Today, I found myself in the environs of Westminster and, on College Green, a film crew were patiently seeking to record a scene with some of the cast from “The Thick Of It” while people like me got in the way. Some passing American tourists, on hearing it was the BBC, immediately jumped to the conclusion that this was a news crew interviewing real politicians with the Palace of Westminster in the background. But no, though it may have the trappings of normal politics, it isn’t. It’s just show business.
The same of course, could be said of the Bradford West by-election, where the bizarrely-titled ‘Respect’ candidate, serial showman George Galloway, trounced everybody else who stood. As it was considered a safe Labour seat, the Labour party will feel it most keenly. Is it anti-politics? Is it what happens when you excite the sympathies of local mosques in an area with a high muslim population? Each explanation seems initially attractive, but soon seems beset with problems. At the moment, no-one really knows.
Against this rather unfortunate backdrop the Police Foundation released a study of what might predictably happen under the normal rules of politics (well, psephology actually) in the Police and Crime Commissioner elections, based on the last general election and recent opinion polls, which is here, with a methodology here, and a previous worked example of guesstimating second preferences here.
The study predicts results in each of the 41 police areas on the basis of 4 scenarios – the 2010 general election with and without the LibDems, and recent polling with and without the LibDems. If your party wins under all four scenarios in your area then you may feel pleased. If it loses on all four, you may be dischuffed, and if it is more mixed, you may not know what to think.
Please note this is not what my former colleagues in the police would call “bobby-proof”. The warnings about interpreting it aren’t the general terms and conditions of software licences that you might be tempted to skip past. In particular, please note that the 2010 General Election may not be the best place to start when making such estimates. My own suspicion is that turnout may be a lot more like a local election and voter distribution is anyone’s guess in a single-issue election. So, by all means accept that this is a good piece of work marred only by the comic timing of its release, but don’t mislead yourself with the pretty graphs, especially when the real information is in the tables. As Labour’s Sussex hopeful Paul Richards argues, those up for election need to win each vote afresh at each election. Take it from him – he loses on all four scenarios.
TopOfTheCops.com 
What poverty of aspiration the police foundation seem to have shown here. They have failed to take into account the role of Independents (which they do at least make clear). This is fine if we are looking at a response to a Parliamentary election but simply misses the point that many within the political parties, let alone those outside them have a feeling that this is not a suitable job for a party politician. What is needed is analysis which assumes that the outcome is certain, only in those locations where the largest party actually polls more than 50% of the votes based on 2010. I have no idea how many locations this locks up but certainly in Sussex, the Conservatives would not get 50% of the vote based on 2010. This then opens up the possibility that a strong Independent is the second largest party (clearly depends on a v strong candidate). If they then appeal to those who would give their second preference to them, we have some targets where a good Independent is a realistic outcome. For me Sussex is entirely winnable on that basis, with or without Lib Dems.