Friday, 18 February 2011

Why the "Vote #No2AV to spite Nick Clegg" meme is 100% mistaken

There is an obvious reason for being unimpressed with the current efforts of the 'No To AV' side in the referendum campaign to turn the referendum into a popularity contest concerning the LibDems in general and Nick Clegg in particular (see e.g. http://www.ft.com/cms/s/0/3cd7b944-3937-11e0-97ca-00144feabdc0.html & http://www.dailymail.co.uk/news/article-1357379/Alternative-Vote-Nick-Clegg-power-years.html ). Namely, that it is the most short-sighted and dirty form of personality politics to try to get people to vote against a seminal proposal for constitutional change -- the reform of our electoral system, the very heart of our democracy – so as merely to be able to give one in the eye to our unpopular Deputy Press Minister.

But there is also a much less obvious – a hitherto-unnoticed -- reason for being unimpressed with these efforts, and with the rash of media stories that continue to accompany them (see e.g. http://www.bbc.co.uk/news/uk-politics-12462662 ). Namely, that there is good reason to believe that it is simply false to claim that AV will benefit the LibDems.

That may seem a very surprising claim. For everyone, from the BBC and Peter Kellner (see http://broadleftblogging.wordpress.com/2011/01/30/av-is-a-game-changer/ ) to Nick Clegg himself, is assuming that the LibDems will benefit from AV. It was an assumed past of the script in the recent Newsnight special on AV (http://www.bbc.co.uk/blogs/newsnight/fromthewebteam/2011/02/wednesday_16_april_2011.html ). And after all, haven't the LibDems in the past suffered a good deal from the 'wasted vote' argument, which AV would put an end to? Wouldn't they get lots of 2nd preference transfers from Tory and Labour voters, thus increasing their haul at the next election, or at least stopping the rot that their haemorrhaging support is creating?

There are two reasons why these assumptions are mistaken:

1) In general terms, First Past the Post (FPTP) suppresses the vote of LibDems in places where they are weak (where they are perceived as a 'wasted vote') – but it artificially bolsters their vote in places where they are already strong (where they benefit from tactical voting). What does this mean? It means that introducing AV would (ceteris paribus) increase the 1st preference LibDem vote in places where they are weak – places where they have no chance of winning – and decrease their 1st preference vote in places where they are strong – where at present they benefit from tactical votes, without which (in many cases) they would not win. ...They might even score a higher vote tally overall, but its distribution would work against them winning (or even holding) seats. Their vote-share would be up in places where it was useless to them, and down in places where it would have been useful to them. Ceteris paribus, AV will cost the LibDems seats.

2) But of course, in the specificities of our political situation under this Coalition government, ceteris is not paribus. For the first time ever, there will be large-scale deliberate anti-LibDem voting, at the next General Election. There will for instance be many Labour (and Green, and Nationalist, etc.) 1st preference voters who quite deliberately do not put the LibDems down as their 2nd or 3rd or even 4th preference…because they are so disgusted with the LibDems' betrayals. (Many may well rather place Tory ahead of LibDem in their preference-ordering, for the Tories at least have not gained a reputation as turncoats and opportunists.) People looking to give Nick Clegg one in the eye at the next General Election will have ample opportunity to do so, under AV. They can deliberately put the LibDems bottom of their preference-list (or simply leave them off it altogether).

The key point here is that, relative to AV, it is FPTP that maximises the seats that are gained by unpopular parties, parties that are seriously disliked by a majority. Because AV requires that you assemble a majority of voters' preferences in order to win. This is demonstrable for example in relation to Council elections in this country: there are many seats that the BNP have won under the present system that they would without doubt have lost under AV: for the second and third and fourth preferences of voters voting for mainstream/non-fascist parties would in very many cases have transferred against the BNP. In seats where it is not obvious who to vote for in order to stop the BNP, FPTP is the system of choice for the BNP. Which may well partly explain why the BNP, somewhat understandably, is calling the AV referendum a conspiracy against the BNP... (See http://isupportav.co.uk/2010/11/the-bnp-are-campaigning-against-av/ )

The LibDems now may start to suffer the same fate… They presumably did not anticipate, when they put their eggs in the AV basket, that they would end up not merely losing votes but losing second and even third preferences too, because of becoming objects of such dislike that Cleggmania has turned into the search (among the #No2AV forces) for how to turn the AV referendum into a referendum against Clegg… What is interesting is that the LibDems still don't seem to have woken up to this fact. The fact that, if they are widely disliked, they will be even worse off under AV than under FPTP. (One's surmise is that they haven't woken up to it, because to do so would be simply too painful – as they would then realise that they have boxed themselves into a truly terrible corner, in this referendum campaign: If they lose it, they will be perceived to have lost the one thing that they really wanted out of the Coalition; but if they win it, they will actually suffer at the next election, as a consequence…)

But what of the academic / polling studies that also tell us that AV will benefit the LibDems? Some of these studies, rashly, simply ignore the second point that I have made! They work, that is, on the basis of what voters' second preferences were last May (thus ignoring that the LibDems will now be garnering many less second etc. preferences, and will garner even less, probably, by 2015 – this is to miss my point (2)). Some studies assume that first preferences will remain the same under AV, thus simply ignoring the first point that I have made! And even the best of the studies that I am aware of (using British Election Survey data: http://www.bbc.co.uk/blogs/markdarcy/2010/10/av_study_reveals_some_surprise.html http://www.oxfordjournals.org/our_journals/parlij/gsq042.pdf), which ostensibly at least takes into account my point (1), does not go far enough in undermining the assumption that those who voted (say) LibDem at the recent General Election actually do have LibDem as their 1st preference, that those who voted Labour actually do have Labour as their 1st preference, etc. etc. . Because they do not take into account that some voters who are taking part in an FPTP election will in effect have been primed by the circumstances in which they are answering the survey questions in most cases into thinking that their first preference should remain the same, and that AV will only affect their lower preferences. (Or at least, that many voters will not have given any deep thought to clear 1-n pure preference orderings beforehand, orderings unaffected by questions of who is likely to win, because they never needed to.) That assumption, as we have seen, is much worse than rash -- it is manifestly false. It is falsified by the existence of large-scale tactical voting, under FPTP. When people actually get to understand AV and to think of it as their actual voting system, then first-preference-votes change more, and continue to change over time.

A case-study for this is what has happened in Australia. There, AV has allowed smaller Parties (and independents) that are not thoroughly disliked by a majority to build up their votes. This is how the Green Party vote has grown in Australia, to the point where the Greens have won a number of seats in the Upper House (elected by PR) through being able to build up their credibility through scoring decent numbers of first-preference votes (through AV) in the Lower House (and thus scoring well enough to be taken seriously, in opinion polls as well as actual elections). And the Aussie Greens have now won their first seat in the Lower House, through second-preference-transfers under AV…

Meanwhile, the closest equivalent to the LibDems in Australia, the 'Democrats', declined under AV, to the point that they have now collapsed altogether…

The key point that one has to understand is that AV changes people's ability to expresses their preferences, and (most crucially – this is my point (1) above) that AV can change quite drastically how people vote ON FIRST PREFERENCES too. This point would quite possibly help the LibDems under PR -- but it will hurt rather than help them under AV. In some cases (as has happened in Australia) it will lead to them being overtaken on 1st preferences by what are currently the 'minor' Parties. And this effect will increase over time.

The big question about the effect of AV on election results then is how the abolition of tactical-voting and of 'wasted vote' arguments (an abolition that AV very largely, thankfully, effects) and the drastic reduction in safe seats that it will simultaneously bring about will (over time) affect the first-preference votes of the LibDems -- and (differentially) of the various smaller Parties. This question, the 'experts' and the LibDems themselves seem largely to have ignored. Most unwisely…

The conclusion, then, is clear. If you are one of those people who is worried about voting for AV because you don't want to do the LibDems a favour, then you need worry no longer…

It is the Right (including the worst elements of tribal Old Labour) who are campaigning against AV, and pouring money into the opposition to it. Anything that is seen as bad by the dog-ears of Labour and by the Tories and the Taxpayers Alliance and the BNP surely has to be a good thing… If the AV referendum passes, these people will be spitting blood. But the unexpected conclusion of my analysis here is that passing AV will in reality be one in the eye for Nick Clegg, too… For when it actually comes to the 2015 election, Clegg's Party will, for the reasons I have explained, lose seats, not gain them, because of AV. In sum: It is time to dump the 'Vote #No2AV if you want to give Clegg one in the eye' meme, not only because it is puerile and small-minded, but also for the even more important reason that it is simply 100% factually mistaken.

[An earlier version of this piece has been published in slimmed down form at http://liberalconspiracy.org/2011/02/18/two-reasons-why-libdems-might-not-benefit-from-av ]

1 comment:

Amanda said...

Nice post, Rupert. :-)