In this post, Tony Travers uses historical data from local elections to analyse the chances of a Labour party victory in 2015. Mirroring the general message from most current forecasting models, he finds that Labour are some way off the level that has been historically necessary for opposition parties to replace the government in UK elections.
The 2015 general election will be determined by the impact of the vote shares of a series of ‘minor’ parties on the two big ones. It is likely the Conservative + Labour vote share in May will be similar to or below the 2010 level (65 per cent). It is the constituency-by-constituency movements of votes between, say the Conservatives and UKIP (or Labour and the SNP), which will determine which seats fall to which party. National poll vote shares are helpful, but only up to a point.
In the four years since 2010, there has been a full round of local government elections. In these ‘real world’ contests, the impact of declining Liberal Democrat vote shares and UKIP’s recent surge has played out on four occasions. Mid-term unpopularity for the government has led to a fall in the Coalition parties’ vote share. Labour has won additional votes and council seats.
In this post, I convert the vote share in each year’s local elections into a ‘national equivalent vote’ for each party. The National Equivalent Vote Share (NEVS) uses the votes cast in a particular set of elections and adjusts them to take account of the pattern of contests in a particular year. The NEVS therefore provides a snapshot of country-wide voting preferences, regardless of where council elections are held.
In the past, parties which were on the way to winning a parliamentary majority would generally win 40 per cent or more of the NEVS in years immediately prior to the general election. On no occasion since 1979 has a party defeated the government without a solid NEVS performance averaging about 40 per cent in the four years prior to the election. Having said this, it is possible to win around 40 per cent on average and not go on to win the election.
In the four years since 2010, Labour has had an average NEVS of 34 per cent, with the 2013 and 2014 figures being 29 and 31 per cent respectively. The figure below shows the national equivalent vote share lead of opposition parties from 1983 to the present. Elections that resulted in a change of government are coloured red, and elections where the opposition failed to win are coloured black. It is hard to see Labour’s NEVS performance between 2011 and 2014 (coloured gray) as a convincing basis for an election victory. No Opposition since 1979 has managed to win a general election with such a low average NEVS lead in the four years immediately before the election. Moreover, governing parties tend to poll higher in the following general election than they did in the previous years’ local election vote share average. Oppositions generally poll lower in the general election relative to their pre-election NEVS average, though Labour in 1997 out-performed the previous years’ NEVS average.
But, it might be argued, things are different now. The Conservatives have seen their vote attacked by UKIP, while Liberal Democrats have switched more to Labour than the Conservatives. A low vote share for Labour might be enough to secure victory.
So what has happened to the Tory tally of councillors since 2010? If the party was being attacked by UKIP, it would surely have lost a large share of its councillors. Oddly not. The Conservatives have lost just 10.4 per cent of their 2010 councillor tally in the five years of the Coalition. By contrast, Labour lost over 20 per cent of their councillors in each Parliament from 1997 to 2010.
The Lib Dems have lost about 40 per cent of their councillors, taking the brunt of the Coalition’s unpopularity. But the overall tally of councillors lost by the Coalition partners is just under 20 per cent, not a particularly bad performance compared to Labour in the Parliaments between 1997 and 2010. Labour has won over 2250 councillors since 2010 and now has about 7100 in total. But this number is well below the 10929 (on a slightly larger national base of councillors) the party held just before their 1997 election victory.
None of this is to say Labour will not win in May next year. But, this analysis of local election results mirrors the general message coming from different forecasting models: history suggests that Labour is still some way from being strong enough to win a majority in 2015.