The PB AV debate: Part 2

The PB AV debate: Part 2

Rod Crosby puts the case for voting YES

Theoretical Case
Arrow’s Impossibility Theorem states that no (#1) voting system can be totally fair when choosing between more than two candidates. Selecting a single-winner voting system therefore depends on which aspects of fairness we consider more important than others.

FPTP’s unfairness has been understood for centuries, and its most deprecated flaws are:-

1) failing the Majority Loser Criterion. FPTP can elect the candidate whom the overwhelming majority of voters wish to defeat.

2) failing the Condorcet Loser Criterion. FPTP can elect the candidate who would lose a head-to-head contest with every other candidate.

3) failing the Independence of Clones Criterion. The well-known phenomenon, also known as “the spoiler effect”, where two similar candidates split the vote, allowing the election of a candidate inimical to both. See Bush, Gore, Nader, 2000.

4) failing the Mutual Majority Criterion. If 59% prefer lemonade, fruit-juice or tea, and 41% prefer whisky or beer, FPTP says it’s democratic for the majority to have to drink whisky.

5) failing the Favourite Betrayal Criterion. FPTP encourages some voters to behave dishonestly, abandoning their sincere choice for the “lesser of two evils.” Philosophers, unsurprisingly, find this troubling. In the UK we know this syndrome as “tactical voting”.

The Alternative Vote completely cures the first four problems, and while not wholly removing the fifth (no system can) it renders it less severe. Last, but not least, AV demonstrates that every elected candidate has the acquiescence of the majority of the voters, increasing their legitimacy. By objective measures, therefore, AV is a far superior system to FPTP (and also superior to other single-winner methods).

Empirical Case
Let us dismiss some canards about the operation of AV in Australia. Firstly, Australia does have a third party. It’s an agrarian party called the National Party (formerly the Country Party), which retains its own identity. It has been in permanent if sometimes uneasy alliance with the Liberals in the lower House but acts independently in the Senate, and it’s probably fair to say that without AV it would have sooner or later been absorbed by the larger party, in a similar way that the Liberal Unionists and National Liberals were absorbed by the Tories here. Australian politics seems richer for the existence of this separate, if somewhat eccentric third party.

Secondly, there are currently 4 independents in the Australian lower House, which would equate proportionally to 17 in the British House of Commons. So it would appear that AV, in practice as well as in theory, is slightly more capable of electing independents than FPTP is. [Another example is the Irish Presidential election of 1990, where against all expectations AV elected an independent instead of the machine politician put forward by the major party.]

Thirdly and perhaps most surprisingly, AV is a semi-PR system(#2), displaying about half the dis-proportionality of FPTP elections, and has actually delivered results in Australia that are more proportional than the so-called PR systems used in the UK (#3) to elect Euro MPs and the devolved Assemblies!

For those reformers holding out for some perfect version of PR – which doesn’t even exist – remember, this is the first time since 1931 that any change to the voting system has been seriously proposed (and that was to AV). Eighty years is a long time to possibly wait for the next opportunity for reform, and our great-grandchildren will rightly look askance if the 2011 referendum is defeated by an unlikely coalition of diehards and dreamers…

1 some voting systems, such as Borda, Approval and Range appear to confound Arrow’s Theorem, but they are cardinal systems which violate the principle of majority rule (the Majority Criterion), so Arrow excluded them from his definition of a voting system. The shortcoming shared by these systems means that they are never likely to be considered suitable in a mass democracy.
2 Taagepera, R. and Shugart, M. S. (1989): Seats and Votes. The Effects and Determinants of Electoral Systems. Yale University Press, New Haven.
3 Kestelman, P. (2005) Apportionment and Proportionality: A Measured View. Voting Matters, Issue 20(4), The McDougall Trust

Comments are closed.