Check Your Boundaries
John Lanchester · How to Vote Tactically
With everyone and his Labour dog calling for tactical voting, the always-good-value Tim Harford made an important point on the Today programme this morning. He says that about 10 per cent of voters report having voted tactically in previous elections. That might not sound like much but in most constituencies it is usually more than enough to determine the difference between coming first and coming third.
The important point, though, was the one he made next: that if you are considering voting tactically, it is important to check the facts about the position in your constituency. Leaflets shoved through the door make all sorts of claims, and so do canvassers on the doorstep, but it crucial to check the facts for yourself, because constituency boundaries have changed all over the country. This means that simply looking up the results for 2005 aren’t enough, you need to check what the results would be if the constituencies had been tweaked to reflect their current boundaries. (It is because of this that there are oddities such as lists which show seats that are currently Tory as Tory target seats – i.e., seats they’re trying to win from Labour. The point in these cases is that post-tweaking, the relevant constituencies would be Labour seats on past voting patterns. So even though the seats are currently Tory, in this election they would count as wins from Labour, because if the boundary changes had been in effect in 2005 they would have been Labour at the last election. I’m glad that’s clear.)
What we badly need is a non-partisan website which takes the 2005 voting and plugs in the 2010 boundaries to tell you what the voting pattern would have been with current boundaries. That is very important information for anyone planning to vote tactically. If only that site existed... but wait! The BBC have done exactly that at their election website. In fact they go a step further and give the 2005-tweaked to-2010 data as their default setting when you look up your Parliamentary constituency. Any tactical voters out there, don’t neglect to check.
Comments
This election is actually 650 mini-elections which will be aggregated, but as this post shows, it's not 650 mini-elections in which we can add people's votes up and say this is what we thought. Rather, the elections are a reflection of the estimation of the best way to achieve a certain outcome; envy the person who doesn't care and can cast a vote unencumbered by any such calculations.
I had an argument with a mate a few days ago about what to do in Brighton Pavillion; as a Green, I was trying to persude him to vote for us, not least because it's the best anti-Tory strategy, but he was nervous. We never actually discussed who might have better policies, or be a better MP. Our dispute was over what the current state of play was and who was currently ahead between the Greens and Labour. I can't imagine that this is what they had in mind as they gathered at Peterloo.
This is matched by the fact that marginal seat polls tend to be polls of a sample of normal size undertaken in marginals. But to get a snapshot of a particular seat, I'm told you need a sample of around 500 to capture the specifics of a seat and measure deviation from national polls. No-one is polling like this, except the parties, who aren't revealing them, presumably because the main beneficiaires are the Lib Dems, who can't afford to poll.
All of which adds up to say no-one has a clue what will happen, because no-one has the data nor the model to predict how people will perceive the race to be and how many will vote accordingly, and how accurate those assessment will be, not to mention what the 2005 stayways will do. If anyone nails this at 10pm on Thursday, they've either been incredibly lucky, or have found the philosopher's stone.
I'm not sure about stay-aways. In what's now my constituency there was a massive swing from Labour to the Lib Dems in 2005, essentially on anti-Iraq War and anti-Blair grounds. When I was writing my profile of the constituency I assumed that turnout would also have dropped as formerly loyal Labour voters stayed at home, but it actually rose from 51% to 55%. I think the 2001 election had a particularly low turnout.
As a young man Cameron looked out on the social carnage of pit closures and mass unemployment, looked at Margaret Thatcher's government and thought, these are my people. When all the debating is done, that is really all I need to know.
Even if it's not news at the LRB, this is a very good point.