The moment for decision is almost here. With nearly all of the relevant information now available, this is when tactical voting moves on from endless speculation and debate to a politically meaningful choice. And since there isn’t a single authoritative source of information on how to reverse-engineer a broken electoral system, it is a choice you’ll need to make for yourself. Here’s how to do it right.
First, it is now probably safe to visit your favourite tactical voting aggregator site to see what the postcode-based advisory websites are saying. I’m not posting a link here as it is my firm conviction that if you aren’t already aware of these things, you’re probably best off not bothering. Because what the postcode sites will tell you is one of three things:
You may be in a safe seat where tactical voting will not make any difference. In which case, vote for your favourite party, as it has a slight benefit in terms of the short money it will receive.
Or it is really obvious which party you should tactically vote for, because everyone agrees on who’s in second place. In which case you probably already knew that.
Or tactical voting could, in principle, make a difference, but it is not obvious which party you should vote for.
In the third case, it really isn’t a good idea to look at the various websites and go with whatever a majority of them say. The fact that they differ at all shows that there is normal statistical error in the forecasting process, and it is likely that as a well-informed voter of reasonable intelligence and diligence, you can outperform them by making use of information that isn’t assembled in a numerical dataset.
More importantly than that, tactical voting is, fundamentally, a coordination problem, not a forecasting problem. What the tactical voters want to do is maximise the chance that they all pick the same candidate. Knowing which candidate is currently likely to be in second place is an important piece of information to solve that problem, but it’s not the be-all and end-all, particularly in cases (such as Kensington), where the margin of forecasting error is large compared to the gaps between parties.
So you may use the postcode sites as one piece of information. But you must also consider these things:
1. You need a very good reason to vote against a sitting non-Tory MP. The fact that someone is the incumbent MP means that there is proof they can win; you don’t have to worry so much about the willingness of other people to vote tactically for them. And they are easier to coordinate on because they have greater name recognition.
2. If you have really good local knowledge, use it. If there’s a special factor in your constituency which means that polling data or extrapolations from national swings might not be reliable, use that to decide close calls. This might be a smart thing to do in the half-dozen London constituencies where the Jewish electorate is large compared to the usual margin of victory. If there are a lot of other people who will not vote Labour no matter what, the tactical thing to do is join them.
3. Pay attention to the ground game. The number of votes that are likely to be swayed by tactical voting is small. It is likely to be unusually large in this election but, according to surveys, we are still talking about just under 20% of Labour voters and 36% of Lib Dem voters. And many of these people will be doing so in constituencies where it is obvious what to do; in close and difficult races, the “effective tactical margin” will be reduced by people voting tactically the wrong way. So the actual effect on the result is going to be similar in magnitude to the roughly 4% effect from a normal leaflet campaign, and even the 0.5-1% advantage from on-the-day canvassing could make a difference. If you think one party has been working particularly hard in your area, take that into account.
4. YouGov MRP is first among equals. The model, which published its final forecast last night, is not necessarily the best, but it’s the one that most people pay attention to. Your fellow tactical voters will be looking at it, so you should look at it too.
5. And finally, recognise that this is a losing hand. When you’re thinking seriously about tactical voting, you’re holding bad cards and hoping for a stroke of luck. The likelihood is that the party that you want to defeat will get in; you might not even have particularly good odds of making the situation better rather than worse yourself.
You’re taking a gamble, and it’s a decision you’re going to have to live with. For me, this is the real reason not to look at voter guides or postcode websites. Figuring it out for yourself, rather than outsourcing your decision to someone else, will be easier to live with if it all goes wrong. Good luck.
• Dan Davies is a former Bank of England economist and investment banking analyst
