Brexit: Yet another example of biased polling
For the weak-minded among us who continue to believe in polling data, the Brexit vote shows – yet again – what a biased absurdity the polls have become.
But delusions and obfuscation runs deep in the mainstream trenches.
Nate Cohn at the New York Times writes that "[i]t was not a cataclysmic polling failure[.] ... [I]t's hard to argue that this was a big polling failure."
Sure it wasn't. Once you are done rolling around on the floor in laughter at claims that this wasn't a massive polling failure, read on.
To review, the Brexit side won by 4%, 52% to 48%.
Not a single one of the well known polling aggregators/predictors picked Brexit in their last-minute final projections.
- Elections Etc. had the Remain side up by 6.6%, 53.3% to 46.7%.
- Number Cruncher Politics had the Remain side up by 6%, 53% to 47%.
- What UK Thinks: EU had the Remain side up by 4%, 52% to 48%.
- The Financial Times had the Remain side up by 2%, 48% to 46%.
- Britain Elects had the Remain side up by 1.6%, 50.8% to 49.2%.
- HuffPost Pollster had the Remain side up by 0.5%, 45.8% to 45.3%.
- The Economist's poll tracker had the race tied at 44% apiece.
Thus, we had a systematic bias in the aggregated polling data that ranged from 4% to almost 11%.
Individual polls leading up to the vote were publishing ridiculous results. In the week prior to the vote, 9 of the 13 polls predicted a victory for Remain ranging from 1% up to 10%. Just three polls had Leave in the lead, but just by 1% to 3% – i.e., still below the actual margin of victory – and one poll had a tie. Not a single individual poll got the result correct, or overpredicted a Leave win.
According to Sky News, "Ipsos MORI extended their fieldwork into the final day of campaigning and updated their final predictions from a four-point to an eight-point Remain win." You mean the same Ipsos that apparently has political connections that would lead to reasonable concerns about a pro-EU bias?
The overall bias in favor of Remain was effectively uniform, which is statistically impossible if the bias was random. The bias was systematic.
There is every sort of faux intellectual handwaving explanations being bandied about over why the Brexit polls were wrong. Follow the money, and you will find the answers you seek.
Tinfoil hat time. Funny how in the last couple years, the polls have been so wrong, and it seems always in the direction against the cause du jour of the leftists.
Just take the last U.K. general election, which polling data spectacularly failed at. As noted by Dan Hodges in the Telegraph at the time:
We do know why the polls were wrong. And it has nothing to do with sampling or turnout or demographics or Shy Tories or Lazy Labourites or a last minute swing. The polls were wrong because the polling industry collectively and deliberately and cynically manipulated its own findings. In other words, the pollsters lied to us all...
Since the election, the narrative has been "the polls were wrong". But what people are conveniently forgetting is that during the election the narrative was different. Back then, it was "why are the polls all over the place?". More specifically, it was "why do online polls show a dead heat, but phone polls show the Tories beginning to build a quite significant lead?".
But then something strange happened. Over the final 72 hours this discrepancy between the phone polls and online began to vanish. The polls suddenly began to cluster. The wide variation between the polling companies -- variation that we had been witnessing for the best part of five years -- mysteriously stopped. All the pollsters were suddenly, and miraculously, in agreement. The election would be a tie, give or take a single point in either direction.
I say "strange", but it wasn't strange. It was wholly predictable. Cynical, disreputable, despicable. But predictable.
It's what known in the polling industry as "herding". And herding, not to put too fine a point on it, is when pollsters cheat. Each polling company knows that however accurate their results are, they will ultimately only be judged on one poll. The final poll before the election.
Which presents them with a dilemma. What if their poll is at variance with all the other polls? What if that vital final poll looks like a rogue? What if they, alone, stand out from the crowd? What, God forbid, if they alone are wrong?
The commercial and reputational risk is too great. So they herd. They deliberately manipulate their results to bring their own figures back into the pack... Polls aren't just used to predict election results, they're also used to try to influence election results.
There are multiple problems in polling, and herding is certainly one of them.
The one at play over the past couple weeks in the United States was pure, unadulterated liberal bias in an attempt to sway Donald Trump's campaign. And it apparently worked.
After correcting for clear liberal bias in all polling data released during the past several weeks, each and every major poll shows Trump either tied with Hillary Clinton or, more probably, out in front. And yet Trump himself was on Lou Dobbs's show this week giving credibility to a flawed CNN/ORC poll that had him 5% behind Clinton.
On Bill O'Reilly's show following the Corey Lewandowski firing, O'Reilly asked Trump if he had internal polling data to help assess whether or not the other polling data presented in the mainstream media is real. Trump's reply clearly suggested that his team had no internal polling data. While that may be fine at this early stage, then at least don't start regurgitating the results of polling data biased against you and claiming that you are losing as if it were real. Then, just a day or two after using CNN polling data as if it were unbiased, Trump blasted CNN as biased over its fact-checking of his speech on Clinton. If, as Trump says, CNN is the "Clinton News Network," why is he giving their polling data credence? Surely if you think global warming is a worldwide scam, that vaccines cause autism, and that the Iraq War was based on false intelligence, in the Maslow's hierarchy of credible conspiracies, a liberal bias in polling data should have been front and foremost in Trump's mind.
We now know that Lewandowski had Trump clearly in the lead as of late Sunday/early Monday, immediately before he was fired, purportedly in part, for leading a losing campaign. And GOP establishment operatives continue to take over Trump's campaign. This may not end well for Trump, not just because the establishment is clearly not competent at winning presidential elections, but more precisely because Trump's base comprises entirely anti-establishment voters who rejected the establishment time after time in the primaries.
Utah, which the leftists and the GOPe were screaming about as in play in the run-up to Lewandowski's removal, was clearly back in Trump's side according to a poll conducted June 8-17, but whose results apparently did not reach the mainstream media until June 21.
One is reminded of the infamous sketch of Lucy, Charlie Brown, and the football. The leftists and GOPe keep holding the ball for the populists to kick, and we know how that story ends. Looks as though it may be the same ending this time around again.
We'll have to wait and see where the corrected polling data over the next couple weeks takes us, but there is preliminary evidence emerging that Trump's lead in North Carolina is shrinking in the post- vs. pre-Lewandowski comparison.
The Reuters Polling Explorer potentially shows a similar trend. On June 19, the five-day rolling average showed Trump behind Clinton by 8.9%, 35.9% to 44.8%. But among the respondents, 50.3% were Democrats, and just 27.7% were Republicans! Since there should likely only have been a slight advantage in the number of Democrats polled – say, about 1% – this would suggest a corrected poll having Trump out front by nearly 13%.
Fast-forward to the June 21 five-day rolling average, and Clinton is ahead of Trump, in uncorrected form, by 9%. But 52% of those surveyed were Democrats, and 35% were Republicans, for a 17% advantage. This translates into a current bias-corrected lead of 7% for Trump, well below where he stood on Sunday night.
Those who live by flawed polling are doomed to die by flawed polling. The polling-based campaign pivot from rebel to establishment that took place on Monday may work out. Then again, we may look back in November and identify it as a tipping point of historic proportions in the wrong direction.
To return to the Brexit vote, the Leave side was also lucky. Had they waited another couple years to have the referendum, the rapidly growing pro-EU immigrant base – who will always overwhelmingly vote to remain – would have made victory impossible for the Leave side. Let this be a lesson for the immigration advocates across the West. Demographics is destiny, and it may not be the destiny you desire.