It’s not giving any secrets away that this May’s elections will be a difficult fight in many parts of the country. One skill people will have to learn very quickly is fighting “defensive” campaigns.
That isn’t a term I particularly like – the best way of tackling a campaign when you are on the defensive is to remain positive, give people reasons to vote for us and look for ways to put your opponent on the back foot.
Another important thing to understand is what is happening to voting intentions across your electorate. It’s in more difficult elections that your ability to analyse and understand what your canvass data is telling you really starts to matter.
There are many techniques that can be used to analyse your canvass data which are covered in the two Campaign Manuals and those chapters should be essential reading for campaigners planning next May’s campaigns. One often overlooked tool is the switch analysis function in EARS.
Basically it allows you to compare two sets of canvass data so you can see how people who were canvassed a particular way at a previous election are recanvassing in the current election. In an election where you are worried about losing support this can be vital as it can give you hard data about the degree to which you are losing support and where it is turning to.
Like all types of canvass data analysis there are certain basics that need to be in place for it to work:
- A reasonably high level of contact in both a previous election and the current campaign. As you are looking for people who have been canvassed on both occasions you need a reasonable contact rate to get enough overlap to be worthwhile.
- A consistent approach to canvassing. I first started using switch analysis in wards where the canvassing was, by and large, done by the same people each year. Whilst some people might be over optimistic or over pessimistic that shouldn’t have any effect on the changes from year to year. If people previously canvassed as soft Labour by an over-optimistic canvasser were being recorded as probable or definite then it was indicative of something significant.
- To draw conclusions across a ward you need a representative sample across the whole area. That doesn’t though mean that conclusions can’t be drawn about a particular area but they may not hold true for the whole ward.
Understanding the switch analysis print out is key to the whole process. The printouts look complex (and to be honest are not a model of simplicity!), but ALDC has produced a simple PDF guide on how to interpret the information. See our guide below. Note though that comparisons of opposition voting intention need to be treated with a fair amount of caution as they will often be based on a very small sample.
I’ve found that the most useful figure to look at is the retention rate of Lib Dem voters. Using the percentage figures you can calculate that 93.5% of Lib Dem voters from last time are recanvassing as definite or probable. You can also take a narrower measure of definite recanvassing as definite.
Obviously this figure can’t ever exceed 100% and will very rarely approach that. This is about as high as I’ve seen this indicator reach and, unsurprisingly this was a campaign which we won!
Where a switch analysis can be very useful, particularly in a “defensive” campaign is in tracking slippage in our support. You may find that the feel on the doorstep is quite good but a switch analysis can pick up more subtle changes such as a drift from definites to probables. That sort of movement is the danger sign that you need to be very alert for, because that is the sort of slide in support that can be reversed if you pick it up at an early stage.
For further reading see ALDC’s Winning Local Elections Volume 1 Chapter 9 and Volume 2 Chapter 10.
Article by Hywel Morgan who was office manager for Gordon Birtwistle MP and is currently a candidate for the Calder ward of Calderdale Council.