Why were the polls so wrong?

[Georgethesecond]

It seems there is some deep seated reason why the polls during the campaign say one thing, but that thing is different from the reality and different from the exit polls (which get it right). The same thing happened five years ago. But why?

Is is because people are reluctant to say they vote Tory because of the "nasty party" image?

Do people fool themselves that they will vote one way but when it actually comes down to it they don't? Some kind of human nature fundamental?

I'm not aware of people being afraid to say they vote Tory. IME, most Tory voters seem to be out and proud.

I think lots of people like to blend in with the crowd and go along with the flow. The media (i.e.BBC, Channel 4) etc pumps out a left wing progressive message non-stop i.e. caring Labour, nasty Tories etc and therefore people keep quiet about what they really think and go with what they think they are meant to say or feel - you can see this on Question Time for example where the audience all applaud progressives and boo UKIP or example. But come crunch time, they vote for what they believe, in the privacy of the polling booth where the BBC can't see them.

If Big Brother was watching, then they would do what they think they are meant to do and what Big Brother wants, but in private they do what they believe.

Because people realised late in the day that a) they were fearful of the SNP influence in Westminster if tied up with Labour, b) that they didn't have confidence in Miliband as a world leader, c) that in unsure times better the devil you know than that you don't, and d) that we're actually an inherently conservative country so again, in unsure times a lot of people will fall back to that stance. These are the reasons I think anyway.

I heard today that pollsters don't like to stick their necks out so if one thought it would end up like it has, because the others weren't predicting it they would stay safe and not predict it either.

Matthew Paris got it right, apparently.

So is it that people realise things late in the day then, not that the polls are just plain wrong?
They mess so much with the raw data to try to make it statistically representative, does this skew things in the wrong direction?
I just can't get my head around it.

Polls are only ever a sample of voting and a lot of people decide who they are voting for last minute.

They were right in Scotland even though even Nicola sturgeon didn't believe them!

I'm Scottish and didn't vote SNP but we all knew that would be the result up here.

The polls might be more accurate if they only questionned people who live in marginal or swing seats.

It's the so called "shy Tories" (or embarrassed Tories as I prefer to call them). Exactly the same happened in 1992 - people were too embarrassed to admit they were voting Tory because of it's image as the party of the selfish and the self-interested.

They keep quiet until they can vote in the privacy of the voting booth.

There is to be an investigation. Otherwise no one will waste money on them again!

I think a lot more people are floating voters. Even my old Mum who was dyed in the wool lifelong Labourite seemed to be wavering at one point.

I also suspect Dave Cameron was hypnotising the electorate by repeating his "stubby pencil" line!

It was a subliminal message that you are alone in that booth so you can go for an unpopular choice.

Most of the polls were conducted online by self-selecting participants. The grey Tory voters are less likely to be on line to indicate voting preference. There is a very vocal left wing online presence which does not lend itself to engagement with more right wing voters.

A mixture of things:
a lot of people were undecided until the last minute,
A certain demographic answers polls and others prefer to keep quiet (we saw this with the Scottish independence polls, the No voters were less vocal),
Some of the stats were badly done. At school they did this in a lesson and the kids worked out a slight Tory majority as their prediction. Simply by taking the Lib Dem seats and seeing who was second last time and they were pretty much spot on!

In every part of life that has human input there is going to be human error.

The thing is, I don't think it can just be the online issue - because it's almost an exact re-run of what happened in 1992. I think PasstheCremeEggs has the best analysis.

1992 polling:

"Almost every poll leading up to polling day predicted either a hung parliament, with Labour the largest party or a small Labour majority of around 19 to 23. Polls on the last few days before the country voted predicted a very slim Labour majority.[8]

With opinion polls at the end of the campaign showing Labour and the Conservatives neck and neck, the actual election result was a surprise to many in the media and in polling organisations. The apparent failure of the opinion polls to come close to predicting the actual result led to an inquiry by the Market Research Society. Following the election, most opinion polling companies changed their methodology in the belief that a 'Shy Tory Factor' affected the polling."

This was published before the election, so very prescient -
www.ncpolitics.uk/2015/05/shy-tory-factor-2015.html/

and it is a very, very good analysis if you are interested.

Yes it is really detailed and very interesting - thank you.

I don't know why they were wrong but I knew they were wrong.

'Why were the polls so wrong?'

Political correctness.

"It is likely that a significant section of the electorate told polling organisations what they believed they wanted to hear. Since people know that certain sentiments are stigmatised today, what they communicate to pollsters is not always their genuine view. A significant proportion of the opinion communicated to pollsters is influenced by a pressure to provide the ‘right responses’. Many people do not want to be thought of as ‘bigots’, or even just as old-fashioned, and in part they experience a demand for their opinions as a pressure to conform.
...
The pressure to conform and the fear of social isolation can lead to what the German social scientist Elisabeth Noelle-Neumann identified in 1974 as a ‘spiral of silence’. According to this theory, people’s assessment of the opinions held by the majority influences and modifies the way they express their own views. Some individuals feel anxious about expressing sentiments that differ from the consensus outlook, as expressed in the political and media realm, and it is thought that, ‘prompted by a “fear of social isolation”’, some are ‘less likely to express their own viewpoint when they believe their opinions and ideas are in the minority’. Typically, the fear of negative social sanctions influences the way people express attitudes about numerous morally charged ‘threats’, such as foreigners, crime or terrorism.

Of course, in virtually every social setting there is always an element of self-censorship. But in contemporary Britain, the all-too-understandable impulse to conform is continually reinforced by the message ‘You can’t say that!’. Regrettably, the only time these insecure voices feel confident enough to express their true opinions is in the secrecy of the polling booth. This is why, from time to time, a seemingly predictable election campaign can result in a surprising outcome.

Sadly, significant sections of the political and cultural establishment would rather that the people always felt that ‘We are not allowed to say that’. This is the 21st-century version of the old oligarchical ideal, ‘Know your place’. Which is why fighting for a culture of tolerance and open debate, and above all for freedom of speech, is so important today."

www.spiked-online.com/newsite/article/why-the-opinion-polls-got-it-so-wrong/16965#.VVPfdaLbKic

I never really believed the 34 versus 34 polls. But I was very suprised they were so wide of the mark. I wonder what odds bookies were giving on
the day before we voted?

Aren't the poll results presented with the undecideds removed? If so maybe most of the undecideds decided Tory when it came to.it?

Also outside of Scotland not many seats changed hands, and the swings were pretty small. So maybe the polls weren't that wrong, its just that the FPTP system magnified things (small swing -> tipping point of where other party leads)

I think there is some truth in the 'lazy Labour' reports. I spoke to a few friends about the results and many said they 'were Labour' but didn't bother voting as they 'are all the same now.' I think Tory voters are more likely to vote and unless Labour can get across to these apathetic voters then they have little chance.

Jellybeans. I am sure there is something in that. Only the exit poll will be trusted at the next election