Wednesday, 31 October 2007

The Polls explained

There was no PMQs today because the house has been prorogued. A fancy word for MPs saying we have been on holiday for a few months, come back for a couple of weeks and need a another short break before Parliament's official state opening in a week's time. Good work if you can get it. Why Parliament can not just officially open when they all come back after summer recess, I don't know. This does give me a chance, however, to use my Wednesday afternoon PMQs battle time for a quick analysis of today's poll...

The new ICM poll out today again muddies the water with trying to understand what is going on with the three political parties and their support. There are two things that are becoming clear:

Firstly: The Tories have regained at least some of their pre-Brown position and are "in the game".

Secondly: The Lib Dems will make a recovery of some degree in the post-Ming era.

While the headlines are that the Tories are five points ahead of Labour in the latest poll, the real story is not so clear. The big mistake (either deliberate or otherwise) that some in the media make is that they compare different pollsters (e.g. a ICM with a YouGov) and that they compare the difference between the parties. This can give a misleading view on what is happening. There are two key factors to take in to account when looking a the polls.

Firstly you should only compare like for like pollsters. Each pollster has it's way of collating its data, hence they are all giving slightly different pictures. To say that Labour have gone up 2% in the polls over night because one day in a YouGov poll they have 38% (for example) and then the next day in an ICM poll they have 40% is simply not true and very misleading.

Secondly, while what most people want to know is how far ahead one party is from another, it is not always (if ever) the best way to view a poll. It is far more important to view what a party's share is in the current poll compared to what it was in the previous poll.

I will put this in to practice for today's ICM poll. Yes the Tories are now 5% ahead of Labour, but it also shows that they have lost 3% of their vote - to the Lib Dems. Labour are also down - but only 1% - again it looks like this has gone to the Lib Dems, who are up 4% to 18%. The previous poll showed the Tories were 7% (they are down 2% in the current poll) ahead of Labour - and while this is of course an important piece of information, in the longer term the drop of 3% in the Tory vote to the Lib Dems is far more critical.

While this poll is a good spin for the Tories, reading between the lines they have dropped in the polls and the votes are going to the Lib Dems at a ratio of 3:1 compared with Labour. This is exactly what the Tories want to avoid. The Lib Dem rise in the polls is likely to continue - possibly up to another 5-6%. This has to come from either Labour or the Tories - whoever it comes to will be in the stronger position. Today's poll suggest that it more likely to go to Labour. However, as I said in yesterday's Politics Decoded column, it is still far too early to say very much at all for sure...

To put it another way, even though the Tories have surged from the low 30s to the low 40s in the past six weeks, the Labour vote has still held fairly strong. So yes Labour were about 8% ahead of the Tories a few weeks ago and are now about 5% behind - they haven't actually lost much of their vote at all. In fact, in 2005 Labour polled 35% in the general election and won by 66 seats. Given the efficient spread of the Labour vote, it is far more important to look at what Labour are polling rather than how far ahead or behind they are the Tories they are.

No comments: