In the coalition negotiations between the three large British parties, both the Tories and Labour offered the Liberal Democrats the possibility of changing to the Alternative Vote electoral system, and in the end David Cameron committed himself to holding a referendum on the issue. This was assumed to appeal to the Lib Dems because the AV generally favours centrist parties: very few Labour or Tory voters would cast their alternative vote for the other but large numbers of both Tories and Labour voters would give the Lib Dems their second preference vote. In particular it is assumed that AV would cement the ‘progressive alliance’ of Labour and the Lib Dems as the majority of both their voters would cast their second preference votes for the other.
In fact AV might well not do that. The heart of the matter lies in the 533 English seats where the first three places are almost universally held by the Tory, Labour and Lib Dem candidates, followed by the odd Independent and Green, UKIP and the BNP, and a Monster Raving Loony or equivalent. AV proceeds by counting all the first preference votes. Should any candidate win 50 per cent of the vote he or she wins the seat outright. But in the large majority of seats where this does not happen the bottom-most candidate is eliminated and their second preference votes distributed to the others. Given that the candidates at the rear are generally Independents with only a hundred or so votes, this would seldom settle the matter, so then the next-to-bottom candidate would have their second preferences distributed and so on, until one of the leading candidates goes over the 50 per cent line.
AV would give a couple of small advantages to the Tories: for a start, they’d pick up most of the BNP and UKIP’s second preferences, a not insignificant 5 per cent of the vote. Also, at present UKIP and the BNP exercise strong leverage on the Tories, forcing them to compete hard for the anti-EU and anti-immigration vote, but under AV they would lose that leverage since the Tories could count on their second preferences anyway. Indeed, many voters, realising that their second preference is the one that counts, might start voting for it first: in France, this is how Sarkozy has bled the Front National vote and how Mitterrand bled the Communist vote.
The big question however is what will happen in those seats where the Lib Dems run third (where they run second to the Tories or Labour one may assume that most of the third-placed Labour or Tory votes would go to the Lib Dems under AV, which is why the system is generally seen as being in their favour). You may think that most Lib Dem voters would give Labour their second preference, because of something which some politicians refer to as ‘the progressive alliance’. But this is taking too much for granted.
Most of the people in Britain who take an interest in politics are partisan, and that partisan political culture has two families, Labour and Tory. There is not a corresponding Lib Dem culture of any size: most Lib Dem voters (and even some of their MPs) are ex-Labour or ex-Tory, or come from Labour or Tory families. The Lib Dems tend to be popular with the non-political, former non-voters and first-time voters – who are often not registered anyway. These were the sort of people who inflated the Lib Dems position in the polls during the campaign but such support is notoriously ‘soft’ and unreliable.
By definition, all the ‘Other’ parties are fishing in the same pool as the LibDems – the SNP and PC, the BNP, UKIP, Greens and Independents. This is why, in Wales and Scotland, one does not generally find the Lib Dems competing with PC and the SNP for seats: once PC or the SNP have taken the third-party vote, there is little left for the Lib Dems. This is also the reason, however, that Lib Dem voters have some strange overlaps. During the rise of the National Front in the 1970s it became clear that there was a degree of interchangeability between NF and Liberal voters. When the NF did well in Leicester, for example, the Liberal vote there collapsed. The same interchangeability is today sometimes found between the Lib Dem and the BNP vote. This is emphatically not a snide way of saying Lib Dem voters are crypto-fascists, merely an observation as to the strange overlaps which inevitably exist if you’re looking to pick up votes from the pre-political, non-political and unattached.
It also makes it hard to predict what would happen to the Lib Dem vote under AV. I can’t help but suspect that the system might suit the Tories surprisingly well, because they would have the virtual certainty of harvesting BNP and UKIP second preferences which, topped up with a decent number from the Lib Dems, could be enough for them to take a fair few marginal seats from Labour.
And the Lib Dems surely know this, which is one reason their leaders have now chosen coalition with the Tories rather than a ‘progressive alliance’ with Labour. Clegg is clearly betting that his voters will not punish him for this, that they are available to be won over by a good Lib Dem performance in government. If Clegg wins his bet Cameron will have done more than just find a reliable coalition partner. He will have banished the chimera of ‘a progressive alliance’ into the shadows for the indefinite future.