It looks like the polls were off by 8% because of this. So what's the percentage here? Whatever it is, it's not zero.The polls immediately before the [Brexit] election leaned “Remain” or showed a 44%-to-44% tie, with 12% undecided. The referendum’s result was 52%-to-48% “Leave.” The idea that the undecideds broke 2-to-1 for “Leave”—against the status quo—is not tenable. That said, the pollsters are not at fault. The pollsters cannot conduct an accurate poll because voters are unwilling to tell pollsters how they intend to vote. It is that simple.Why are the voters doing this?The UK now has its own version of America’s Bradley/Wilder effect: people—decent, ordinary, hard-working, law-abiding people—are afraid of abuse and reprisals if they reveal their true political preferences. People are tactically deceiving the pollsters because the pattern-and-practice of British politics is now:(i) to demonize opponents;(ii) to invade and shut down their peaceful political meetings and conferences (and then to “justify” speech suppression as “free speech” or other lawful protest);(iii) to threaten and physically assault their party leaders and members;(iv) to destroy their political posters and/or to paint over them (and then to call their vandalism “art”); and,
(via)
I don't know how, but I read the introduction as "the polls are a dogs erection..." in my RSS reader.
ReplyDeleteI didn't know where you were going with it, but it seemed as reasonable as anything else this year.
Would you put a Trump bumper sticker on your car? I would not. But the Bernie supporters don't feel threatened. I see lots of them.
ReplyDeleteThe outcome will be interesting.
The 'polls' are simply meant to tell the masses what to think before they go to the polls.
ReplyDeleteNo need to vote, we already know the right result.
Diversity means everyone MUST think alike about 'diversity'.
I have a Trump sticker on my car. But that car is a rust bucket of no real value. and so far no problems.
ReplyDeleteWe do not have a trump sticker on the "nice" cars.
But yes...... I suspect that in most cases the "polls" are either total fabrications .... or they have been "adjusted to compensate for perceived errors." If it's good enough for AGW then it must be OK for political data.