Home / Dave Brockington / Tactical Voting Under the Alternative Vote (UK Edition)

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.

  • Facebook
  • Twitter
  • Google+
  • Linkedin
  • Pinterest
  • Daragh McDowell

    I’m not sure how ‘achieving maximum realistically possible electoral reform given overall political strength in parliament’ counts as a ‘sellout.’ Especially since the Tories are still grumbling about the Lib Dems wringing the Referendum out of them.

    • NonyNony

      As an outside observer, if this was the best Clegg was going to be able to get the Lib-Dems would have been better off remaining the “pure” party and not sullying their hands with an alliance with the Conservatives (and giving Lib-Dem voters a reason to think that maybe the Lib-Dems aren’t worth voting for at all, if they’re just going to enable the Tories to do what they want).

      Perhaps that’s wrong, but the alliance between Lib-Dems and Conservatives seemed to anger a good number of the Lib-Dem voters when it was made, and if it appears that the Lib-Dems made a Devil’s Bargain to get a half-measure it’s going to sour people even more.

      • Davis X. Machina

        The problem is that together with the Great Wall of China, and a few other man-made structures, Nick Clegg’s ego can also be seen from low earth orbit…

        • dave brockington

          Indeed. Which might be a reason why the LibDems are polling between 8% and 10% at the moment. They sold out.

          • dave

            Taking the view that entering the only coalition arrangement that could produce a working majority in the House of Commons after the 2010 election is ‘selling out’ requires a view of the nature of parliamentary politics so infantile that I cannot, at this precise moment, come up with a suitably scathing metaphor to describe it. Perhaps you would have preferred to watch a LabDem government attempt to survive its first year, stuffing the mouths of the devolved nations’ nationalist representatives with gold, while slashing budgets on English matters, plunging below 20% in the polls, and awaiting the first by-election defeat to trigger a Tory landslide?

    • Incontinentia Buttocks

      The larger problem than the theoretical question of whether such a compromise is a “sell-out” is the practical question of whether this compromise will work at all.

      With so many of its former supporters furious at the LibDems, my sense is that this referendum is much less likely to succeed now than in the past.

      Incidentally, Chris Bertram started an interesting Crooked Timber thread a couple weeks ago on whether the left in Britain should back AV. Here’s how he framed the question:

      Tempting though it might be, I suppose I shouldn’t decide my view on the basis of my desire to stick it to the vile Nick Clegg.

      In the ensuing conversation, it became clear that many, many people were more than happy to decide their vote on the basis of sticking it to Clegg and his party.

      Whether or not the end of AV justifies the means of the ConDem government will become a rather moot question if the end is not even achieved, especially if it fails as the result of voter anger over the ConDem government.

      But perhaps this is a distorted view from afar. What’s the polling data on the referendum look like?

  • partisan

    I was just wondering if you know where I can find a good critique of Theodore Dalrymple, an English doctor, and actually the alias of one Anthony Daniels, whose writing on the welfare state and the horrible British underclass are popular in American conservative circles.

    • dave

      A critique in what sense? The man writes from close personal experience of tending to the bodily and psychological needs of prison inmates, welfare recipients and other victims of the dependency culture over a career spanning several decades.

      What he reports is terrible, and deeply discomfiting to those who would prefer to believe that the people, released from the shackles of capitalist oppression, could at once proceed to build a shining new world.

      He has a conservative view of life, but you need no special expertise to ‘critique’ that. If you are hoping for evidence that he is a liar, I think you will be out of luck.

  • Scott P.

    I don’t understand the complaint about tactical voting. Every time I vote for someone other than myself, I am voting tactically, since I am voting for someone who represents my views worse than I do. Government itself is run on the principle of tactical voting, as can be seen by the Liberal MPs voting for Cameron for PM instead of Clegg.

    • Excellent point!

      There are voting methods which ARE 100% tactics free. But you’re not going to like it. It’s called “random ballot”, it goes like this: everyone writes down the name of one candidate, and the winner is the name on a ballot drawn at random. Presto, no tactics! If you think you’re the best choice, the best tactic you can make is to vote for yourself.

      There are, unfortunately, no deterministic (read: non-random) tactically-free election methods.

      Which is why the argument over what election method is best shouldn’t be about which is least-tactical, but which is most-effective. And yes, IRV/AV is better than plurality… but worse than every other option.


  • Captain Splendid

    Count me as one of those people that think this is very much a good thing. Worst case scenario is that a lot of people who’ve never even thought of their elections being done differently will get some education on the topic.

    Is there anyone out there doing work on the psychological aspects of different election systems? It’s one thing to go in the booth and vote for all your favourite candidates, quite another, IMO, to have to go in and rank them all in order of preference.

    • Ranked ballots (for IRV and for various forms of proportional representation) are used in several countries around the world. You can find examples written in English from the Republic of Ireland (uses single transferable vote for proportional representation) and Australia (uses single transferable vote for their senate and instant runoff for their house). Voters in those nations, generally, hate their politicians and their government as much as those of us using simple plurality.

      There was a computer simulation study done, testing the *effectiveness* of various voting methods (including IRV) by a measurement called Bayesian regret. It found that, after plurality, IRV is the worst possible election method.


  • wengler

    Actually Instant Runoff Voting(IRV) is the hobby horse that Ralph Nader and many in the Green Party have been riding for the last decade. I think in a small country with a parliamentary system like the UK, proportional voting would be better, but IRV is not nothing. This could seriously give the lib-dems a boost, their crappy alliance with the Torys aside.

    Also, can state government in this country do something else with their Senates other than make them some grouping of house districts? They seem like the perfect place to push for proportional voting.

    • Incontinentia Buttocks

      Correction: while the Green Party has long supported IRV, Nader has not.

      Nader is deeply invested in the power of “spoiling” elections, which IRV actually eliminates.

      And to answer your second question: IANA constitutional lawyer, but my understanding is that states are someone constrained by the principle of one-person-one-vote, but that a properly framed proportional system of electing state Senates might pass constitutional muster (they would also make the State Senate a much less redundant institution than it currently is in bicameral state legislatures).

      • wengler

        Nader endorses IRV in his book Crashing the Party. He also talked extensively about the two institutional parties’ stranglehold on ballot access in many states.

        Nader would’ve gotten a hell of a lot of first choice votes if IRV was around in 2000, and Gore probably wouldn’t have run to the right of Clinton.

        • Incontinentia Buttocks

          Nader has certainly talked and written extensively about ballot access reform, but I honestly don’t remember him endorsing IRV in Crashing the Party…and a quick electronic check of the book on googlebooks and amazon.com doesn’t come up with any passage in which he does (if you have a page cite, please give it).

          My memory, as a Green convention delegate in 2004, is that Nader’s camp was decidedly lukewarm on the IRV plank.

        • Scott Lemieux

          Gore probably wouldn’t have run to the right of Clinton.

          This is false. Gore’s campaign in 2000 was more liberal than either of Clinton’s. (He did distance himself from Clinton personally, but that’s a different issue.)

      • IRV most certainly does NOT eliminate the “spoiler” problem from elections; that’s what the original article was all about! (Tactical voting *IS* voting to avoid a spoiler.)

        It reduces it, from where even one single voter can cause a spoiled election, to the point where at least 25% plus one single voter are needed to spoil an election.

        The LibDems got 23% overall last election, and between 25% and 50% in a large number of constituencies.

        Switch to IRV/AV, and you’ll see still see plenty of spoiled elections.

    • DocAmazing

      IRV has worked very well in the Bay Area. By means of it, San Francisco has managed to overcome some of the money advantage of developer-backed candidates, and Oakland had a very welcome uspet victory in its mayoral race.

      • It “worked” in that no one decided to burn down the town hall in protest, so yes, it wasn’t worse than plurality… but it hasn’t really changed anything either. Third-party representation has gone DOWN on the SF board since IRV was instituted, and Quan’s win in Oakland STILL wasn’t with 50% of the votes (she only got 45% in the final tally, due to exhausted ballots.)

  • Davis X. Machina

    This could seriously give the lib-dems a boost, their crappy alliance with the Torys aside.

    Apart from that, Mr. Lincoln…

    • wengler

      Well it is somewhat meaningful that the UK has two significant left of center parties and only one nationally viable rightwing party. So the calculation of Tory leadership will be whether this will hurt them or Labour more.

  • Marc

    The real question is in which choice is the worst for the loathsome Clegg.

    • Incontinentia Buttocks

      While I certainly don’t like Clegg, I think this is absolutely the wrong question.

      But I fear that enough voters will agree with Marc that the referendum will fail.

  • MikeB

    Frankly, any subject or cause with Cleggs name attached to it in any way is now going to go down in flames. When the joke ‘Why did Nick Clegg cross the road? Because he said he wouldn’t’ becomes a standard, you know there’s a problem.

    Speaking from the viewpoint of a form Lib Dem member and activist, I cant see how AV is going to get passed. The Cons hate it, Labour really doesn’t like it, and its tainted by association with the LD’s in coalition. Even if you do support it, its no more than a half way house.

    The good thing if it fails? Then even the quisling LD parliamentary party will get the message that their screwed – and someone might show some backbone and do something other than cravenly support the Tories. The sad thing is that so many of them continue toeing the party line, which was certainly never a problem in my day!

  • TomClark

    Good post, and although I think I can defend most of the claims in my piece which you mention, there was one flaw in my reasoning. When I get a moment I will publish a revised version!

  • dave

    If AV doesn’t pass, given the other changes that the Tories are pushing on the electoral map, you can kiss goodbye to a progressive government in the UK for the next generation. The question you have to ask is, is ‘sticking it’ to an individual you’ve decided to dislike worth that outcome?

It is main inner container footer text