only counts in horse shoes and hand grenades — and Silver was not even close.
The only thing that matters is the final prediction just before the election and on that 538 was just plain wrong. That the prediction for Clinton and Trump converged a couple times during the cycle is meaningless.
People make a very big mistake when they assume that these prediction models are “science” — when they are really little better than the methods of palm and Tarot card readers and psychics (Senor Silver, gazing into his crystal ball before he makes a prediction.)
Unlike voodoo election psychics, scientists do not keep the details of their models hidden under magician’s cloaks.
Polling was unusually wrong this year because of the Undecideds. About 9% of them. Some were silent because ashamed of voting for a bigot, others just don’t tell; some women were intimidated by men in their families. I know some of this from phone-banking.
Whatever the reason, polling was not “wrong” so much as necessarily incomplete.
You seem to be overlooking another factor: cheating. Clinton *might* have carried Wisconsin and Michigan had voter ID laws and Republican intimidation not been in place. See also Greg Palast’s reporting on how Ohio deliberately removed security devices from voting machines that would’ve kept them from being tampered with.
In making their calculations, the pollsters underestimated voter turnout for Trump. Trump talking about “rigged election” egged on his followers to vote.
Sam Wang is not part of “corporate media”- it’s a Princeton University group. Neither is “larry sabato’s crystal ball- he is at UVA. don’t assume that because somebody got something wrong it was because of a bias. the available information pointed a Trump loss- also, did you notice that he did in fact lose the popular vote? I feel like something has gotten very weird lately at FAIR- are you trying to make salient points or just looking to rip “corporate media” for any reason at any time? Are you enjoying the Trump victory since he isn’t part of the “establishment”?
I would like to see the highly likely fraud element of this election discussed by media. I see now that one responder here, Gary Kleppe, comments on this too. I wrote a letter to the editor and submitted it to LA Times this morning about this concern, but it probably won’t be printed. Hillary and the Democrats were set up to not question the results of the election, and it can’t really be verified anymore in states where electronic voting machines were used. In states that use electronic voting machines, and where the governor and secretary of state are of the same party, investigations have occurred into using these machines to flip votes. One can read about investigations by Greg Palast and also by Harvey Wasserman, who was interviewed by Amy Goodman on DemocracyNow! February 23, 2016. He has a book that has since been published, co-authored with Bob Fitrakis, called The Strip & Flip Selection of 2016: Five Jim Crows & Electronic Election Theft.
I personally think that Donald Trump did not actually win this election with enough votes, and that is why the polls were way off and far more than half the people are absolutely devastated and shocked by the results.
These investigative researchers have written about how Presidential elections (and other offices as well), starting with 2000, have had incidents of tampering with electronic voting machines and tampering with voter registrations. Wasserman asserts in his interview with Amy Goodman in February (available on democracynow.org in the archives) that George W. Bush was never actually elected by the American people, either time. Republicans tried to keep Obama from being elected, but he had such overwhelming support, it was too much for their plan to work. Trump is not the legitimate leader of this country, though he is the recognized President-Elect. Extremely unfortunate outcome.
Regarding media and elections, I would like to see no paid political advertising at all. The airways are supposed to belong to the people. I would like to see candidates be able to get on free and inform the voters of their plans and positions, just them alone talking, or perhaps being interviewed.
i think debates with all candidates from all parties, with folks who are educators, not media people, asking questions would be helpful for voters.
This media circus that the elections have become is gaudy, crude, depressing, and an assault on democracy. I think democracy does not really exist here anymore, but I am keeping hope alive, have to. We may have just experienced a political coup of the type never seen before in this country, moving toward fascism and repression. Everyone is reeling, except the racist crazies who are sending hate mail to Move-on and their ilk.
Thank you for your responsible writing, FAIR. I would be curious to see your group’s comments on the fraud issue. If no one is covering it, maybe you can’t really write anything but that fact, since you cover accuracy in reporting. Any comments on the new book by Wasserman and Fitrakis? I haven’t read it yet. Greg Palast is amazing. I saw one of his films on this issue a few years ago and have read some of his writing.
I agree that one should look for evidence of fraud — in all elections.
But one has to be careful to distinguish between polls taken before the election — which is what the election psychic’s “predictions were based on and which are uncertain and therefore unreliable for many reasons, not least of all because the polling sample may not be representative of the overall voting population — and voting day exit polls, which are better, but still uncertain, nonetheless.
Pre-election polls are just not very good indicators of possible “vote flipping” and vote tabulation manipulation.
Exit polls can be used for that purpose, but only in cases where the difference between poll and outcome is fairly large (say > 2 or 3 %) because of uncertainty inherent to the exit polls.
The only way you can REALLY tell if post voting fraud has occurred is by looking at the actual voting data.
And the only SURE FIRE way one can do the latter is if one has paper ballots (that have been subject to a strict chain of custody) that can be manually recounted and tabulated manually.
There are also statistical methods that can be used on the election results (particularly on the results at various stages of the election) that can be used to highlight “anomalies” — unexpected patterns in the data.
Statisticians have done this and actually found “anomalous” (unexpected and even suspect) patterns in voting since about 2004 — and not insignificantly in the 2016 primaries.
See the report at Electoral System in Crisis
The findings of Bev Harris and others that are reported at Blackboxvoting.org about the GEM’s tabulation software (referred to by some as “Fraction magic”) which is used to tabulate about 25% of all votes in the country make it very clear that it would be VERY EASY to fix the outcome of a Presidential (or other) race for an entire state after the fact .
In other words, despite assurances by the media and even our President, the possibility of “election rigging” is VERY REAL and of great concern.
Having said all that, I would really have to see all the election data as the election developed (particularly for the states like Penn and Michigan) before I would even conclude that fraud was likely.
But that was not what the above post was about at any rate. The pre-election prognostications are really a lot of baloney, though folks like Nate Silver — who is not a scientist and not even a real statistician (he’s an economist pretending to be a statistician) would have us all believe otherwise.
Nate Silver is not a statistician, but he is most certainly a statustician: one who seeks media attention.
I think, US People selected the Trump because he is looking fair man. Many people saying that his behavior was not a democratic. He always like reality and do some thing in the same way.
If there’s one silver lining that came out of Trump’s victory, it’s this. The mainstream media tried so hard to vilify Trump and prop up Hillary. So the fact that they failed to convince the electorate is a huge step forward because it means they no longer have the power over the public that they used to, which is a very good thing.
As for polls, I’ve always thought them to be a gauge used by media to see how successful they’ve been at propagandizing the people.
It’s even worse.
Because most polls are not open to inspection (essentially black boxes), they are subject to manipulation for the actual PURPOSE of propagandizing.
And I suspect that many of them are. Another reason not to trust the results.
Pollsters and the media have become very ACTIVE participants in our elections which, I am pretty sure, is not part of their job description*
*though I may be wrong in the case of Fox News and NPR. In those two cases, it might very well be written into their contract that they have to support whoever the Republican or Democrat candidate is. respectively.
The Kennedy/Nixon debate isn’t infamous, it’s FAMOUS.
At some point in this nation’s history TPTB decided the people could no longer be trusted to pick a president, or even pick a presidential candidate for that matter. Since that point, probably somewhere pre 2000 presidential election, the interests of the American people have given way to the interests of a very select few secret hidden miscreants.
You say that Trump had little or no support from the corporate media but this simply isn’t true! Murdoch’s News Corp has been backing Trump all the way.
And what is it you libtards love so much about Communism?
Eh? What’s communism got to do with it? Libtards? What are you smoking?