Amateur PCC election number crunching

Since I blogged about the PCC election, I’ve been doing a little number crunching. This is frankly the first time I’ve looked in any detail at election results and my GCSE in Maths didn’t include any statistics beyond mean, mode and median, so take the theories and conclusions below with a pinch of salt and do check my manually copy-and-pasted numbers. You can download the whole Excel 2010 spreadsheet and do your own number crunching as well, and it is mostly normalised, unlike the Guardian spreadsheet.

I had proposed that small parties did and would have done rather well at the PCC elections. To back that up, I came up with a statistical measure of my own based on the thinking behind page-rank. Google’s Page-Rank algorithm attempted to measure how likely it was for a user to arrive at a web page if they started on a random page and clicked on random links. My measure assumes that if voters voted at random they would normally vote for one of 192 candidates about 0.5% of the time, a share of the total vote better than 0.5% indicates a candidate making disproportionate impact. Obviously, a shortage of candidates to vote for would impact how often voters could vote for one group or another in a constituency, so I’ve noted how many candidates stood for each party. I treated independents as if they were in one large party, although that gives them an unfair advantage with 54 candidates in 41 constituency areas.

So what happened?

So the first surprise is that the big two Labour and Conservatives are still dominant despite all the talk about how well independents did, and despite the Tories uncomfortable term in power. They are the only two parties that beat 0.5% per candidate. The Justice and Anti-Corruption Party are encouraging . They made a bigger splash per candidate than UKIP, the Lib Dems and even those new fangled independents. Economically far left Green and racist parties bring up the tail. Is there evidence here for the theory that small parties did well too? Well, that stellar result for JAC on this measure is extremely encouraging, and UKIP beat the LibDems into 6th (or 4th depending on how you look at it).

Then again, that JAC result is a 5th place in an area with 15% turnout (slightly below the 15.8% average) if JAC were stellar performers, might we have expected them to win something? What about better known UKIP? Labour and the Tories still dominated with their name recognition, why did UKIP fail to outperform JAC with better name recognition. The Justice and Anti Corruption name is perhaps a better name for the ballot paper, but perhaps it’s time to look at another measure:

Based on a simple share of the vote, it’s obvious that the coalition has been thoroughly spanked compared to the 2010 General Election result. The economic far left continues to trail, with a reassuringly poor result for anti-human Greens. Their anti-colour counterparts the British Freedom Party do well to drive them into last place from nowhere, or so it seems. If you compared them to the BNP their result would be a -91% spanking. While the racist boogeyman was talked about in advance of the election, there is no obvious explanation for the Green’s being ostracised in an election about law and order.Their candidate didn’t even mention cannabis.

JAC look good again with a share of the vote just half that of better known English Democrats but with just one fifth of the candidates. The result for independents is ludicrous 12,000% higher than at the general election. I believe this is a reflection of the different thinking that goes into selecting someone to oversee the police, where party politics is seen as a weakness of the new system. What is interesting though, is that UKIP and the English Democrats both received a share of the vote that was multiples of the General Election share.

So is this somewhere that a libertarian party can thrive? I think it depends. If a libertarian party is seen as “pro” drugs and vice then I don’t think it would do well. It would need to emphasise the Peelian principles, and talk instead of victimless crimes. My favourite example of a victimless crime is the one I committed by driving at 70 at night, sober, on a straight and empty section of dual carriage way with an unusual 40 mph speed limit. Motorists of the world might rejoice if that kind of crime stopped being a crime. Nationalising or abolishing the private company ACPO, scrutinising the use of speed cameras and privacy invading Automatic Number Plate Recognition are other unique selling points for a libertarian commissioner. A recent poll found that 5% of the population were economically minarchists, if just half that proportion translated into law and order libertarians then we could share the middle of the table with the Liberal Democrats.

Here’s the kicker. There is no reason an independent could not also campaign on those policies, and if a watermelon can get more than 8 thousand votes by proposing a green audit on the police then rebalancing the relationship between police and public should make a splash. As an independent, our ideal candidate would be eyeing up that 12,000% leap in vote share enjoyed by other independents and wondering very seriously if a party label of any kind is simply unhelpful. Even if 400% is better than a poke with a stick, a 12,000% leap is thirty times better than that. For a candidate professional enough to sell at an election of this kind, that sort of challenge will be difficult to overlook.

If only there were a way to sell independence and libertarianism at the same time.