Any psephologists on the Dope? The opinion polls seemed to indicate a clear victory for Remain. So why wasn’t there a clear victory for Remain? On top of that there was the failure at the last election.
Or were the results within the margins of error?
Sometimes people don’t want to admit their bigotry to others, even a pollster. And the less emotional are less likely to vote.
The Leave vote highly correlates with higher age. Higher age highly correlates with voting.
As said above, some opinions are more easily admitted than others. You get the same under-reporting of FN voting in France.
Are Trump voters more likely to vote than others? If so, perhaps there is a fair amount of overlap between Trump voters and Leave voters and some of the same explanations apply.
Sigh.
But aren’t pollsters supposed to adjust for that?
I keep hearing this, but I certainly don’t see it. That poll of polls does have Remain winning by 2, but if you look down at the individual polls, they’re all over the place, with two of the six polls on the final day of polling indicating a Leave outcome. Why people were so confident in a Remain victory, given the polls, is beyond me…
Many of the polls were online polls–and it is well known that the elderly and less educated are less likely to be online.
I thought the polls were always too close to call, not “a clear victory for remain” at all.
Sigh, yourself. In every single poll I’ve seen, the vast majority of people on the exit side state “control of Muslim immigration” as the primary reason for their choice. Sure, there may be others reasons here and there, but claiming that the bigotry argument isn’t the primary motivator for the result is disingenuous.
Another example to add to my previous ones: The Economist’s final poll-tracker report on June 23rd. Tied 44-44.
The only person you can absolutely trust in a vote is the one who looks you in the eye and says, “I’m not going to vote for you”. Graham “Richo” Richardsonformer ALP NSW Right numbers man.
When? For the last three weeks before the election, they were indicating a tie or a narrow victory for Leave, apart from a very slight, within-the-margin-of-error swing toward Remain during the last three or four days.
They did indicate a clear victory for Remain for a couple of months prior to tightening, but this just shows that it’s a bad idea to get too complacent about polls that take place a month or more before the election. (This is a lesson that US voters should heed right now.)
I would like to point to the article below from Peter Kellner. Polling companies are always weighting for age, turnout etc. They also try to predict how the undecideds will go. I believe this Kellner article predicting how undecideds would split went viral among the great and the good. It became conventional wisdom in the weeks running up to the vote. It skewed pollsters opinion, bookies opinion and market opinion. Polling companies probably weighted most undecideds as Remainers.
I gave up my landline over a decade ago, so I haven’t been polled lately. Before that I had an attitude problem with the concept of polling and would intentionally never answer truthfully. How do they control for that, and is it a statistically significant concern?
This.
[QUOTE=PastTense]
Many of the polls were online polls–and it is well known that the elderly and less educated are less likely to be online.
[/QUOTE]
Actually, the online polls tended to show Leave doing better than other polls. The conventional wisdom was then that they were the ones that were wrong. That was usually based on the assumption that the Leave supporters were more passionate and therefore more willing to take part in those polls. That assumption was in turn coloured by the fact that it was widely thought that a not-dissimilar distortion was one of several reasons why the polls got last year’s general election so wrong.
[QUOTE=Fuzzy_wuzzy]
I believe this Kellner article predicting how undecideds would split went viral among the great and the good. It became conventional wisdom in the weeks running up to the vote. It skewed pollsters opinion, bookies opinion and market opinion. Polling companies probably weighted most undecideds as Remainers.
[/QUOTE]
It didn’t require that particular article to establish that as the conventional wisdom. That was well-established as a cliché of much of the commentary on the subject right from the outset, usually coupled with the related assumption that high turnout would favour Remain. Neither assumption was obviously implausible. It was just that they turned out to be not the case.
[QUOTE=TimeWinder]
In every single poll I’ve seen, the vast majority of people on the exit side state “control of Muslim immigration” as the primary reason for their choice. Sure, there may be others reasons here and there, but claiming that the bigotry argument isn’t the primary motivator for the result is disingenuous.
[/QUOTE]
You can’t have looked far. The most high-profile of the opinion polls conducted on the same day found that immigration was only the second most important reason why people voted Leave. Then there’s the point that there were plenty of xenophobic Leave voters who were at least as concerned about immigration from places like Poland. Brits and their politicians have been arguing about the EEC and the EU for over forty years, so views on the subject tend to be really rather varied. Even bigots could think of it as being about more than a single issue.
Preferring that over uncontrolled Muslim immigration is now defined as bigotry?
Do you think it’d be more or less bigoted to prefer “control of immigration in general” over uncontrolled immigration in general?
As the OP, please let the bigotry issue rest in the other thread.
Do we have no psephologists on the Dope?
Yeah it seems the case that the polls weren’t off by much but the betting sites and other poll watchers made bad guesses.
Could you provide a cite for that, please? I have seen some references that show immigration control was in some cases a primary concern of a plurality of Leave voters. I have never seen anything supporting your statement as you literally phrase it.
Part ot the problem is that referendum elections in Britain are very rare, so it was hard for pollsters to come up with a good “likely voter” model.