I did a little piece on the General election recently for Betfair. Underlying it was some work I’d looked at recently on turnout for elections.
When the election was called the markets shifted significantly anticipating a Conservative landslide. I immediately wondered if there would be a correlation between certainty and turnout. But first I had to work out what a ‘normal’ turnout would look like.
I looked back at elections since the second world and munched through some data. The first thing to note is that turnout is falling. In 1950 nearly 84% of people voted and that had fallen to under 60% by 2001. It’s climbed a little since then but the trend is definitely down.
The recent turning point came with the re-election of the Blair government in 2001. They were expected to get a landslide victory and did, 246 seats more than the Conservatives. Surefire wins by big margins seem to bring out voter apathy. What’s the point in voting when you know the result?
My immediate thought when the election was announced, was OMG not again! I welcomed the fixed term parliament as a way to avoid political contrived votes and more elections. Yet here we were, midterm, with a new election. Having been dragged back to the polls every five minutes I was a bit pissed off. My initial thought was ‘none of the above’. I still think I should start the ‘none of the above’ party as I’m sure it would get lots of votes.
You can pretty much split the era of decent turnout, to apathetic quite easily. Up to 1997 turnout was generally pretty good. An average of 73.50% with a standard deviation of 6.50%. From 2001 to present it slumps to 63% with a standard deviation of 3%.
Scotland had a particularly low turnout, till the surge in support for the SNP lifted the turnout from just 58% in 2001 to 71% in 2015. So it does appear that having something to really vote for really drives turnout. Contradicting that though is the low turnout in 2015 when things looked ‘too close to call’. They weren’t.
So I conclude that people being fed up with politics is probably the only consistent theme. Getting the voters out may be key in this election. A certain re-election seems to drop turnout.
Having researched the historical data, I mapped the historic’s onto the current and came up with numbers. I also snapshotted the odds at the time I did it, to see how they moved in the interim.
Assuming a millennium style turnout, then this brings odds of a 60-65% turnout in at 1.76. When I looked initially, this was 3.40 but has since contracted to 2.62. Here are the full list and the moves since I took a snapshot.
If you plug in pre-millennium figures the numbers, obviously, push to the far right, 80%+ comes in at odds 6.20 and that looks completely unrealistic.
Given that most don’t expect a close election, my bias it towards to lower end of the above graphic. The only thing that makes this look a weaker assumption is the shortening of ‘no overall majority’. This would imply a slighter higher turnout. But I guess most of the electorate don’t look at Betfair markets. I’d still prefer to be on the low turnout side as voting fatigue hits home.
Thursday night is going to be interesting.