Constitutional Non-Moment

Glen Newey on the Referendum

A.J.P. Taylor memorably remarks in The Course of German History, written to mark the centenary of the failed 1848-49 revolution, that with the Frankfurt Parliament, ‘German history reached its turning-point, and failed to turn’. A similar non-moment in the UK’s constitutional history now lies before us, in the form of the alternative vote referendum on the voting system.

The quality of the AV ‘debate’ has been staggeringly low. Some contentions, particularly on the ‘No’ side, have been risible, such as its ‘dead babies’ case against electoral change. Those of their opponents have been little better: for example, that in some elections AV would have given the result they wanted, whereas the (misnamed; see the excellent discussion by Tim Gowers) ‘first past the post’ system wouldn’t. It’s hard not to regret that there has to be a winner.

Beyond the plastic zoo of politicos and celebs, the standard of argument has not been much better. It’s often assumed that the popular will is somehow out there, and that the task is to design a suitable butterfly net in which to catch it. The point is not just that people often tweak the way they vote to fit the system; there are different formal and political criteria for aggregating individual votes. So AV has been charged, correctly, with breaching formal criteria like monotonicity – which says that a candidate should not be disadvantaged by being promoted on some ballots, if the other rankings are held constant. ‘First past the post’ is monotonic, but so are other methods, such as the Borda count. Failure of monotonicity is mildly counterintuitive. But there is still a debate to be had about how heavily these factors weigh against such political desiderata as representativeness and ‘strong government’. We certainly haven’t had it over the past few weeks.

The lasting fall-out from the 5 May elections will be in Scotland, where the SNP has romped home: the Nationalists may well get the 65 seats they need for an overall majority at Holyrood. Unlike in Wales, where it is about to return to power, the Labour Party in Scotland has been hobbled by its longstanding reputation for cronyism in local government and having its haggis shot by Ed Miliband, who suggested on a brief campaign stopover that the Scottish parliamentary elections were a prelude to the Big One in Westminster in 2015. Meanwhile, the Nats’ triumph testifies to the political uses of vacuity. On its website the SNP styles itself ‘Scotland’s moderate left-of-centre pro-independence party led by Alex Salmond’. At least one element of that description is beyond dispute. On the party’s manifesto site, even the saltire has been ousted by a shot of Salmond’s jowls. He’s easily edged it over rival party bosses Tavish Scott and Iain Gray, who muster all the charm and visibility of an underground car park.

Since 2007 Shrek has run his minority administration with a rubberoid elasticity of principle. He’s made noises against Trident in the knowledge that this gratifies the nationalist constituency but lies beyond the executive’s devolved powers. Meanwhile, he’s cosied up to Donald Trump and presided over regressive fiscal policies and public spending cuts that have been voted through Holyrood each year with the help of Annabel Goldie’s Scottish Tories. You have to search hard on the SNP manifesto site to find any mention even of independence, though they go large on wind farms, presumably in case they need Green support in a coalition. The Tories are more natural allies. Salmond funked the independence referendum promised in the last election, and it’s not mentioned in the current manifesto. But if he has a majority in the new parliament, he can hardly not put it to the vote. The Tories are of course officially unionist, but the chance to offload the regular 58-odd non-Conservative MPs returned by Scotland at each UK general election must be tempting. In 1848-49 the story of reform’s failure, as Engels noted, was the impressive co-ordination of pan-German reaction to stifle revolt. In this year’s farcical repeat, the role of Bismarck will be played by the genial green man with trumpet ears.


  • 6 May 2011 at 3:15pm
    alex says:
    Interesting piece, but Taylor's work wasn't written to mark the centenary of the failed 1848-49 revolution, being published in 1945.

  • 6 May 2011 at 3:45pm
    Charles Addison says:
    Shreck? Green rubberoid man with trumpet ears?
    Glad to see your cutting analysis and keen argumentation didn't require you to resort to ad hominem attacks.
    Clever you!

    More metropolitan elitism please!

  • 6 May 2011 at 9:03pm
    Robin Durie says:
    If the "quality of the AV ‘debate’ has been staggeringly low," what does this say about the quality of Newey's "analysis" of the Scottish election?

    The only thing more shabby than the sub-tabloid abuse of Salmond is the utter paucity of Newey's understanding of, & "analysis" of, Scottish politics.

    Poor effort, Glen.

    Really poor.

  • 8 May 2011 at 5:39pm
    Richard Wyn Jones says:
    Having read this 'analysis' of the Scottish result I'm hoping that someone might remind Newey that he's writing for the LRB blog rather than the Daily Mail...

  • 9 May 2011 at 8:12am
    SeamusDolphin says:
    One can, of course, be too po-faced about these things.

  • 10 May 2011 at 9:49am
    semitone says:
    "The quality of the AV ‘debate’ has been staggeringly low ... It’s hard not to regret that there has to be a winner."

    This is exactly the kind of political comment (let's not call it analysis, because it aint) that I hate. The idea of a democracy, Glen, is to vote for policy outcomes you want, or for the candidate you like most (or fear least). You don't vote to reward the campaign you thought was better run. If you really don't care whether the UK uses FPTP or AV, write an article saying so and explaining why the counting system doesn't make any difference (good luck with that).

    The rest of us have a view on which is better, and we voted accordingly. Please, please, can't the LRB blog do better than this?

  • 11 May 2011 at 7:58am
    Glen Newey says:
    Well, you can’t please all the people all the time. One or two distinctions that seem to have been missed in the above comments:

    A blog, whether in the LRB or elsewhere, doesn’t have to be given over to “analysis”. Commenters have scare-quoted the word, as if I had described my 5 May contribution in these terms. It is perforce not an analysis of the election results, because it was written before the elections took place. Like most blogs, this one is a chronicle of comment on current events, not an article in the Journal of Election Studies. I assume the LRB editors who commissioned and accepted it as fitting what I take to be a brief to produce topical comment that is by turns provocative, flippant, informative – or, on occasion, analytical.

    I bemoaned the quality of the debate over AV and suggested it was a shame that one side had to win. At the risk of stating the obvious, there is a difference between taking sides on a debate and commenting on the quality of arguments of the debaters. Again fairly obviously, to say or imply that neither side has given decisive reasons for its view is not to refuse to take a view oneself, still less to say that it doesn’t matter what view anyone takes. As it happened I didn’t have a vote on 5 May, but for whatever interest it has, I would have cast it against AV. I gave some arguments against AV in a blog on 11 April.

    One problem mentioned there was that in a slate of n+1 candidates, it gives one elector’s nth preference as much weight as another 1st preference. A better system is a version of the Borda count, such as the quota preference score method devised by Michael Dummett, or a system allocating a fixed quantum of k points to each voter, to be approportioned among the candidates at the voter’s whim. If k is made large enough, say k=100, this obviates the insensitivity to measures of preference strength in pure ordinality systems like AV, and (for those who care about such things) also preserves monotonicity. It doesn’t obviate tactical voting, but as I argued in the 11 April blog, nor did AV, and more important it’s not obvious that doing so is an asset in an electoral method. The wider point remains that desiderata like equality and representativeness may conflict with each other and with political values like efficiency, and there is no good reason to think that some one system will prove to be, from the perspective of all of them, optimific.

Read more