Writing on the LSE’s British Politics and Policy blog, Steven Ayres looks at where UKIP’s vote came from in the 2014 local elections, and what this might mean in next year’s general election.
It’s an interesting piece, but it brings me back a thought I’ve had recently with regard to UKIP’s chances next years and whether analysis is factoring in the effects of differential turnout and the motivation to vote of UKIP voters.
One assumption that we tend to make in projecting local election results forward to a general election is that the different turnout at the two elections can be ignored as a factor. We know that general elections have larger turnouts that local elections, but we assume that the voters who turn out at local elections are a proportional sample of the voters who’ll turn out at the general election. As an example, imagine a constituency with four parties where 20,000 people will vote at a general election, but only 10,000 of those will vote at the preceding local election. At that local election, we get the following result:
The tendency is to assume that the percentages of votes received in that election represent the share of opinion amongst the wider electorate and that thus amongst the general election electorate of 20,000 Party A will have 8,000 voters (40%), Party B will have 6000 voters (30%) etc. We assume that the 50% who will turn out for the local election is drawn equally from the parties – a Party A voter and a Party D general election voter both have a 50% probability of voting in the local election.
The reason I’m writing this in relation to UKIP votes is that it seems to me that projections of what UKIP might do in 2015 assume that their voters turned out at the same rate in 2013 and 2014 as did supporters of the other parties. I’d need to go back to some of the polling from before May’s elections to confirm this, but my recollection is that it showed UKIP supporters were much more likely to intend to vote than those of other parties.
To go back to the example, we assume it represents voter intention amongst the wider population, but what if each party’s supporters had a wildly different likelihood to vote at local elections? For example, if Party A’s supporters were 80% likely to vote at local elections, while Party D’s were only 20% likely, both could actually have 5,000 supporters amongst the general population. Extrapolating a general election result from a local one when the parties have had differing levels of turnout is not going to produce accurate forecasts.
To return to the UKIP question, one thing I have noticed is that while they have had relatively high percentages in some elections, they’ve not gone beyond around x% of the total electorate. By my calculation, the share of the electorate UKIP have got in high-profile elections recently is:
My suspicion – and it’d take a lot of time with a lot of polling and election data and a stats package to explore that in more depth – is that UKIP’s peak support within the electorate is somewhere around 15%, but they are much more likely to turn out at all elections than voters for the other parties are. Thus, in elections with a lower turnout, their higher propensity to vote and our tendency to assume that those elections are a mirror of voters as a whole makes them look more of a threat in the general election than they will actually be. However, in a general election with turnouts of over 60%, it’s very hard for a party that can’t get more than 20% of the electorate to win a seat. That would require winning a seat with 33% of the vote or less, and while that’s possible, it’s very rare.
UKIP’s dilemma for next year is that their best chance of winning seats comes if a) there’s a low overall turnout in a seat, allowing the higher motivation of their voters to be a factor and b) there are multiple parties competing for a seat, thus making it possible to win with a low share of the vote. However, seats with strong competition between multiple parties are unlikely to have a low turnout because of the amount of campaigning that’s done in them. It seems that UKIP are very good at getting out the vote, but they’ll need to broaden the number of people willing to vote for them to have a serious chance of winning a Westminster seat.