Local Elections 2013 – Seat Projections Too ‘Conservative’?
May 2nd – Preview
The current slate of local elections are mainly a refresh of the June 2009 County Council elections. Since then we’ve had a sea change in the UK political scene. The formation of the Coalition and a decline in support for those parties, a huge improvement in polling for Labour and the emergence of UKIP on the national stage – perhaps in part a product of the Coalition itself.
An average of national polls in the month before the June 2009 local (& European) elections put the parties in the following positions for Westminster voting intention:
CON: 39% LAB: 23% LD: 19% UKIP: 6% Others: 13%
In the Survation poll, published today – May 1st, Westminster voting intention is as follows:
CON: 29% (-10) LAB: 36% (+13) LD: 12% (-7) UKIP: 16% (+10) Others: 8% (-5)
(Change since the 2009 pre-election polling average)
So huge changes in support for the parties since then.
Political academics Professors Colin Rallings and Michael Thrasher have provided their excellent annual analysis of the parties’ potential vote share based on their performance in council by-elections in 2013 and provide the following projections for Thursday:
Rallings & Thrasher have a history of accurate projections, but this year have indicated that a surging UKIP, combined with UKIP standing, often for the first time in almost 73% of seats up from less than a quarter in 2009 make this more difficult.
In 2009 UKIP stood in 593 council seats, winning 8 and finishing second in a further 125. If we apply a swing to those 593 seats that is a little less extreme than the change in national polling since 2009 (UKIP up 8% in their best seats compared to 10% nationally) then the results are shown below.
(In seats where UKIP stood in 2009)
2009 average vote share
2013 Survation projection
A uniform swing projection on this basis results in UKIP winning 79 of the 593 seats where they stood in 2009, already ahead of the Rallings & Thrasher projection. If in addition they were to pick up seats in a few places where they did not stand last time (and some local council by election results from the last year suggests that they might), their total number of seats won could well be in the 80-120 range.
If UKIP do perform better than the Rallings & Thrasher projection then the Conservatives would be expected to suffer equivalently worse losses. The Liberal Democrats meanwhile could slightly outperform expectations, as a high UKIP vote share will help them hold onto seats that the Conservatives would otherwise have threatened, in the style of the Eastleigh by-election. The Liberal Democrats’ strength has always been in their local party organisations and since last November they have gained 8 council seats through local by-elections as well as holding 6 more whilst suffering no losses. Overall, then, Conservative losses might be looking closer to 400 rather than the 300 or so seats predicted.
If UKIP does even better than expected, however, then their gains may start to eat into Labour and Liberal Democrat seats as well. The table below shows an estimate of how UKIP’s vote composition changes as their vote share rises. At first it is overwhelmingly Conservative, but these are the easiest votes to pick up and form a smaller share of UKIP’s vote the more popular it becomes. The share from Labour switchers starts very low but increases steadily as UKIP grows in popularity and starts attracting traditional working class voters concerned by issues such as immigration. Meanwhile there is a significant block of LibDem defectors early on that are quite easily won over (presumably of the ‘protest voter’ type) but then very little additional switching from the Liberal Democrats after this (as the remaining core LibDems who genuinely support their liberal, pro-European and environmental policies are quite far politically from most of UKIP’s policies).
If this breakdown is accurate (and we are in the process of adding more data) then the interesting conclusion is that the optimum UKIP vote share as far as Labour is concerned is actually around 16%. After that point, more of the marginal switchers to UKIP start to come from Labour than the Conservatives (even though the overall UKIP vote is still predominantly Tory into the 20%s, after 16% further growth in UKIP vote is offset by the falling Conservative share).
At this maximum point UKIP increases Labour’s lead over the Conservatives by about 5 percentage points, a potentially election-wrecking margin. Interestingly 16% is the UKIP vote share recorded in today’s Survation poll. In other words, UKIP is currently polling at the worst possible level for the Conservative Party’s prospects and is responsible for over half of Labour’s current lead in the opinion polls.
Analysis by Patrick Briône and Damian Lyons Lowe. Scatter chart compiled by Survation Researcher William Mosseri-Marlio. Majority of historical data points taken from various British Polling Council member polls. Data for vote share above 20% uses Eastleigh by-election polling, though trends lines are similar if just extrapolated from the other data.