Should David Cameron regret campaigning against AV?

In May 2011, as part of the deal that brought together the coalition government, the UK was asked if it wanted to change its electoral process to use the Alternate Vote system as opposed to the current First Past The Post.  Recall that the AV system involves voters ranking the candidates, and essentially eliminates the bottom candidates and redistributes their votes until one candidate gets above 50% of the total.  For a number of reasons, some based on merit and some based on other things, the idea was rejected by a margin of more than two to one.

In the lead-up to the referendum, the Conservatives were campaigning to keep FPTP, while the Liberal Democrats were the driving force behind AV.  However, over the last year, we see that UKIP has been taking a considerable share of the votes away from the Conservatives.  So was opposing AV a mistake on the part of the Tories?

Let's first go back in time and look at where the polls were sitting in May 2011.  At this stage of the parliament, voters were starting to reject the government and its austerity measures.  The Liberal Democrats were being particularly punished for turning their backs on their tuition pledge.  At the time, the opinion polls pointed towards a working Labour majority in the neighbourhood of around 56 using the First Past The Post system.

There was a lot of argument about whether moving to AV would have had an effect.  Now I looked long and hard to find data on preferences, but the only thing I've been able to find is the second preferences for the London Mayoral election last year.  For lack of better data, I've applied these preferences to come up with an AV estimate.

Among the highlights of this preference data:

• Those who voted CON as their first preference voted LD second around 30% of the time, GRN around 17%, and UKIP around 16%.
• Those who voted LAB as their first preference voted GRN second around 37% of the time, LD around 23%, and CON around 19%.
• Those who voted LD as their first preference voted about evenly between LAB and CON as their second choice (31% and 30% respectively), with GRN around 21%.
• Those who voted GRN as their first preference voted LAB second around 52% of the time (though this is largely due to a direct endorsement), with CON and LD each around 15%.
• Those who voted UKIP as their first preference voted CON second around 39% of the time, with BNP around 19% and GRN and LAB each around 12%.

Using these preferences, here's roughly how the projections in 2011 would be different under AV:

We'd see a handful of very close marginals move to the Conservatives and the LibDems, but not really enough to say any substantial difference had been made.

Fast forward now to 2013.  UKIP has reached record levels of support, coming second in some high-profile by-elections and gaining hundreds of seats in last week's local council elections.  Opinion polls generally have the Conservatives lower with Labour holding a majority of around 84 under FPTP.

In the last two years, using First Past The Post, the Tories have lost 27 seats, split around evenly between Labour and the Lib Dems (gaining 14 and 13 respectively).  However, much of these seats are gained back when we stick it through the AV calculator.

Assuming the preferences are uniform across the country, much more UKIP's preferences would go to the Conservatives than they would elsewhere.  This would give the Tories 28 more seats under AV using today's projections than they'd have under First Past The Post, mostly at the expense of Labour.  In terms of the change since 2011, the Conservatives will only have lost six seats, much fewer than the 27 lost under FPTP over the past two years.

Let's just make one thing clear, though.  Even despite this, Labour still finishes with a majority of 44 under this model.  AV would not be enough to keep David Cameron in Downing Street.  But it would be enough to protect a considerable number of Conservative MPs from a UKIP surge which ultimately leads to a Labour seat gain.