Do the mid-conference season polls mean anything?
The comments I've received on the post-conference Labour polls contain a lot of encouraged Labour supporters thinking that the mood may have changed, sometimes seemingly quite bewildered about why people are quite so dismissive of the polls showing a smaller Labour lead. I've been phoned up by some journalists this morning asking what weight we should give polls showing Labour catching up with the Tories. The answer is - for now - not much at all.
It is, of course, possible that Labour caught the public mood at their conference and things have indeed changed, but looking at polling around Labour conferences in the past, it seems more likely that it is a publicity boost. In the past when party conferences do have any meaningful effect, it has been extremely short lived.
The graph below shows Labour's ratings from YouGov (since they tend to do more polls around conference than other pollsters) at each party conference since 2002. The first column for each year is Labour's average rating from the start of August that year until the beginning of conference, the red column is the YouGov polling done immediately after Labour's conference (in a few cases the average, when YouGov did more than one poll in the few days afterwards). The final two columns are the Labour score in YouGov's monthly polls at the end of October and November.
Labour enjoyed boosts in the polls from their conference in 2005, 2006 and 2007, and in every case they fell back down within a couple of months - in 2006 and 2007, they went back down almost immediately. Only looking at 2003 does the conference season appear to have given Labour a long term boost - but that is actually the effect of the Conservative party collapsing into internal leadership ructions after their own conference.
Of course, it is hard to draw rules as every conference is in its own way its own special case: in 2007 there was the non-election announcement, 2006 was Blair's valedictory conference. It's possible that this one will be different and will have a lasting effect, but looking at recent conferences, there aren't many encouraging precedents. No doubt there will be plenty of polls next weekend - if nothing else we are due YouGov's monthly poll for the Telegraph. If Labour have still narrowed the Conservative lead then, then we can start talking about the game having changed. If not, then only when any Conservative boost from that conference has died away will we actually know if the conference season has made a difference - but it would be quite an unusual conference season that did.
