The commonly used yardsticks of how well the national parties are doing in local elections can be confusing or deliberately misleading, and there is likely to be a lot of ‘spin’ after the elections of Thursday 2 May. Lewis Baston argues that studying the votes cast in marginal parliamentary constituencies, particularly in areas where there have been other local elections since 2010, is a better guide to electoral trends than council control, seats won, or even – sometimes – aggregate votes cast.
The primary purpose of council elections is to provide democratic local government, but this is often overshadowed by the way the media, politicians, analysts – and of course voters – see them as a way of taking the temperature of the national political scene. Before the election, politicians manage expectations downwards, for example Conservative Party Chairman Grant Shapps has said that the party expects to lose 500 council seats (they don’t, and won’t). If Labour really does not expect a net gain of more than 200, they are anticipating a result significantly worse than the county elections in 2005, when Labour had a three-point national lead in the concurrent general election. This also seems dubious.
There are four yardsticks are used to assess performance in local elections. Much the worst way of doing it is to point to results in councils where your party did well as being somehow a better reflection of the national mood than areas where the party does badly. The Conservatives were able to spin the May 1990 council elections as some sort of success because of good results in a handful of London boroughs, against a disastrous national background. The Liberal Democrats might be able to pull this trick in 2013, if they manage to gain Somerset from the Conservatives.
More seriously, one can look at changes in council control. This will vary from year to year based on which councils are up for election and when those seats were last contested. Labour, for instance, pulled in a good haul of councils in 2012 because the previous time those seats were fought was in the party’s disaster year of 2008. In 2013 the councils being contested are Conservative-inclined in an even year, and most of the seats were last fought in 2009 which was anything but an even year. The Tories won all the shire counties except Cumbria and will certainly lose some. Derbyshire should be in the bag for Labour and Nottinghamshire probably likewise, even if the party is having a disappointing day. Labour should be able to gain Lancashire (where there are several key marginal seats) as well, but Staffordshire is a more uphill struggle – from 32 seats and a narrow majority in the 2005 elections Labour crashed to a humiliating 3 councillors in 2009, damage that may be too much to recover from in one go. This is pretty much the entire target list for Labour as far as control is concerned, although the party can hope, on a good day, to be the largest single party in hung councils in Warwickshire, Northamptonshire and possibly Suffolk. Labour can also hope to gain two mayoralties (Doncaster from Independent, North Tyneside from Conservative), and possibly Northumberland unitary council. But it will be slim pickings.
Two things need to happen for the Conservatives to lose control of a lot of the county councils. First, the party needs to be significantly behind Labour in national polls. Second, there needs to be a big challenge to the Tories in their heartlands from a third or fourth party. There are, even in good years for a party, some areas where they just cannot win more than a handful of seats – the Conservatives were unviable even in 2008 in places such as Barnsley and Manchester, and even in a stellar year for Labour they are not going to sweep West Sussex. But the Tories can lose overall control if they shed seats to the Lib Dems, as they did spectacularly in 1993 when they lost everything except Buckinghamshire, or – as they might in 2013 – to UKIP or Independent candidates. But 2013 is almost bound to be less extreme, because the big towns and cities were taken out of the county areas in 1997. For example Hampshire, without Southampton and Portsmouth, would have been Tory even in the 1993 deluge.
Some of the best UKIP areas are having elections this year. The party did reasonably well in 2009, benefiting from the Euro elections being on the same day and the climate of the expenses scandal, but it is riding higher in the polls now. The difference between a breakthrough in terms of seats, and merely getting respectable second places in lots of wards, is whether UKIP gains in votes are concentrated in winnable seats or whether it increases its votes evenly. It could also tip seats to the Lib Dems in the South West and South East, and Labour in the East of England and the Midlands, as it did last year in places such as Great Yarmouth.
It has been increasingly common to use parties’ net gains and losses of councillors as benchmarks. This is trickier than it seems. Vastly different numbers of seats are contested – in 2009 there were 2,318 seats up, while in 2011 there were 9,460. The starting point will also vary – a party defending seats fought in a peak year must expect to lose seats without it indicating particular unpopularity. Results will also not be completely comparable with elections from four years ago, because sometimes council electoral cycles change – for instance, there were no elections in Durham or Northumberland in 2009 but there will be in 2013. There are boundary changes in some authorities (Northamptonshire council shrinks by 16 seats), and also by-elections and defections have changed the composition of some councils since 2009. There have been a number of recent defections from Conservative and Independent to UKIP, for instance. Some councils have lots of seats and this affects the numbers – there are 126 in Durham, and if Labour bounces back strongly since 2008, the party could gain 40-50 here alone. An equally strong swing in Northamptonshire might net about 10. Although it is the most used measure, net gains and losses is also one of the most confusing and unsatisfactory.
The popular vote is a much better way of assessing and comparing local elections. But when people talk about the share of the popular vote in local elections, sometimes they mean different things. One obvious way of doing it is to simply count up the votes cast in one set of local elections and then work out the party shares. However, this will not be comparable with what happened in 2010, 2011 or 2012 because different parts of the country have local elections in different years. The English shire counties have long been the part of local government where Labour has been weakest. Even in the 1997 county elections, which took place on the same day as Labour’s landslide general election win, the Conservatives were still the largest party in the counties in both seats and votes (by a margin of 5 percentage points).
Table: Actual votes cast and national equivalent vote in county elections since 1997
Source: Rallings & Thrasher Local Election Handbooks – note that except in 2009 the NEV figures are the general election outcomes.
After the elections psephologists calculate the ‘national equivalent vote’ (NEV). This essentially means asking what the shares of the votes cast would have been had there been elections everywhere, rather than just a sub-set of local councils. It is a perfectly valid exercise, but it is a model – and sometimes different psephologists come up with slightly different national numbers depending on their sample and methods. And there are some questions one cannot answer – there will be no way of assessing the fortunes of the SNP in these elections because there are no Scottish local elections. It also takes a little time to calculate accurately, with final figures not available for a few days – long after the media-political spin cycle of covering the results has finished.
This set of elections are, much more than any of the other four years of the normal local election cycle, very much small town and rural, and concentrated in the south of England. This suggests that extrapolating from these council elections will tend to understate Labour’s national strength, because these areas are precisely where the party’s recovery since 2010 has been weakest in polling and in previous local election results. The three broad regions of England have performed differently in 2011 and 2012. Putting it crudely, the north swung heavily to Labour in both years, the south more weakly in both years, and the midlands had low swing in 2011 and high swing in 2012. Even if Labour’s national strength is entirely unchanged since their good year of 2012, it will look less good because there were elections in 2012 in the north and the cities, while there is not in 2013.
If Labour appears to be doing slightly less well than in 2012, this may be illusory – a product of the very sharp regional and urban/rural divides that have opened up in British politics that gave us a northern, urban snapshot last year and a southern, rural one this year. The most valid comparisons will involve the results in marginal seats that have had elections in 2011 or 2012 as well as this year – Ipswich, Harlow, Hastings, Redditch, Tamworth, Pendle and Carlisle, for example. It is in these sorts of places that general elections are decided, and the votes cast in these areas are the most accurate indicator of whether Labour is on course to return to power in 2015.
Note: This article gives the views of the author, and not the position of the British Politics and Policy blog, nor of the London School of Economics. Please read our comments policy before posting.
Lewis Baston is a writer on politics, elections, history and corruption. He is Senior Research Fellow at Democratic Audit, and was formerly director of research at the Electoral Reform Society.