« Supreme Court of Canada on restitution of invalid taxes | Main | Supreme Court of Canada - Decisions - Security Certificate Case - Charkaoui v. Canada (Citizenship and Immigration) »

2007.02.20

Lords reform again ...

Guardian Unlimited Politics reports that the UK government has caved ... not on the substantive question of House of Lords reform canvassed in the White Paper but on how MPs were to vote on the options. The White Paper had (sensibly) proposed an alternative voting system (ie something like the preferential voting familiar to Australian voters) but the opposition seems to have regarded this as a "dangerous constitutional precedent" ... :

The government backed down on controversial plans to use a new voting method for MPs to decide on House of Lords reform.

Jack Straw, the leader of the Commons, told MPs that they would not have to use his preferred method of an alternative vote ballot.

Instead they will use the traditional division system and Labour MPs would get a free vote.

The decision revives the possibility of a stalemate on reform with MPs rejecting all the options as they did in 2003.

Mr Straw had faced criticism when he set out proposals for MPs to list the various options - ranging from a fully elected to a fully appointed second chamber - in order of preference.

Theresa May, his Tory shadow, said that the move was a "victory for common sense".

To carry on with the original plans would have set a "dangerous constitutional precedent", she claimed.


February 20, 2007 in Constitutional commentary | Permalink

Comments

What a bizarre situation. MPs elected by simple first-past-the-post plurality insist on retaining the seriatim motion-and-amendment system for their own decisions in Parliament - despite the fact that the seriatim method could lead to a complete deadlock on important policy matters.

Some conservative writers think this is great, since it supposedly ensures only a "Condorcet winner" is chosen:

'... Political scientists use the Condorcet technique, in the form of a "Condorcet winner", to estimate the strength of a policy, or a political individual, or an ideology that has won a battle but not yet a war. Mackerras turned to Australia's trudge toward a republic for a vivid, somewhat tongue-in-cheek illustration by putting the Queen forward as a Condorcet winner. "We have already seen the Queen versus Malcolm Turnbull," Mackerras says, "and she won easily. If we matched her against an elected, executive presidency, she would have another walkover." The psephologist claims that a plebiscite on whether Australians want a republic would have limited significance. The question was whether any republican model acceptable to voters would vanquish Her Majesty, the Condorcet winner...'

- Frank Devine, "Taking a useful measure," Weekend Australian (25-26 November 2000), Review p 11.

The problem is that either Devine or Mackerras misunderstands how Condorcet's principle can be manipulated when used in a series of consecutive yes/ no votes. Condorcet's principle requires only that the winner defeats all rivals one-on-one. It may be that this "beats-all" winner could have only a low first-preference vote - too low to survive early elimination under an Australian-style AV preferential system - but first-preference votes are even less decisive under the Condorcet form of preferential voting than they are under the AV version.

In a referendum, however, the "vote no this time because it goes too far, and vote no next time because it doesn't go far enough" tactic ensures that any change proposal has to have an absolute majority of YES votes to get through - which equates to an absolute majority of first-preference support if one can convince voters (a) to reject any proposal that isn't their very first preference, and (b) that voting against their second or third preference this time will help to first preference get on the referendum ballot in future (historically, a very dubious assumption) - and this is a far steeper hurdle than the Marquis de Condorcet ever proposed.

Posted by: Tom Round | Mar 5, 2007 1:29:55 PM

Post a comment