Are the betting markets a good guide to by-election results?

With two by-elections coming up on Super Thursday, let's look back at how the betting markets have fared in forecasting previous contests.

There have been twenty two parliamentary by-elections in Great Britain since the 2015 general election. A couple of them didn't have betting markets available due the nature of the contests, but that still leaves us with an interesting selection of results. In summary, of the other twenty by-elections, the betting market favourite won on fifteen occasions (75%).

Winner OTD% = percentage chance given to the winner by the betting markets at 8am on polling day.

What are the chances of a shock result on Thursday? Well, if Labour fail to win Wakefield, that really would be a huge upset. Despite Labour having failed to make a by-election gan since Corby ten years ago, the odds see them as 98% likely to triumph. The two polls of the constituency that have been published put them 20 and 23 points ahead of the Tories and it's hard to think of anything that's happened since that could shift the dial by anything like enough. The most surprising result of recent times was probably George Galloway's win in Bradford West in 2012 - you could have backed him at 100/1 at one stage, although by polling day his odds had contracted to about 6/1.

Tiverton & Honiton looks much more interesting. There haven't been any opinion polls published here unless you count the "internal survey" reported by the Liberal Democrats which reported that the Tories were ahead 46%-42%. The markets are understandably very sceptical about such information - parties only release this sort of stuff when it suits their purposes and a narrow Tory lead is exactly what the Liberal Democrats would like to broadcast to the electorate, as it might serve to motivate any floating voters or remaining Labour supporters to lend them their votes. Hence, the Lib Dems remain favourites to overturn a 24,239 majority, although the odds have started to tighten up a little over the last few days, with their win chance dropping from a high of 90% down to 76%.

There isn't much encouragement from our historical record for the Conservatives here. Of the five occasions when the favourite lost a by-election, four of those were when the Tories failed in contests they were expected to win and the other was a Brexit Party flop in Peterborough. There are no recent examples of the Liberal Democrats failing to win by-elections in which they were competitive, so it'll be a massive disappointment for them if they can't pull this one off. On the face of it, this looks a more straightforward task for them than their victory in North Shropshire, although it may be that the Tories have picked a better, local candidate this time.