The Eastleigh conundrum
by Peter Kellner in Commentary and Politics
Mon February 11, 9:02 a.m. GMT
YouGov President, Peter Kellner, analyses the Eastleigh conundrum ahead of its by-election
Most by-elections are events of only fleeting interest. Some are sufficiently dramatic to linger a while in the memory. Only a few truly matter. Eastleigh could be one of these. To show why, we must start with a trip down memory lane.
One of the institutions that wields real clout at Westminster is the 1922 committee. This is the body that represents Conservative backbench MPs and was, of course, founded in 1922. It is the circumstances of its creation that are relevant to our story. Tory MPs had been summoned to the Carlton Club at 11am on October 19, 1922. The summons had come a few days earlier to discuss the outcome of a by-election in Newport, where the party expected to do badly.
The contest was caused by the death of the MP, Lewis Haslam. He was a Liberal, and a supporter of the Con-Lib coalition government led by David Lloyd George. Four years into the 1918 Parliament, feelings were running high in both the Liberal and Conservative parties. The Liberals had already split between Lloyd George’s supporters and the anti-coalition group led by Herbert Asquith. Most Tories had backed their leader, Austen Chamberlain, in supporting the coalition but, by 1922, increasing numbers wanted to bring it to an end.
The Conservative by-election candidate in Newport, Reginald Clarry, belonged to the anti-coalition faction. The Liberal candidate, Lynden Moore sought to keep both wings of his party happy. Labour’s John Bowen, who had come second in 1918, hoped to take the seat, and many Conservatives thought he would. The meeting at the Carlton Club on the morning after the election was expected to discuss the consequences of a Labour victory, and how to fight the socialist menace.
Instead, at 2am, Clarry was declared the winner.* Moore came a poor third. By the time the Carlton Club meeting started, it was clear that British politics had changed. The Liberals were on their way out. The Conservatives no longer needed them. Chamberlain argued for remaining in coalition with the Liberals, but in vain. On a vote he was defeated by 187 to 87. He resigned as party leader; the coalition collapsed; a general election was held four weeks later; the Conservatives won a clear majority; the Liberals were finished as a party of government.
Eastleigh is unlikely to produce such dramatic consequences; at least, not immediately. But if the Tories do win, then we may we’ll look back in years to come as the contest that marked the beginning of the end of the current coalition. Many Tory backbenchers are likely to conclude, like the predecessors 90 years ago, that they no longer need to cosy up to the Lib Dems; and Lib Dem activists who feel queasy about their relationship with the Tories will urge the party leadership to end it well before the next general election. But if the Lib Dems hold the seat, then this will strengthen Nick Clegg’s leadership and sharpen the questions about the Tories’ ability to retake enough Lib Dems seats in 2015 to give them a real chance of winning a clear majority.
What, then, are the prospects? To save readers the time of reading to the end of this blog, my answer at this stage is a resounding ‘don’t know’. If the Lib Dems fight a good campaign and succeed in getting their vote out, then they should hold the seat; but those are huge ‘ifs’. By-election campaigns often develop a dynamic of their own. Anyone who says they know for certain who will win is fooling you, and possibly themselves.
So what can we say? Two polls have been published since Chris Huhne resigned as MP. The first, commissioned by Lord Ashcroft, reported a three-point Conservative lead; the second, conducted by Survation for yesterday’s Mail on Sunday, reported a three-point Lib Dem lead.
In fact the two polls disagree less than those figures suggest. Ashcroft’s figures are based on those considered likely to vote. The Tories say they are far more likely than Lib Dem supporters to turn out on the day. If we take everyone in the poll who indicates a party support, then Ashcroft’s poll produces a three-point Lib Dem lead, just like Survation. And one thing we do know is that the Lib Dems are able to fight a strong ground war in by-elections. They will do all they can to get their supporters to turn out on the day.
This is part of a wider point about early by-election polls. Campaigning tends to sway votes far more than general elections. The biggest reason for this is tactical voting. It has fuelled some of the biggest by-election upsets in the past. When one of the main parties has been seen as out of the race, many of its supporters have switched to their second choice in order to keep out the party they hate most. Early polls tend to miss this. The last time the Tories were in government, Labour supporters were happy to shift to the Lib Dems once they were sure that this was the best way to beat the Tories – while Lib Dem supporters were happy to return the compliment in Labour-Conservative battles.
What about today? With the Lib Dems in coalition, will Labour voters be as willing to vote tactically for them to stop the Tories? Will disgruntled Tories vote UKIP as a sign of protest – or return home to ensure a Lib Dem defeat?
To throw some light on this, YouGov has questioned a national sample on its willingness to vote tactically in a by-election. We asked about a variety of tactical circumstances. The one relevant to Eastleigh was this: Imagine that political commentators and opinion polls were saying that only the CONSERVATIVE and LIBERAL DEMOCRAT parties had realistic chances of winning the by-election in your seat - how would you then vote?
This is what we found:
in LD-Con contest %
As those figures show, the Labour vote does still get squeezed, albeit not as much as in some past by-elections. On our figures, the people who say they would vote Labour in a general election divide like this in a by-election contest that Labour couldn’t win: Labour 55%, Lib Dem 18%, UKIP 5%, Conservatives 2%, Greens 3%, other parties 7%, don’t know/would not vote 10%.
At this point a word of warning is essential. A national survey of this kind cannot predict what will happen in an individual seat. Anyone who attempts a precise extrapolation from these figures to the Eastleigh vote is wasting their time.
However, all else being equal, the by-election is for the Lib Dems to lose. They have a strong local party and an excellent record in recent local elections in the area. Our figures suggest that they should be able to squeeze Labour’s support to some extent. The Tories will hope to match this by squeezing UKIP support; but they may be disappointed. Past by-elections in the current parliament have shown that UKIP is able to win impressive numbers of votes even when they have had no realistic chance of winning the seat. Our poll also shows UKIP’s vote holding up in a Con-Lib Dem battle.
To win Eastleigh, then, the Conservatives must overcome a number of handicaps. But if they do, Nick Clegg could be in trouble. David Cameron would be delighted, so long as he can stop his backbenchers making the same immediate demand as their predecessors in 1922. On the other hand, a Lib Dem victory would suggest that the party is able to defy national trends in the seats they are defending. Either way, politics after Eastleigh will not be the same as it has been for the past 33 months.
* One quirky historical note. As John Ramsden points out in his excellent account of the Newport campaign in By-elections in British Politics (UCL press, 1997), the availability of alcohol was a key local issue. The Labour and Liberal candidates were non-conformists who held their meetings in the town’s Temperance Hall, while Reginald Clarry, the Tory, had the backing of the Licenced Victuallers Association. One contemporary account said that this swayed a number of working class voters. It seems that the final, fatal blow in the ‘strange death of Liberal England’ was struck not in England at all, but by the drinkers of South Wales.