From Political Betting –
Some detailed analysis on ICM polling stats shows that the weighting may be wrong.
If we take the weighted 2005 figures in the ICM table: Con 150 Lab 173 LD 92, and bung in the Others 8.0% from 2005 it gives 2005 vote shares of Con 33.3 Lab 38.4 LD 20.4%. That is overstating Lab by 2.2 points at the expense of the LDs when compared to the actual results. The transfers suggest a small net Conservative gain compared to 2005 and a small net Labour loss, but weighted to the actual result the situation is reversed. According to this table Labour’s lead among serial voters has fallen from around 5% in 2005 to 4.8% now. But Labour’s lead in 2005 was only 3%.
See the thread number 87. Kevin L.
If true the Brown bounce is puny indeed, and nothing like the 10% YouGov unweighted result given in the weekend press.
On top of that, Brown faces the possible collapse of Labour’s vote in Scotland, if the SNP continues to outmanoevre Labour. That could cost him up to 40 MPs especially if Conservatives were to vote tactically.
No one knows how much impact the BNP will have as they open up in 500 Constituencies this year. ‘Others’ are up in all polls since 2005 from 7% to at least 11%. That too could upset Brown’s calculations.
It seems most unlikely Brown would risk an election under these circumstances. He will try to ram the Constitution through Parliament as his key number one objective. There too he could be in trouble as 40 or more Labour MPs are saying they are willing to rebel.