Would you rather...?

Glen Newey · The Referendum

Mid-April, and Britain attends to the 5 May referendum on the Alternative Vote with all the rapture of a gutted cod. Voters will be asked: ‘At present, the UK uses the “first past the post” system to to elect MPs to the House of Commons. Should the “alternative vote” system be used instead?’ This version of the question is a redraft, made at the Electoral Commission’s bidding. When the government published the original, one-sentence version last year, which rendered MPs in unabbreviated form, the commission worried that people would be too thick to understand it. Now campaigners merely worry that people are too thick to understand AV itself.

In pre-campaign polling, people were quizzed about the question. Debate raged over whether the new commission-approved wording favours a Yes response; results published in January by the pollsters ICM and Angus Reid suggested that it might. The outlier was YouGov, the anoraks’ playpen, where opinion divided 41 per cent to 32 cent against the view that the question’s phrasing was pro-AV. But this may well be because, in its polling, YouGov had inflicted on its subjects yet another version of the AV question, of no fewer than 139 words. This was always liable to reduce even hardened wonks to a state of dribbling accidie.

On 4 April the Telegraph, never one to miss the panic button when change looms, eagerly parroted the prime minister’s alarm that Britain might be ‘sleepwalking’ into AV. In fact, as recent polls and – usually a sounder guide – the bookies’ odds show, it’s far more likely that Britain is sleepwalking into more of the same old crap. Even the Telegraph had to acknowledge the next day that public ‘support’ for AV was ‘waning’. In truth, it had never waxed.

Not that, on any plausible measure of natural justice, the Yes campaign deserves to win. Their opponents’ no-change propaganda has certainly been asinine, suggesting that a range of worthies, from Afghanistan-based squaddies to neonate cardiac patients, will get hit if AV goes through. But the pro-AV case has been hobbled, to put it mildly, by the fact that its highest-profile champion is treated by his fellow Yes campaigners as the political equivalent of the Ebola virus. By contrast with the hapless deputy PM, Ed Miliband has played something of a blinder, continuing at least nominally to back reform while leaving the campaign, and its probable flop next month, as very much Nick Clegg’s baby.

Understandable schadenfreude about Clegg shouldn't blind anyone to the wider vacuity of the Yes campaign. Nobody has come up with a decent answer to the question of why a voter’s (n+1)th preference, where n may be rather large, should be able to count for as much as others’ first preference. The psephologically null concept of the 'wasted' vote has been dragged out like an ageing debutante for a final coquettish gallop. And the fallacy has been put about that AV gets rid of tactical voting. Not really. If my ranking of the candidates is A > B > C, my overriding aim is to stop C at all costs, and I’m confident that A will garner enough first-preference votes to survive the first round, I have reason to put B as my first preference. Not that there’s anything obviously wrong with tactical voting, anyway – why is it better to vote for the candidate you like most than against the one you like least?

No doubt all these issues will be revisited in the fervid intellectual debate that beckons over the coming weeks. If the issues pall, there are always celebrities to fall back on. Can’t be fagged to grasp the niceties of instant run-off ballots? No matter. Who would you sooner have a jar with: Eddie Izzard or Norman Tebbitt? Joanna Lumley or Jacob Rees-Mogg? Helena Bonham-Carter or Nick Griffin? Then you know how to vote. Or not.


  • 11 April 2011 at 8:02pm
    Joe Morison says:
    The argument from celebrity endorsement has never seemed so strong.

  • 12 April 2011 at 4:37pm
    Geoff Roberts says:
    It all sounds like a very bland sop to placate the liberals. The arguments for a change in the system (outdated by a century) are too great to be wasted on this circus act.

  • 13 April 2011 at 1:06pm
    semitone says:
    "Nobody has come up with a decent answer to the question of why, with a slate of n candidates, where n may be rather large, a voter’s (n-1)th preference should be able to count for as much as others’ first preference."

    Well, Glen, here's one. Is the point of an election to get the most popular candidate (really with FPTP it's the least unpopular, I think, given that it's possible to win such an election with much less than half the votes cast)? Or is it to represent, as closely as practicable, the general will of the voting population? That's right, it's the second one.

    At the moment, anyone not feeling well-represented by a major party (and that's most of us here, I think) either votes for a minor party, votes for a party they don't like much, or doesn't vote at all. Under AV they get to vote for the party that represents them best, but can still have a say in which of the two or three contenders for the seat actually wins.

    They're not forced to have a say; they have a choice. You can still put "1" in the box next to your favourite, and leave the rest blank. Opponents of AV should be invited to do just this; if the referendum gets up, I wonder how many will.

    • 15 April 2011 at 11:06pm
      Phil Edwards says: @ semitone
      AV doesn't come close to representing the general will of the voting population, because it imposes an artificial single-member constraint on the proportional system of STV. Under STV the 'quota' - the share of votes required to gain a seat - is one vote higher than 1 / n+1, where n is the number of members being elected; what this means is that, in a three-member seat, the highest number of voters who can be totally unrepresented is 1/4. Under AV, with single-member constituencies, the 'quota' is set at 50% of the votes cast plus one vote. Which, as night follows day, means that anything up to 50% minus one vote have no influence on the result.

      Now, the difference between winning a seat on (say) 35% of the vote and a minimum of 50% isn't nothing, but it's bought at the cost of uprating second and third preferences, which in the nature of things are likely to be less strongly-held and more opportunistic than first preferences. Under STV this doesn't have much of a distorting effect, as most people whose second preferences are counted will have had their first choice discounted on the grounds that the candidate has already made his/her quota - in effect they will have got their first choice anyway. This clearly won't be the case under AV. And, given the woeful lack of public education on the system, we're likely to see a lot of people automatically ranking the candidates in order of preference, not realising that the question they're being asked is "do you want your vote to count for these candidates?" It's likely to produce a massive inadvertent tactical vote, in other words.

      AV is not PR: on a national level, AV is no more likely to produce a proportional result than FPTP. Supporters of PR should bite the bullet and make common cause with the know-nothings of the No campaign.

    • 19 April 2011 at 6:12pm
      semitone says: @ Phil Edwards
      As good an argument for PR as I've heard, though of course lots is left out. But just because AV is no more likely to produce a proportional result than FPTP (see what you did there?) doesn't make it less likely to produce a representative result.

      Which is the point.

      Anyway Alex it sounds like you're talking about something a bit like the Hare-Clark system, delivering stable government to Tasmania ... most of the time. It's PR, but it keeps the consitituency link. Actually I think the UK could use a big fat dose of Hare-Clark. The next thing we know we'll be electing the House of Review, and suddenly it'll feel like democracy.

    • 19 April 2011 at 6:26pm
      semitone says: @ semitone
      oops sorry, meant to say Phil not Alex. But I think this conversation has lapsed anyway.

    • 20 April 2011 at 11:20am
      Phil Edwards says: @ semitone
      Well, I'm still here. I think (or Wikipedia thinks) that Hare-Clark is what I call STV, so yes, that is what I'm talking about. AMS would be OK, too, although it does seem to confuse people (you get two votes, one FPTP and one for a regional 'top-up' list).

      As for AV producing less representative results than FPTP, I think it would tend to overstate support for runner-up parties - and bland technocratic runner-up parties in particular - for the reasons I gave above.

    • 21 April 2011 at 10:08am
      semitone says: @ Phil Edwards
      I wonder if there's any research on whether STV works best for small consitituencies. It does OK for Tasmania I think, and that's a very polarised state (especially over environmental issues) and they used it in the A.C.T. (that's Canberra and environs, for you non-Australians still reading).

      I'm voting yes because I think AV is better than FPTP, not because I think it's the best voting system. I'd love to be able to propose my favourite model then vote for it, but if we all wait for that to happen then nothing will change.

      One way the current proposed AV is better than Australia's is that it doesn't force you to number all the boxes. So support for bland technocratic parties (a description that seems positively made for the Australian Labor Party) won't be overstated; people will preference them if they like them, and won't preference them if they don't. In Australia I vote Greens 1 Labor 2, and have done since I was old enough to vote. That doesn't mean I support Labor, just that I want to be able to vote for the Greens without risking "splitting" the left-ish vote so that the dreaded Liberals get in. AV lets me do that, FPTP doesn't. Under FPTP I'd consider voting for the bland technocratic party that doesn't represent me, just to make sure I wasn't saddled with the evil technocratic party I abhore and fear. In this scenario (and I can't be alone in this), it's FPTP that overstates support for centrist bland technocracy, while AV allows people to protest against such blandness but also oppose the dangerous extremism of the other major party.

      UK politics ain't so different, I think.

  • 15 April 2011 at 1:42pm
    alex says:
    If the concept of a wasted vote is psephologically null, what kind of null is the concept of an ageing debutante? Or, since we've started, that of a coquettish gallop (especially when performed by the aforementioned oxymoronic agent)?
    Our political situation is shit enough as it is: hurling mixed metaphors at it doesn't help much.

    • 18 April 2011 at 5:32pm
      Glen Newey says: @ alex
      It's from Tolstoy, so blame him. I don't see why the notion of an ageing debutante is oxymoronic. Maybe she just didn't get out enough earlier in life.

    • 19 April 2011 at 8:41pm
      alex says: @ Glen Newey
      A debutante is customarily on the threshold of adulthood. After her debut she is a former debutante. So tell that to Lev Nikolaevich (or his translator) when you next see him.

    • 20 April 2011 at 11:22am
      Phil Edwards says: @ alex
      "Ageing" qualifies "debutante" - it implies the woman in question is older than usual for a debutante, or perhaps that she's been unsuccessfully launched on society once already.

    • 20 April 2011 at 8:31pm
      alex says: @ Phil Edwards
      In the latter case - and that's the context implied by 'dragged back' - the debutante's debutantiality seems to be not so much qualified as nullified.

Read more