Populus cross party splits	CON 2010 %	LAB 2010 %	LD 2010 %
CON by-election	46	2	3
LAB by-election	12	89	31
LD by-election	34	3	55

Or was the party churn more complicated than that?

In the aftermath of the OES result a widespread assumption has developed that all the Tory votes lost last night went to the Lib Dems. This looks a simple straightforward answer and apparently explains everything.

But does it? For the Populus poll of the constituency, taken last week and published last Sunday, seems to point to a more complex range of behaviour which I have reproduced in the table above. The columns represents how respondents to the survey said how they voted at the general election and what they planned to do in the by election.

Clearly this is not an accurate representation of the voting dynamics yesterday but given the broad overall accuracy of the poll it does provide a pointer to what might have happened. In any case it’s the best data that we have got.

So the first column shows that 46% of 2010 Tory voters were planning to do so again; 12% were planning to vote Labour while 34% said they were going Lib Dem. The balance went to other parties.

Clearly this is all guestimates but broadly about two thirds of the Tory losses went to the yellows. That’s a lot but is not everything. As can be seen as well the Tories the losses were party off-set by small gains from the other two main parties.

The bulk of the Lib Dem losses went to Labour but not, by any means all. And the yellows benefited by a small group of Labour general election voters switching.

Hopefully there might be further research on this.

Mike Smithson