What You Can Get Away With » Elections

Officials count ballot papers in WitneyWriting 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:

  • Party A: 4000 votes
  • Party B: 3000 votes
  • Party C: 2000 votes
  • Party D: 1000 votes
  • 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:

  • Eastleigh by-election: 14.7%
  • Newark by-election: 13.7%
  • 2014 Euro election (East of England): 12.4%
  • 2014 Euro elections (East Midlands): 10.9%
  • 2014 Euro elections (national share): 9.4%
  • Rotherham by-election: 7.3%
  • Wythenshawe and Sale East by-election: 6.7%
  • South Shields by-election: 6.4%
  • 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.

    , ,

    These are the results for Colchester in the Essex County Council elections. Full results can be found on the Essex County Council website.

    Abbey: Lib Dem (Margaret Fisher) 1273, UKIP 786, Labour 519, Green 467, Conservative 395. LD hold
    Constable: Conservative (Anne Brown) 2075, UKIP 1471, Labour 504, Green 387, Lib Dem (Carolyn Catney) 303. Con hold.
    Drury: Conservative (Sue Lissimore) 1957, Lib Dem (Nick Cope) 1127, UKIP 951, Labour 527, Green 314. Con hold.
    Maypole: Labour (Dave Harris) 1665, Lib Dem (Lyn Barton) 933, UKIP 573, Conservative 475, Green 143. Labour gain from Lib Dem.
    Mersea and Tiptree: Conservative (John Jowers) 1913, UKIP 1134, Labour 629, Green 216, Lib Dem (Gill Collings) 181. Con hold.
    Mile End and Highwoods: Lib Dem (Anne Turrell) 1417, Conservative 888, UKIP 725, Labour 408, Green 180. Lib Dem hold.
    Parsons Heath and East Gates: Lib Dem (Theresa Higgins) 1259, UKIP 809, Conservative 609, Labour 489, Green 192. Lib Dem hold.
    Stanway and Pyefleet: Conservative (Kevin Bentley) 1723, UKIP 929, Lib Dem (Jessica Scott-Boutell) 829, Labour 491, Green 247. Conservative hold.
    Wivenhoe St Andrew: Labour (Julie Young) 1895, UKIP 599, Conservative 562, Lib Dem (Shaun Boughton) 383, Green 248. Labour hold.

    I don’t have the full result yet, but I have been informed that, along with Janet Nunn and Nigel Quinton, I’ve been elected as one of the members of the East of England Liberal Democrat Regional Policy Committee. Thank you to everyone who voted for me – I look forward to having the chance to get on with the things I wrote about here.

    There are other places on the committee still to be filled, and there are three places available for Liberal Democrat councillors in the region, as well as two from the region’s parliamentarians and two from the regional executive. If you are a councillor and interested in one of those positions, then you should have received details from the regional secretary about how to put yourself forward as a candidate.

    I’ll hopefully be providing some updates and feedback here during my time on the RPC, and I’m always open to hear your thoughts and comments.

    ,

    There’s an interesting article on Buzzfeed about American right-wing bloggers and their determination to prove President Obama was somehow unfit or unqualified for office.

    (Spoiler: they failed)

    It’s interesting because it’s an examination because even though it doesn’t use the word, it’s an examination of political groupthink. We have a group – however informally constituted – who have decided on a plan of action and then continue to press on with that course of action despite evidence that it isn’t working. The article goes through a lot of the ideas that this group were pushing, having committed themselves to the belief that Obama was a dangerous radical and that all they needed was the single piece of proof that would bring him down. (In that light, the belief in, and desperate searching for, the seemingly mythical video in which Michelle Obama used the word ‘whitey’ becomes something like a grail quest)

    The consensus that soon emerged on the right was that if Americans were fully aware of Obama’s relationship with extremists like Reverend Jeremiah Wright and the former Weatherman terrorist Bill Ayers, they never would have elected him. And since tank-dwelling mainstream reporters couldn’t be trusted to expose The Real Obama, it would be left to the crusading online right to get the job done.

    The reality – that Obama is a moderate Democrat, whose political views would likely place him on the centre-right of European politics – just doesn’t get a look in. It’s very easy for us to point and laugh at the Tea Party types because their errors are so extreme. Outside of the bubble. the idea that he’s a radical socialist, a secret Muslim or Kenyan-born is obviously nonsense, but does that help us to forget that we’re sitting in our own bubbles?

    It’s easy enough to point to groupthink on the extremes where it’s obvious – the belief that if the Tories swung hard to the right and embraced the UKIP agenda, they’d get a majority, for instance, or the old Left belief that Labour’s mistake was not being revolutionary socialist enough in 1983 – but I think that there are many examples within the mainstream of politics too.

    In the closest parallel to Obama, consider the attempts to depict Ed Miliband as some kind of socialist firebrand dominated by the unions. As with Obama, the idea that ‘Red Ed’ wants to take the country back to some cartoon version on the 1970s is barely plausible in the real world but is an article of faith on the right. (The same applies to an extent on the left, though, where the caricature David Cameron drinks the tears of starving children with his nightly caviar)

    The problem is that the web has made it much easier to slip into groupthink mode. It’s very easy now to launch an attack on a political opponent, get lots of support and back-slapping from an army of Twitter warriors and congratulate yourself on a job well done, despite the fact that your attack has never registered with the public at all. However, you can point at the blog hits you’ve got, the retweets you’ve received, the likes and +1s you’ve achieved while not drawing attention to the fact that all these are coming from the same pool. It’s a classic reward for groupthink – do something that appeases the group and reaffirms their central idea and get praise, criticise it and get ostracised. (Or at least, not linked to.) Compare that to the work the old political operatives had to do to create their networks.

    Of course, you could argue that in order to exist and thrive, political parties have to practice some form of groupthink, otherwise they’ll splinter too easily over internal divisions.

    And no, I’m not excluding myself and my fellow Liberal Democrats from falling victim to political groupthink. Indeed, I think much of the party is falling into groupthink mode over staying in the coalition where lots of evidence is being ignored or twisted in order to proclaim that it’s a good thing and that we must stick it out for the long term. Slivers of good news get praised to the skies, while bad news is ignored or rationalised away. Don’t worry about a lost deposit in Corby, praise some local by-election victories instead!

    , ,

    Despite the fact I think that we shouldn’t be having Police and Crime Commissioners, I’m finding it hard to ignore the announcement of results that’s going on today. The electoral geek in me has come out, especially as the news I’m hearing from the Essex count is very interesting – the Conservative candidate has a lead after the first round, but the independent Mick Thwaites might be able to close that gap with second preferences. Results from across the districts show that the other independent candidate, Linda Belgrove, has also done well.

    The problem, though, is that this information is coming from what people who are at the count are tweeting. Officially, none of this is available to the wider world until the result is announced, and even then the announcement will just be the basic result for the whole county, not the breakdown by district. (As happens with European elections, the district-level results may be released later)

    This is the way all our elections get counted, with all the votes cast for a post announced together, and from what I understand it’s another way in which Britain stands alone. In other countries, votes are counted by where they’re cast, these results are announced and then aggregated together to give an overall result. This is what we saw in the US election a fortnight ago, with results being declared by precinct (roughly equivalent to a British polling district), and most of those announcements being made online on an official election site. This is why US media have the ability to call states before counts are completed – from seeing the results as they come in, they can project the result for the rest of the state.

    Over here, though, that information isn’t announced, and we all must wait until the full result is announced. Surely it’s not beyond the ability of returning officers to arrange counting and announcement by polling district, and for the Electoral Commission to create a site or sites for these results to be announced on? (Indeed, researchers and academics would probably find a single database of all local election results very useful, rather than having to scrape them from individual council sites)

    Declaring results by district would give everyone a lot more information – not just who won where, but how turnout varies across an election – and would likely make election counts and declarations more interesting. What would we need to do to make it happen?

    , ,

    Because history repeats itself, except for all the times it doesn’t.

    Barack Obama’s re-election this week meant that three successive US Presidents (Bill Clinton, George W Bush and Obama) had all been elected to two terms in office. The only other time this has happened in American politics was 200 years ago, with the Presidencies of Thomas Jefferson, James Madison and James Monroe.

    The interesting thing about this is that in both of these sequences, the President who served immediately beforehand was a holder of the office for a single term (John Adams and George HW Bush), who’d previously been Vice-President for two terms (for George Washington and Ronald Reagan). In the Jefferson-Madison-Monroe sequence, the next President was a son of that former President who served for a single term: John Quincy Adams.

    As perfect symmetry can’t be achieved unless someone finds a way for George W Bush to run again for the Presidency, the burden of history falls on the other politically-active son of George W Bush, former Florida Governor Jeb Bush. To keep up with history, he should win a bitterly disputed election that ends up decided by the Congress, after the ruling party has split several ways. The winner of the popular vote and the most electoral votes (Andrew Jackson in 1824) will then swear revenge, get elected for two terms immediately afterwards and make radical changes to the way the political system of the country works.

    Of course, the parallels break down when you look too closely, not least in how James Monroe’s period as President was known as ‘the era of good feelings’ with so little domestic strife that he was re-elected without serious opposition to his second term. When the historians write about this period of US history, I somehow doubt ‘good feelings’ will be used much. However, Jeb Bush is being mentioned as a potential Republican candidate next time round, so maybe history is preparing for the tragedy or the farce.

    ,

    Though it is a comment, from a Lib Dem Voice thread on Conference accreditation:

    It sometimes strikes me that some people see winning elections as an end in itself, without giving very much thought about why you are trying to win them or what you should do when you have. And paradoxically, it’s that kind of attitude that has paved the way for electoral disaster.

    ,

    Someone’s started up a petition on the Number 10 website to introduce STV in local elections in England. I’ve long thought that this would be a decent step to improving local democracy and accountability, so if you feel the same, why not go and sign it?

    ,

    I wrote last week about the odd situation in Gwynedd, where a lack of candidates forced a non-election in one ward, and that ward turned out to be crucial for overall control of the Council.

    Nominations for the ensuing by-election have now closed, and the electors of Bryncrug/Llanfihangel sadly haven’t shown a continued spirit of anarchism by refusing to nominate anyone again. Instead, having not been able to find a single candidate a few weeks, this time they’ve found five, including three different independents.

    However, it looks like the election won’t be as crucial as originally thought, as Plaid Cymru have now done a deal with Labour to run the Council. I’ll keep an eye out for other updates as this campaign rolls on, though.

    ,

    A couple of days ago, we had the story of the council ward in Wales with no candidates, and now we have the story of the ballot box in Glasgow that may or may not contain no votes.

    It seems that Glasgow Council has just discovered that one of the ballot boxes used for the Langside ward was recorded as having contained no votes, but that, on closer examination (or perhaps just looking in the box) it seems that there were some in there. How many there are, I’m not sure, but from what I can find on Glasgow Council’s site, the polling district in question (Battlefield Primary) contains up to 3434 voters. (The report on polling districts is a little confusing, and I’m not entirely sure if it’s proposing creating one district with 3434 voters, or two sub-districts at the same polling station with 2,405 and 1,029 voters. If the latter is the case, then the box could be for the smaller of the two districts and only be for 1,029 voters.)

    The official result is here, and Lallands Peat Worrier breaks it down in more detail here, complete with a graph that makes it easier to see where transferred votes (Scotland uses STV for local elections) have gone. What’s clear from that is that while the number of votes in there might not effect the SNP and Labour candidates who won on first preferences, the battle for third place is very close, and a few votes could change the outcome dramatically. And even with an ‘empty’ ballot box, they still managed a 35% turnout.

    Beyond the election itself, there are some important questions to ask, starting with just how a box with votes in it got recorded as being empty. However, that then opens up a whole set of other questions, notably starting with how no one raised an eyebrow at a ballot box (and the BBC report refers to it as ‘the’ ballot box for the polling station, not ‘a’) coming back empty. At that point, someone ought to have checked with the staff running the polling station to see if no one had come in to vote during the day. There’s also questions about how none of the agents or candidates spotted that a ballot box wasn’t being counted, too.

    Officially, though, that ballot box has no votes in it – that’s what the declaration of result says, and that stands as fact, despite what the evidence might say, until someone gets an election court to say otherwise.

    ,