Last September, the politics team at Smarkets decided to start offering a market on one local by-election every week. We’ve had a look back at the results to see if we could learn anything about whether the betting markets provided any useful information about the outcomes.

On average, there are about five local by-elections across the UK every Thursday. Unsurprisingly, we’re usually going to pick whichever one we think looks the most competitive. (One other factor we sometimes consider is how much information we can find on social media about the election and the candidates involved). 

Anyway, let’s see the general patterns in all local by-elections recently:

Almost half of all local by-elections resulted in a seat changing hands! That compares to around 12% of constituencies that changed hands at the 2019 general election. So, it shouldn’t be a surprise that there were a lot of gains in the wards we picked, given that we were already trying to find ones that were going to be close.

Here’s what happened in the thirteen contests we have had a market on so far.

Favourite LTP is the last traded price on the favourite where a different candidate won

So, the favourite won in nine out of the thirteen votes. The biggest “shock” was the most recent by-election in Cavendish. It looked like it should have been an easy Labour hold if one knew nothing other than the result last time and the national opinion polls. The betting did give some clue as to what might happen, with the Liberal Democrats halving in price before going on to win by 39 votes. In fact, the biggest betting move came on a loser in the opening contest in Hetton. The Lib Dems first traded price was 40 and they were backed all the way down to 4.6, but ended up missing out by just 27 votes.

Matthew Shaddick

Matthew runs political markets for the Smarkets betting exchange. He used to be at Ladbrokes