Liberal Democrat Newswire #40 is out: predicting Lib Dem seat numbers for 2015

Each morning at home when I open my wardrobe I am reminded of the vagaries of predicting how many seats the Liberal Democrats will win based on stuffing mid-Parliament opinion polls through a uniform national swing (UNS) calculation.

Such predictions are the easy, tempting calculations to make. But they have often led to people hugely under-predicting the likely Lib Dem haul of seats, and a result skewing the betting odds available to generous levels – something which has kindly financed much of the smarter end of my wardrobe in previous electoral cycles.

The 1997 general election was the classic case of how vote share and seat numbers can be only very loosely, if at all, connected for the third party in a first past the post system. The Liberal Democrat vote share fell between 1992 and 1997 (down 1%), but the number of Parliamentary seats won by Paddy Ashdown’s party and Chris Rennard’s target seat campaign soared (from 20 to 46).

It made 1997 a decisive break from the past electoral history of the Liberal Democrats and its predecessor parties, as shown by a key ratio that is rarely talked about.

It’s the seats:votes ratio.

For a long time prior to the formation of the Liberal Democrats, the ratio for the Liberal Party and then the Alliance was stuck at less than 1. In 1970, for example, the Liberal Party won just 6 seats on 8% of the vote, for a ratio of 0.75. Even a ratio of 1, finally achieved in 1992, was not all that much – after all there are 650 or so seats in Parliament (the number has varied a little over the years) but only 100% of the vote to go round. For the winning Conservatives in 1992 the ratio was, by contrast, 8.

In that 1997 general election, the third party finally broke the long record of only winning roughly as many seats as its percentage share of the vote, getting 46 seats on 18% of the vote for a ratio of 2.7. That ability to turn votes into seats got even more efficient in 2001 and then 2005, peaking at 2.8. In 2010, however, it dropped to its worse showing since 1992, back down to 2.5 (thanks to the Lib Dem vote going up, but the number of seats won going down).

This ratio, therefore, provides a different way of looking at how many seats the party is likely to win in 2015.

It has become accepted wisdom amongst political pundits that the Liberal Democrats are likely to do fairly well at the next general election in bucking the national trend in the seats that matter most to the party. However, just how far does it have to buck the trend to return a reasonable number of seats? And what if the 2010 slippage in the party’s ability to buck the trend continues?

That is where looking at the seats:votes ratio is so useful as a tool for understanding the likely range of outcomes, but first let’s look at the evidence for the party’s ability to buck the trend at all and then at what sort of share of vote the party can reasonably aim for.

Liberal Democrat MPs are the only ones to get a thumbs up from their constituents in new research by Phil Cowley and Rosie Campbell.

Lib Dem MPs scored a net +14 satisfaction rating, compared to -5 for Labour MPs and -13 for Conservative MPs. Crucially for their re-election prospects, Liberal Democrat MPs are also the only ones to come out with a positive rating amongst people who say they are not currently intending to vote for their party: +7.

Labour MPs get a -16 rating from non-Labour supporters in their seats, whilst the huge difficulties the Conservatives face in appealing outside their existing vote is shown by the massive -31 rating their MPs get from non-Conservative supporters. That may not matter too much for existing Conservative MPs, but it is indicative of the big struggles the party faces in seats where they need to win over new support to win.

The logic behind the Liberal Democrats’ 75 seats strategy is that although the next general election will be overwhelmingly a defensive matter for the party, there are some seats the party can look to pick up, particularly from the Conservatives. As local government election results have shown, Lib Dem / Conservative battles are much easier, and indeed sometimes even very promising, territory for the Liberal Democrats than Lib Dem / Labour fights, which are much tougher (though keep an eye on councils such as the London Borough of Haringey and Parliamentary seats such as Ashfield).

In those Conservative / Lib Dem fights, Nick Clegg’s party has been helped by UKIP eating more into the support of other parties than into its own support. However, the Lib Dems are also managing to eat into Conservative support in such places. Indeed, Lord Ashcroft’s latest polling in a series of Con-Lib Dem marginals found that, “extraordinary though it may seem – nearly as many Tory Defectors have gone to the Liberal Democrats as have gone to UKIP”.

Overall, the net effect is that the Lib Dems continue to poll strongly in such marginal seats compared to the Conservatives.

Ashcroft’s polling finds that the Tory lead over the Liberal Democrats in eight key marginal seats held by David Cameron’s party has only gone up by 1% since 2010.

In 2010 the Tories won the eight seats of Watford, St Albans, Oxford West & Abingdon, Harrogate & Knaresborough, Camborne & Redruth, Truro & Falmouth, Newton Abbot and Montgomeryshire with a 2% lead over the Lib Dems. Both Lib Dem and Tory vote shares have fallen since 2010, but the net effect is to leave the Tory lead little changed. (Tory support has fallen from 41% to 32%, whilst Lib Dem support is down from 39% to 29%.) The Tory lead is now 3%, only 1% up and so well within the poll’s margin of error. There are also a significant number of Labour voters still willing to consider voting tactically for the Liberal Democrats.

One tricky finding from the poll for the Lib Dems is that the four most important issues to voters in those seats are economy/jobs, the NHS, immigration and education. That makes the party’s current review of its immigration policy all the more important – and is a reminder of why the party focused so heavily on jobs at its Glasgow conference.

Also worthy of note is the edge – but only small edge – in volume of local campaigning activity that voters in these seats report between the Lib Dems and the Conservatives. For example, 41% recall having had a Lib Dem leaflet in the past few months with the Conservatives on 36%. That gap is small – and the 41% is low for a marginal seat with a full-on operation. The campaigning edge in this poll is smaller than in a YouGov poll in marginal Con/Lib Dem seats in 2008. Although not directly comparable (different pollster, different question wording), the contrast between the results then and now is a warning sign to the Liberal Democrats that the party’s local campaigning reputation is not what wins votes, it’s the manifestation of that reputation in actual work on the ground that does.

More positively, however, the recall of having been leafleted by the Lib Dems is higher in the 2013 Ashcroft poll than in the 2008 YouGov one, and recruitment patterns for Conservative constituency organisers suggests the party will only be putting modest amounts of efforts into seats it hopes to gain from the Liberal Democrats.

Its 308 pacy pages cheerfully zig-zag between marketing manual, self-help book, and campaigning A-Z — with dollops of political history, pop-psychology, and behavioural economics thrown in for good measure – Stephen Tall

101 Ways To Win An Election is available from Amazon (paperback and Kindle editions).

For Apple fans it is available on iTunes as an iBook for iPad, iPhone and iPod.

Users of Kobo readers are also catered for with the Kobo ebook version.

Although senior party figures are not talking in public about any particular share of the vote they are aiming for, high teens is the implication of many of their comments, both in public and in private. It fits with public and private polling showing around a quarter of the electorate are willing to think seriously about voting Liberal Democrat in 2015 – not all of whom of course can realistically be turned into actual votes on polling day.

Mid to high teens is also the figure suggested by history. The current party poll rating averages out this year at 7-8% lower than this time in the last Parliament. A 2015 result on a similar scale would mean the party securing 16-17%.

However, another way of looking at it is to see how the party’s poll rating has changed between this far out from polling day and polling day itself in previous Parliaments. The movements have been:

• September 1990 – 1992 general election: +6
• September 1995 – 1997 general election: +1
• September 1999 – 2001 general election: +4
• September 2003 – 2005 general election: -1
• September 2008 – 2010 general election: +6

That averages out at +3 or (if you exclude the special circumstances of the Iraq war Parliament) +4. That points more to a mid-teens result of 13% or 14%. (All these figures come from the largest public database of UK voting intention opinion polls, which I maintain at www.markpack.org.uk/opinion-polls.)

How realistic you think a high teens target is depends in part of course whether you place more faith in current opinion polls such as those by Opinium (latest Lib Dem rating 7%) or ICM (latest Lib Dem rating 14%). In the past, ICM has been a very good guide to the party’s fortunes. Will history repeat itself?

One respect in which history will not repeat itself is the popularity of the Liberal Democrat leader. A canny reading of the evidence, though, suggests this will not matter.

Nick Clegg’s overall public ratings are far lower than they were for him in 2010 (even pre-TV debates), or for Charles Kennedy in 2001/2005 or Paddy Ashdown in 1992/1997. Yet in the past high ratings for the party’s leader have only been very weakly correlated with voting for the party.

Previous third party leaders specialised in being popular with large numbers of people who were also determined not to vote for the party. Nice but not useful. (See the book Putting Voters In Their Place for a good analysis of the evidence.)

Therefore Nick Clegg’s much lower post-2010 ratings matter much less than might appear at first. Moreover, party strategists are very bullish about how Clegg is viewed by that quarter or so of the electorate that is willing to think seriously about voting Liberal Democrat.

YouGov’s most recent poll for the Sunday Times, for example, gives Nick Clegg a -45% rating overall. But there are 23% who give him a positive leadership rating, and amongst Lib Dem voters his rating is plus 48 (better than Ed Miliband does amongst Labour supporters).

Used judiciously and carefully targeted, Nick Clegg can be a significant electoral asset for the party. (Which, as an aside, makes the paucity of his social media presence all the odder, given how many opportunities social networks offer to target such possible supporters.)

Turning all these factors into seat predictions boils down to two questions: what will the Lib Dem share of the vote be, and how good will the party be next time at converting national vote share into seats?

That is two variables to juggle, and so the following exclusive table lets you pick what you think likely on both and see what it means for seat numbers (click on the table for a larger version). Pick the likely percentage share down the left and the seats:votes ratio along the top, and then the table tells you how many seats that would produce.

For example, if the party polls the 14% from the latest ICM poll and is as good (or bad, depending on your point of view) at turning votes into seats as in 2010, then that would give the party 35 seats (highlighted in gold). However, if the party manages to creep up to 16% and repeats its 2005 performance at turning seats into votes, than translates into 46 seats, down on 2010 but still above the 1997 breakthrough figure.

Remember, 2.9 was the 2005 ratio and 2.5 was the 2010 ratio – and have fun predicting away…

How much did you like this newsletter?

lowest 1   2   3   4   5   6   7   8   9   10   highest
Sorry, voting is closed.

I hope you’re found this edition of Liberal Democrat Newswire interesting, informative, useful – or all three!

Like this newsletter? Then let your friends know about it too. Just point them at this sign-up page.

Best wishes and thank you for reading,

Mark