Conservative activist Tom Waterhouse has written a fascinating pair of blog posts about why his party lost Enfield North at the general election, looking at both what he thinks the team got right and the reasons that Nick de Bois nonetheless lost.
Many of his points are relevant to Liberal Democrats too, such as the use of the air war and ground war to deliver complimentary, but different, messages:
We left the Conservative national campaign to fight the “air war”. Locally we used our leaflets and doorstep message to present people with the choice about who they want as their MP: Nick or Joan. Given Nick’s well-established reputation as an independent-minded MP who entered politics after spending 25 years running the small business he founded, we felt this would contrast strongly against a career politician who put party whips before residents (and had an appalling record on MPs’ expenses). Almost everything we did played into this “Nick vs Joan” frame.
Flicking through a range of local and national Conservative publications from the campaign around the country, it’s clear that this wasn’t unique to Enfield by any means.
Yet looking through Liberal Democrat ones, the national and local campaign messages were much more entwined. Was that higher level of integration better – or was it a problem because the two weren’t each playing to their own potential strengths? More detailed discussion on that is more suited to private locations, but it’s an important question to ask.
There’s also the question of the role of canvassing:
Election campaigns have three phases: voter identification, motivation and GOTV (get out the vote). A good campaign will begin by using the first phase to identify the people minded to vote for you. In the second phase you spend your time and resources ensuring those supporters are sufficiently encouraged to turn out and vote for you (hence all the leaflets, letters, door-knocking and phone calls). Finally, you need the logistics in place so that GOTV runs smoothly and all your supporters know when polling day is.
From the number of people you identify in the first phase, Conservatives work on the assumption that you can turn out roughly 75 – 80% of them. I think Labour work on about 65% or something. So if you’re trying to win in a seat where 45,000 people will vote (being spread among 4 or 5 candidates), you’re going to need at least 20,000 votes to have a good chance of winning. So we needed to identify about 25,000 – 26,500 supporters.
We started the short campaign in January with just 16,000.
Our efforts during the campaign found another 3,000 – but we should have been spending this time just motivating identified supporters instead of finding more.
However that rather prompts the question: in a high profile election such as the general election where just about everyone knows it is polling day, how useful is accumulating lists of supporters to knock up on the day* compared to accumulating lists of swing voters to persuade through the campaign?
Tracking the growth of your list of supporters without tracking what’s happening to your list of swing voters is the norm, but a questionable one too. Again, there’s more to discuss on that more privately.
But meanwhile (and before you go and read Tom’s posts in full I hope) one other thing that may interested Liberal Democrats in particular:
Vote Source is the worst election software I have ever worked with.
It looks like leaving it so late was indeed a mistake.
UPDATE: For a Conservative perspective also on how to win a Parliamentary seat, see Gavin Barwell’s book.
* There are times when I so hope Americans are reading this site.