Politics

AV, extremism and the median voter

Much of the debate between advocates of #Yes2AV and #No2AV is not exactly edifying, particularly some of the tactics employed by the latter group. The quality of some of the debate is pretty feeble. Yesterday’s Question Time was a classic example of an occasion where the topic was raised but the discussion generated more heat than light.

One of the issues that’s attracted a lot of attention is whether a move from First Past The Post to the Alternative Vote will benefit extremists or not. The #No2AV campaign is adamant that it will, while the #Yes2AV campaign is equally certain that the change will have the opposite effect. That the BNP are not in favour of AV is taken as an indication that it is less likely to assist extremists. As I understand it, the BNP’s position is based on a preference for a more proportional system. That at least makes some sense – I’m not sure that either AV or FPTP really offers them a great deal.

This got me thinking about the venerable idea associated with the median voter theorem, which was popularised by Anthony Downs in the 1950s.

The median voter theorem starts from the idea that the political preferences of an electorate can be represented on a single spectrum – let’s call it “left to right” – and are single peaked – that is, there are a few extremists on both the left and the right but most voters’ views are in the political centre ground. If a politician has to win at least 50% of the vote to be elected, then the theory predicts that for politicians to be successful they have to place themselves close to the middle of the political spectrum. A centrist politician marginally to the left can assume that everyone who is politically to their left will vote for them as the candidate with views that most closely match their own. The politician then only has to capture enough of the marginal voters of the centre/centre-right to get elected. The arguments for a centre-right politician are symmetrical. The main prediction is therefore that those parties with any chance of success will be difficult to distinguish as they fight across the centre ground over political positions embodying minimal differences.

Clearly the theorem is a simplification. In practice, political positions cannot easily be captured on a single dimension, they are frequently multi-dimensional. Liberal Democrats know this only too well. But the theorem is seen as having some power in explaining US presidential elections, for example.

I think it also has relevance to the AV discussion. Under the FPTP system in some constituencies it is possible to get elected by a proportion of the electorate substantially below 50%. The median voter theorem has limited traction.

But in an AV world where the successful candidate needs to secure a positive preference from 50% of the voters it starts to become more relevant. Politicians will be encouraged to offer a policy platform that appeals to the median voter.

Clearly, in Tory or Labour strongholds it will be possible to pitch for voters well away from the political centre because the median voter is positioned towards one or other end of the political spectrum.

But in constituencies with a wider spectrum of views the median voter is going to be somewhere nearer the centre ground, albeit a centre ground that has shifted to the right over time. To secure sufficient support – if only in the second or subsequent round of the count – a politician is going to have to adopt a political position that will appeal to a broader base. It is therefore unlikely that AV will lead to the election of extreme candidates. The AV mechanism does mean that the second preferences of voters whose first preference is extremist will count towards the final result, rather than being symbolic and ‘wasted’, but they are not likely to determine the outcome.

With the possible exception of the “Blue Book” mooted by David Davis, stories circulating in the press at the moment of “continuity Labour”, the takeover of the government by “Orange bookers”, the increasing influence of Progress within the Labour party and the Social Liberal Forum in the Liberal Democrats, suggest the political compass is currently spinning quite violently. The various parties are mostly jostling for position in the centre ground. Distinguishing which party ‘owns’ some of the ideas being discussed is rather difficult at times. Precisely what the median voter theorem would predict. But not much consolation for the extremist. And a red herring from the #No2Av camp.

Print Friendly, PDF & Email