A highly unpredictable set of elections

Now that nominations have closed for the Police and Crime Commissioner elections we can have a better stab at trying to work out where the battlefield might be on 15th November, a difficult task as this is the first time elections have been fought on these boundaries, turnout is likely to be derisory, the election system is Supplementary Vote (voters cast a 1st and 2nd preference), which institutionalises tactical voting, and there is an unusual pattern of candidates.

The table below shows the notional winner and second place party if the PCC boundaries had existed at the 2010 General Election, as calculated by the Police Foundation. I have also indicated whether or not there is a Lib Dem candidate and who the other candidates are.

The table is broken down into notional Lab vs Con, Lab vs LD and LD vs Con fights, ranked in order of marginality.

Some points become clear from looking at the table:

• Labour only “won” 14 of the 43 PCC areas under First-Past-the-Post in the 2010 election. Four of these were marginals with majorities under 10%. This is because the boundaries tend to be drawn so that rural Tory areas outvote urban Labour ones, and two of Labour’s strongest areas, London and Scotland, won’t be electing PCCs.
• The next most winnable PCCs after the ones we notionally “hold” are Derbyshire, Lancashire, Humberside, Dyfed-Powys (where we start notionally third but the Lib Dems are not standing), Cheshire and Cumbria, all with notional majorities under 10%, and then the more tricky Staffordshire and Leicestershire.
• The swing in the opinion polls since the last General Election is about 7.5% based on the latest YouGov lead of 8%. If replicated on the 15th November with a uniform swing and with the Supplementary Vote system and the independent candidates having a limited impact, Labour would take all the areas down to Leicestershire, but just miss Bedfordshire and Warwickshire. In the unlikely event that this happened it would give Labour 22 of the 43 seats.
• The Lib Dems only “won” one of these areas in 2010: Avon & Somerset, and that by a tiny margin over the Tories. This was when they were getting about double the percentage support the polls suggest now.
• There are 17 areas with no Lib Dem candidate: Cleveland, Derbyshire, Durham, Dyfed-Powys, Essex, Gwent, Kent, Leicestershire, Lincolnshire, North Wales, North Yorkshire, Nottinghamshire, South Wales, Staffordshire, Suffolk, Sussex, Warwickshire, and West Mercia.
• The pattern of Lib Dem withdrawals doesn’t seem to indicate any tacit deal with the Tories in that they are not concentrated in Lab vs Con marginals.
• There is a geographical pattern though – the gaps in the Lib Dem field are concentrated in Wales (4 gaps), the Midlands (7 gaps) and the counties to the south and east of London (3 gaps).
• Extraordinarily the Lib Dems are not running in Dyfed-Powys where they are notionally a close second, leaving this as a straight fight between Labour and the Tories, in an area Labour could win. There are three other areas which have 1950s-style elections with only the two candidates, Labour and Tory: Staffordshire (which may increase the chances Labour can win it), Durham, and North Yorkshire.
• Some areas with strong pockets of Lib Dem support don’t have a candidate e.g. Kent where only last week they got over 2,000 votes in a county council by-election in Maidstone, or Essex where they have sitting MPs (Colchester).
• You have to go back to the 1970 General Election to find a time when so many voters in an (almost) national election were not presented with a candidate from the Lib Dems or their predecessor parties to vote for. It makes the Lib Dems look like they are not a national party, a problem they had from the 1930s to 1970s, and means many Lib Dem voters will be faced with a “forced choice” between other parties which may get them into the habit of voting for another party.
• UKIP, with 19 candidates, seem to have deliberately targeted the four most marginal notionally Tory-held areas, which may damage the Tories given how well UKIP are polling in current opinion polls.
• Whilst the BNP have not managed to field candidates (perhaps too many are disqualified as they have criminal convictions?), the EDL are running in their Bedfordshire heartland and the English Democrats have six candidates.
• There are no Plaid Cymru candidates, which may help Labour in the two Welsh marginals.
• Labour may also be boosted by the absence of Green candidates everywhere except Cleveland. For instance there are the thousands of Green votes in the Brighton and Norwich areas up for grabs.
• The “wild card” is the large number of independent candidates in what some voters may perceive to be elections for posts that should be non-partisan. They range from extremely credible candidates with policing backgrounds to no-hopers, and their ability to campaign ranges from just being a name on a ballot paper to extremely well funded and professional operation.

All-in-all this makes for a highly unpredictable set of elections, similar in feel to the old pre-1999 FPTP European Parliament elections, when large numbers of seats would change hands on a low turnout.