Two Cheers for AV
TWO CHEERS FOR AV
By Professor Dennis Leech, Department of Economics
It has been nearly a year since the last General Election and the coalition government was formed, but what of some of the policies that were promised? Professor Dennis Leech explains further the idea of the 'Alternative Vote' (AV) system, and highlights the positive and negative impacts it may have on some political parties.
It seems to be taken for granted by a lot of commentators that changing to the Alternative Vote (AV) will be a move towards Proportional Representation - resulting in more hung parliaments and coalition governments.
It is said that the Liberal Democrats will do well out of the change because their candidates are normally the second choice of Labour and Tory voters and therefore, since AV counts second preferences, more of them will be elected. So, while we all acknowledge that AV is not strictly proportional, it is at least a compromise that will get us part way there. If the last election could be re-run under AV, the LibDems would improve on the 8.8 per cent of seats they won, while not achieving the full 23.3 per cent their share of the poll would warrant.
This is a profoundly mistaken view. In fact, careful analysis shows that a parliament elected under AV will not necessarily be any more proportional than one elected under the present 'First Past The Post' (FPTP) system. The reason is that MPs will still be elected separately in each constituency by a voting procedure that takes no account of the national party results.
So, while we all acknowledge that AV is not strictly proportional, it is at least a compromise that will get us part way there.
But AV would undoubtedly be an improvement on FPTP – which is just about the worst election method ever devised, because it does not require that the winner gain a majority. The winner can be elected on the votes of a determined minority of committed supporters even though he or she might be intensely disliked by the vast majority of voters. AV avoids this problem by requiring the winner to have a majority, if not on first preferences alone, then once the second, third and so on of the weaker candidates have been counted after they are eliminated.
FPTP has allowed extremist candidates to win local government seats in some areas. For example, the BNP candidate who was elected to represent a Burnley ward with only just over 30 per cent of the votes because of a three-way split among the three main parties. A majority of voters would probably have preferred any one of the main parties to the actual winner.
We saw how AV works, very graphically, in the French presidential election of 2002, which was conducted under a runoff system somewhat similar to AV. The runoff was between the right wing Chirac and the extreme rightwinger Le Pen (who had the support of a fanatical minority but was detested by the majority). The result was an apparent 80 per cent landslide win for Chirac. But Jospin, who was probably preferred to any other candidate by a majority of voters, and therefore would have been a democratic choice, was eliminated because he did not get enough first round votes. So, while it avoided the worst result, the system produced a second-best outcome.
However, while AV prevents the kind of undemocratic result that often occurs under FPTP, and always ensures that the winning candidate has at least some kind of majority support, it has nothing to do with Proportional Representation. It does not follow that centrist or compromise candidates such as the LibDems will be more likely to win – only that the least popular cannot win. A centrist candidate such as a LibDem will still need enough first preference votes to stand a chance.
In most English constituencies the LibDem candidate would typically receive the third highest number of first preferences among the big three parties (after the minor candidates have been eliminated). That means they would be eliminated nearly every time under AV even if all voters ranked them as their second preference. So the second preference votes of the LibDem voters will be transferred and make either the Labour or Conservative candidate the winner.
In order to gain more seats at Westminster under AV, the LibDems would need to have enough candidates who are the natural first preference of enough voters to benefit from AV. That is a tall order, leading me to conclude there will be no natural tendency to elect parties of the centre. In Australia, where AV has been used for over a century to elect the House of Representatives, it has rarely led to a hung parliament.
Academics at Essex University1 have analysed the 2010 general election results together with survey data on voters’ second and third preferences, and estimated that the Liberal Democrats would have gained an extra 32 seats under AV. However this probably overestimates the scale of the effect because it ignores tactical voting.
In the general election the LibDem vote will have been swelled by natural Tory or Labour supporters voting tactically in some constituencies to try to keep the other main party out. Under AV there will be little opportunity for tactical voting and the LibDems will suffer as a result. Labour and Tory supporters will be able to vote sincerely for their party, with LibDem as their second preference, safe in the knowledge that if their candidate is eliminated their vote will still count.
Labour and Tory supporters will be able to vote sincerely for their party, with LibDem as their second preference, safe in the knowledge that if their candidate is eliminated their vote will still count
My conclusion is that there is little convincing evidence that the LibDems will do better (even ignoring the drop of support they have experienced since joining the coalition government).
There is one quirk of AV that is not shared by FPTP. A candidate who would win under AV (whether or not he would win under FPTP) with a particular set of voter preferences, could actually become a loser if voters changed their preferences so as to make him or her more popular. This paradoxical possibility seems perverse and profoundly at odds with the democratic principle. AV can fail to reflect voters’ preferences in a direct and positive way. This is a well-known theoretical property of AV but we do not know how likely it is to be a problem in practice.
A yet further problem that AV shares with FPTP, and most other voting methods, is that the results can be influenced by the presence of candidates who have no hope of winning. Thus, for example, if a minority candidate drops out before election day, this can change the result because he or she takes votes differently from the major candidates. We see this with FPTP when a no-hope candidate takes votes away from a possible winner, and splits the vote. This happened in 2010 where a number of seats were denied to the Tories by the presence of a UKIP candidate. It seems likely this effect will be a lot weaker under AV.
I will be voting in the referendum for AV because, while it is flawed, it is better than First Past The Post for two good reasons. First, it rules out the possibility of an unpopular extremist being elected due to the vote being divided among the main parties. Second, it frees voters to express their true preferences without having to think about voting tactically. Voters can vote for the candidate they prefer, rather than for the candidate they think is most likely to keep out the one they least prefer, in the knowledge that second preference votes will count if his or her preferred candidate is eliminated before the last round of counting.
I believe that adopting AV will lead to a transformation of politics by making the true wishes of the electorate clear at every election. It will then be apparent how unfair our system is and highlight the need for Proportional Representation. And having got rid of FPTP, it will make genuine reform easier.
The University of Warwick’s Centre for Competitive Advantage in the Global Economy (CAGE) and the Coventry Telegraph hosted a public debate on the referendum on the Alternative Vote. Coventry Telegraph’s Political Correspondent chaired the debate between 'Yes to AV' supporters Professor Dennis Leech from the University of Warwick’s Department of Economics and Labour Councillor Kevin Maton; and a panel of 'No to AV' supporters which included Coventry Conservative Councillor Tim Sawdon and Coventry Labour Councillor Phil Townshend.
1. Simulating the Effects of the Alternative Vote in the 2010 UK General Election, by David Sanders, Harold Clarke, Marianne Stewart and Paul Whiteley, Parliamentary Affairs, 2010, 1-19
Professor Dennis Leech, Professor of Economics at the University of Warwick, Co-Director of the Voting Powers and Procedures Programme and Research Associate at the Centre for Philosophy and Natural Sciences at LSE. Professor Leech's research interests centre around voting power analysis, with special reference to better understanding weighted voting systems through the use of power indices.
Related WRAP Articles
Dhillon, Amrita (2003) Political parties and coalition formation. Working Paper. University of Warwick, Department of Economics, Coventry.
Grant, Wyn (2008) The changing patterns of group politics in Britain. British Politics, Vol.3 (No.2). pp. 204-222. ISSN 1746-918X.
Xefteris, Dimitrios and Matakos, Kostas (2009) Endogenous choice of electoral rules in a multi-party system with two dominant parties. Working Paper. University of Warwick, Economics Department, University of Warwick.
Also on the Knowledge Centre...
As the changes by the coalition government start to take effect, it is interesting to look back at how UK politics has developed.
Free schools are supposed to create opportunities for choice and relevance in education, but how do these principles work in practice?