Tactical Voting Under the Alternative Vote (UK Edition)

The UK is scheduled to hold a referendum on May 5 to determine whether or not to adopt the Alternative Vote for Parliamentary elections.  This is the first UK-wide referendum since 1975, and to my knowledge (ergo my students’ knowledge) only the second referendum to apply to the entire United Kingdom.

I still hold the opinion that this was a sell out on the part of Nick Clegg and the Liberal Democrats.  AV is not proportional representation.  While it’s also not a plurality system, it is in the end a majoritarian system.  For those unfamiliar with how AV works (and I can’t imagine that’s a particularly large percentage of the LGM readership), the blog over at the LSE has an excellent “simple guide” to electoral systems.

While AV is not PR, it does greatly diminish the incentive for a voter to vote tactically.  Of course, this incentive doesn’t entirely disappear, as Tom Clark writes at The Guardian.  However, Clark severely overstates his case, and seems to confuse the empirical presence of “wasted votes” with the behavior of tactical voting.  While AV does not ensure 100% sincere voting as a residue of incentive to vote tactically is left behind, his argument implies that this residue will translate into tactical voting in those contexts where it is present.  I don’t think this will be the case.

Tactical voting requires that a voter possess  a certain level of information about the choice set, and the relative standing of the various options.  The current configuration of Westminster elections is perhaps the most fertile ground for tactical voting: single member constituencies, only two parties with a realistic chance of forming a government, and the combination of a nationally strong third party (for now) with pockets of strength for regional nationalist parties (e.g. Plaid Cymru, the SNP).  Under AV, most of this goes away.  The vast majority of choice sets requiring a tactical calculation on the part of the voter are Conservative-Labour, Conservative-LibDem, or LibDem-Labour marginals.  Yet the following is typical of the examples offered by Clark:

Third-placed Labour supporters in Dwyfor Meirionnydd would very probably have suffered a similar fate in 2010. Chances are Ukip, independent and Lib Dem transfers would have handed Plaid Cymru a majority before they had any say.

This is “very probably” empirical reality, but for tactical voting to actually occur in this constituency under AV, it would require Labour supporters to have near perfect information regarding the chances of each of the six candidates running for the seat: Plaid Cymru, Conservative, Labour, Liberal Democrat, Independent, and UKIP, as well as reasonably solid information regarding the second choices of their fellow electors in this constituency, in order for these supporters to draw the conclusion that neither their first choice would matter, nor would any subsequent choices be counted: a classic wasted vote.  Such an assumption regarding the level, reliability, and validity of information is staggering.  Yet under the old rules, 14% of the electorate in this constituency still voted for Labour, presumably a sincere vote, when information should have been relatively cheap to acquire indicating that the seat was solidly Plaid (at 44%) and the Tories did finish second at 22%.

While AV will not completely eliminate the incentives for tactical voting, it does greatly diminish them.

I’ll have more to say about this in the near future, but it’s been a magnificently busy academic year for me and now I must rush off for another responsibility.  However, the end to the madness is in sight: term is over in a few weeks, and by the end of April two fresh conference papers will have both been written and presented.

Ideally, in that order.