Why Cameron would be smart to follow Fink's advice on AV

Danny Finkelstein wrote a Times piece this morning where he suggested that David Cameron could use a change in the electoral system to AV to his advantage. Because of the paywall, there's little point in me linking to The Fink's piece directly but James Forsyth on the Spectator blog has summarised his idea:

So Danny Finkelstein’s blog this morning suggesting that ‘AV might provide the answer to the otherwise impossible question - if the parties stay together, how can they fight the election apart?’ has caused quite a stir.

The argument is that the Tories would urge their voters to put the Lib Dems as second choice and vice-versa. If this ploy worked — and the Australian evidence Danny cites suggests it would — then AV would hurt Labour not the Tories.

I wondered how the most recent election might have played out had this sort of game been in play. Obviously some serious academic analysis would be needed in order to do this properly but I thought I'd have a stab at a crude way of discerning what could have happened if all voters for each of the coalition partners had gone along with this under an AV system.

I took a spreadsheet that I received a few weeks ago which contains the voting figures for every party in every seat from the general election last month. I then went through it and created a new "Coalition" column which contained the total combined figures for Conservative and Lib Dem votes in every seat. I then made the assumption that whichever of the two coalition parties was in the lead out of the two of them in the seat would receive the other's votes to get a total hypothetical figure for the "leading" coalition partner after transfers.

Before I reveal the figures, some caveats. Firstly in reality all voters are not going to go along with this plan even if both parties in the coalition try really hard to persuade them to do so (which in itself is questionable). Secondly under different electoral systems people vote differently so if 2010 had been run under AV the numbers would have been different anyway. Also, my analysis does not take into account how other transfers from e.g. nationalists, Greens, UKIP etc. to other parties may have worked out. It simply focuses on the coalition partners.

Effectively what this method is doing is taking The Fink's idea to its extreme assuming a perfect set of transfers for the coalition partners from each other in 2010.

In the actual election the figures were:

Conservative: 306

Labour: 258

Lib Dem: 57

In my reworked extreme AV scenario the figures would be:

Conservative: 394

Labour: 141

Lib Dem: 91

So the Conservatives would have a thumping majority. The Lib Dems would have done much better but ironically ended up with much less influence. Labour would be the big losers with just over half of the number of seats they actually got.

Despite all the problems with this analysis I think it was worth doing because it demonstrates that it would be possible for the Conservatives to benefit from AV. In an actual AV election I would not expect it to end up anything like what we see above but I think a good case can be made that a fair number of voters would be willing to put the partner of the coalition party that they put in first place in second assuming that the coalition is reasonably popular in 5 years time. It is certainly true to say that all those certainties that so many on all sides once thought about how the Lib Dems and Labour would be likely to be the automatic beneficiaries of an AV system are not necessarily true any more in the new political reality.

If Cameron could get even just a slice of the sort of action we see in play above then it could make a change to the electoral system a much easier sell for his party. Who knows, out of self interest some of them may even feel able to campaign for it, or at the very least not campaign against it!

This is another example of how everything changed on May 6th.