Does AV lead to more representative governments?
AV has been put forward in this referendum as an electoral system that creates a fairer outcome at elections. Campaigners say that under AV people's votes are better represented in the election outcome locally and nationally.
No to AV: "AV is not a proportional system. In three out of the past four British elections AV would have produced more disproportionate results than First Past the Post. And in landslide elections, AV would have dramatically exaggerated the winners' majority. For example, in 1997, Tony Blair would have had a majority of 245 MPs under AV, despite only getting 43% of the vote."
Models of what past election results would have been in the UK under a system of AV give a mixed result. They show that in 1997 when Labour gained a big majority, AV would have produced more unrepresentative results.
On the other hand, the three elections previous to Labour's 1997 landslide would have seen more representative results under AV.
Both AV and First Past the Post (FPTP) produce unrepresentative results, in terms of how closely the make-up of MPs in parliament would represent the way people vote.
This is in comparison to full Proportional Representation (PR) systems of election which produce very representative results, but loose the constituency link that AV and FPTP maintain.
On the issue of whether AV is more proportional in its results, the Constitution Society briefing paper states that: "There is nothing in the operation of the AV system which increases the correlation between each Party's percentage of the national vote and the number of Parliamentary seats it secures."
The Jenkins Commission report, which looked into the advantages of different electoral systems in 1997, was also sceptical of AV being more representative than FPTP. It said that: "There is not the slightest reason to think that AV would reduce the stability of government; it might indeed lead to larger parliamentary majorities."
The report goes on to cite research conducted on the 1997 general election, which shows that under AV the Labour party would have had an even bigger majority of MPs.
They said: "A 'best guess' projection of the shape of the current Parliament under AV suggests … the following outcome with the actual FPTP figures given in brackets after the projected figures: Labour 452 (419), Conservative 96 (165), Liberal Democrats 82 (46), others 29 (29). The overall Labour majority could thus have risen from 169 to 245. On another equally reputable estimate the figures are given as Labour 436, Conservatives 110, Liberal Democrats 84 and others 29, an overall majority this time of 213. On either basis an injustice to the Liberal Democrats would have been nearly two-thirds corrected (their strictly proportional entitlement was 111 seats) but at the price of a still greater injustice to the Conservatives."
It went on to say: "In the three previous elections, those of 1983, 1987 and 1992, AV would have had a less distorting effect on proportionality between the two main parties."
In conclusion, the report argued that: "So far from doing much to relieve disproportionality, [AV] is capable of substantially adding to it. Second, its effects (on its own without any corrective mechanism) are disturbingly unpredictable."
An interesting piece of context to this issue is provided in the Institute for Public Policy Research report, which looks at the ratio of votes to MPs, under FPTP, for the main parties in the 2010 general election. This shows that it took:
33,468 votes to elect a Labour MP
35,028 votes to elect a Conservative MP
119,780 votes to elect a Liberal Democrat MP
Under AV this distortion of the number of votes required to elect a Liberal Democrat MP is likely to be reduced slightly. However, the core reason for the distortion is that Liberal Democrat support is fairly evenly spread across the UK, with few areas of strong support. In contrast, Labour and the Conservative both have more concentrated support, and so win more seats relative to votes gained.
Data from Australia, which uses AV, shows that the make-up of elected politicians is not very representative of voting.
The House of Commons Briefing Paper on AV points to the fact that in Australia: "The AV system for the House of Representatives has helped to deliver majority government in Australia. There is effectively a two party system, with a majority held either by the Liberal/National Party coalition... or the Australian Labor Party.
The British Academy report on AV also points out that in Australia: "the two largest parties/coalitions won 85.5% of the first preference votes but 98.7% of the seats: with 14.5% of the votes, the smaller parties won just 2 of the 150 seats in the House."
While AV does reduce the unrepresentative results produced by FPTP in some election scenarios, such as where no party gains a clear majority, it does so unpredictably and to a limited extent.
Whether or not AV leads to a more representative parliament would also depend on how many people use their second preference votes on the ballot. The more people who fully rank the candidates on their ballot paper, the more representative the final outcome of the election is likely to be.