Why the Tories Won

zagfles

chewmylegoff wrote: »

Most of the Tory gains were from the lib dems where there was a significant swing from lib dem to Tory.

If e.g. In twickenham, the people abandoning lib dems had all voted green, or labour then Vince cable would have held his seat, but they principally voted Tory.

Anti Tory voters don't tend to vote Tory.

Twickenham wasn't typical. In loads of SW seats the Tory vote dropped yet they took the seat off the LibDems. Overall the Tory vote only increased 0.8%. The LD vote dropped 15.2%.

The idea that significant numbers switched from the LDs to the Tories is ridiculous. Any Tory leaning LD voter last time should be delighted they went into coalition with the Tories.

The LDs lost votes from the "anti-Tory" types. At the same time other swings were in operation, such as to UKIP from all parties.

In England, Labour's vote went up 3.6% despite the rise of UKIP. The Tory vote only went up 1.4%. Given that UKIP probably took as many votes off Labour and the Tories, and probably not many off the LibDems, excluding the UKIP swing Labour's vote would probably have risen about 8% in England.

So I reckon main changes in votes in England & Wales were:

LD to Labour
LD to Green
Labour to UKIP
Tory to UKIP
Possibly Labour to Tory

Be interesting to see a proper analysis of vote changes, though we probably can't trust pollsters any more!

zagfles

onlyroz wrote: »

No mechanism at all. In 2001 I voted twice.

In a national election? Isn't that illegal?

zagfles

bigheadxx wrote: »

There are two electoral regisrers. One is for parliamentary elections the other for council/EU/Assembly votes.

I only get the annual registration details for one and I get polling cards for both local and national elections...

onlyroz

zagfles wrote: »

In a national election? Isn't that illegal?

I'm sure it is. But like you said, nobody ever checks.

antrobus

Thrugelmir wrote: »

Perhaps 1,200 people isn't enough to provide a good prediction.

Whereas the exit poll was some 22,000 in key constituenices.

Basically, yes. The sample size you need to be 99% certain that your results are right with +/-1% is about 17,000. Which was about the size of the exit poll in 2010 - they did a bigger sample this year in order to get Scotland right, apparently.

With a sample size of 1,000 you can only be 95% certain that the result is correct within +/-3%. So you wouldn't expect an opinion poll to be that accurate, but (1) the people who do them like to the think they have a 'methodology' that removes a lot of the random error, and (2) if you carried out a lot of 1,000 sample size polls at about the same time, you'd expect them to average out to the correct result.

So, when you have a situation with 8 (?) different polling companies and they're all flippin wrong about the same thing, it kinda implies that there is something wrong with the methodology. They were not that wrong in 2010. There is something new out there, that they don't know that they don't know.

antrobus

zagfles wrote: »

Twickenham wasn't typical. In loads of SW seats the Tory vote dropped yet they took the seat off the LibDems. ....

Are you sure about that?

The Con vote increased in;

Taunton Deane, Yeovil, Bath, Cheltenham, Chippenham, North Cornwall, and Wells

(I only checked those 7. Then I stopped. That must be a big enough sample.)

zagfles wrote: »

..Overall the Tory vote only increased 0.8%. ....

Be interesting to see a proper analysis of vote changes, though we probably can't trust pollsters any more!

I don't think any of that really matters. What matters is that the Conservatives got the the right number of votes in the right places.

onlyroz

So why were the pre-election polls wrong but the exit poll right? Was the methodology different?

wotsthat

onlyroz wrote: »

So why were the pre-election polls wrong but the exit poll right? Was the methodology different?

What people say they will do is often different to what they actually do.

michaels

onlyroz wrote: »

So why were the pre-election polls wrong but the exit poll right? Was the methodology different?

The exit polls were anonymous and only questioned those who had actually voted. Polls were predicting a 70% turnout, in reality only 67% voted.

onlyroz

wotsthat wrote: »

What people say they will do is often different to what they actually do.

But if they lied in the first poll why did they tell the truth in the exit poll?