AV: For and against

On 5th May, Britain will vote whether to change the rules for electing MPs. James Purnell, former Labour minister, says we need change. James Forder, economist, says we don’t
March 23, 2011
What do we want? Protesters for electoral reform besiege Lib Dem coalition talks last May




James Purnell: In its 1950s heyday, first-past-the-post (FPTP) produced strong and stable government, aided by high turnout and elections in which many seats changed hands between Labour and Tory. Since then, the proportion of people voting for the main two parties has fallen from 97 per cent to 65 per cent. Parties other than the big three got one in ten votes at the 2005 and 2010 elections. Research by John Curtice at the University of Strathclyde on these trends shows that—as with the most recent election—hung parliaments and coalitions will become increasingly likely whatever system we have.

This demise in two-party politics makes FPTP a busted flush. First, elections are increasingly decided by a handful of people in marginal seats (the Institute for Public Policy Research estimates it was just 460,000 voters at the last election). Second, the number of safe seats has increased; half of all seats have been held by the same party since 1970. Third, an increasing number of MPs—two thirds in 2010—are elected without majority support in their constituencies, with some getting as little as 30 per cent of the vote. Fourth, the system encourages candidates to target rivals who are their ideological neighbours, rather than working with them to defeat their real opponents. Fifth, FPTP encourages large numbers of people to vote tactically or risk wasting their vote since, as campaign leaflets are always reminding us, “X party can’t win in this area.”

The alternative vote (AV) is a small change, in keeping with Britain’s evolutionary approach to constitutional reform, but one that addresses these problems. We should vote “yes.”



James Forder: In choosing an electoral system, what we want is a means of securing effective and accountable government. Effectiveness argues for single-party government, but accountability insists upon it. FPTP is the best choice because it usually gives us single-party government, as British election results since the 1950s have shown. Curtice’s research has been contradicted by work by Pippa Norris of Harvard and Ivor Crewe, now at Oxford.

The advantages of single-party government are usually seen in terms of it resulting in “strong” or “stable” government. An even bigger advantage is that it empowers voters and gives us a proper chance of holding politicians to account. First, it means that elections decide who governs. Where coalition is necessary—as it surely would be in a Britain with AV—it is politicians who decide who governs, and they do so after the election. Second, campaign promises amount to more or less nothing. The parties have too many excuses for abandoning their commitments due to the “need to reach agreement.” Third, when the next election comes, the governing parties have excuses for whatever has gone wrong—blame the other party! And even when voters know they want a party out, there is no guarantee that will happen.

Effective democracy requires that voters, not politicians, decide who governs, and that they do so knowing who is responsible for what has been done, and what is promised. For the sake of that we should stick to FPTP.

Purnell: Yours is an argument against proportional representation (PR), not against AV. I agree our electoral system must aim to form effective governments. But that can’t be the only goal, otherwise you would have no preference between choosing governments by FPTP or by the toss of a coin.

May I suggest an addition to your principle? The goal should be to form a single party government when that is the voters’ wish, but a coalition in other cases. FPTP meets that goal better than PR. But when an election is close, FPTP is a bit like Russian roulette. Sometimes a party will get a majority because its support is geographically concentrated, other times you will get a coalition.

AV allows voters to make that choice more consciously. It’s impossible to know for sure what results would have obtained at past elections. But we can be fairly confident that Labour would have won a majority under AV in 1997, 2001 and (more contentiously) 2005, and that the Tories would have won in 1979, 1983 and 1987.

The voters might have elected coalitions in 1974 and 1992, and would still have done so in 2010. But they would have been able to do so without having to vote tactically in the ballot box, on the basis of guesswork about how everyone else is voting.

Moreover, AV allows parties to campaign on what coalition they would form, so that voters are involved in the decision of what happens in a hung parliament. That’s much harder under FPTP because all parties have to pretend they are going to win an outright majority so that voters don’t think that a vote for them is wasted. AV gets rid of that problem too.



Forder: Mine is an argument against PR, but it is also an argument against AV, and with a couple of extra twists. One is simply that, since AV does nothing for the small parties, it threatens us with a three-party system in which the Lib Dems are perpetual kingmakers—this would be the antithesis of democratic accountability.

The second is related to your idea of the electorate wishing for a coalition. What you seem to mean by this is that roughly equal numbers of voters want different parties to govern alone—not the same thing at all. In 2010 no one voted for the coalition. It, and its programme, are the creation of politicians.

But even on your basis, AV is no good since it is a capricious system. Imagine the case where three parties have about equal support. At one extreme, if the Lib Dems were third in every seat, they win nothing. But another possibility—and a less extreme one—is that they come second in enough seats so that after the redistribution of preferences they could even gain an overall majority. That can easily happen even if they came third in the number of votes cast overall. This would be a travesty, and it shows that your goal—of delivering coalitions in close elections—is not something AV offers. Far from it—it is AV that is electoral Russian roulette.

So I will stick with my principle. An electoral system should deliver effective and accountable government. PR will not do that; AV plays dice with it. FPTP gives us the best chance of keeping voters in charge and politicians where they belong.

Purnell: Equally, FPTP could deliver an all BNP parliament on, say, 26 per cent of the vote. Let’s judge by what happens in the real world. Canada has FPTP and has delivered a hung parliament in each of the last three elections. Australia, which has AV, has tended to deliver single-party rule (1940 and 2010 are the honourable exceptions). In the election last May, under FPTP, the Tories needed to win just 24 more seats to cause a hung parliament but 115 to win an overall majority. You simply cannot get around the fact that all electoral systems will sometimes result in hung parliaments. In a multi-party system like Britain, the public will sometimes want parties to work together in government.

If I ask for a pint of Thwaites in the pub and am told it’s off, I don’t walk out, I ask for my second choice instead. AV is more in keeping with most people’s experiences of life than FPTP, and it is a better way of choosing who governs. It gives MPs an incentive to reach out to more voters and means everyone has a reason to vote, since they can do so with both their heart and head. It combines the advantages of FPTP without most of its disadvantages. We should vote for it.

Forder: But what would we expect for Britain? In Australia, there was a more or less permanent coalition between Liberal and National parties so that it operated almost as a two-party system. If you count them as separate there have been plenty of hung parliaments. Canadian politics is riven by the Quebec question (and AV has been tried and abandoned in three states).

Look at the recent past. During the 2010 election, Paddy Ashdown said the Lib Dems could not form a coalition with the Tories; and the Lib Dems made no end of fuss about their promise on university tuition fees. After the election, none of that mattered because voters were out of the picture.

Nor does AV do anything to meet the principled concerns of the advocates of PR. The Lib Dems might do OK under PR, but what about those one in ten people who voted for none of the big three? It offers the worst of all worlds: coalition, freeing parties from voter control; one party regularly in position to dictate terms to the others; capricious representation of the big three; and nothing at all for the smaller parties.

In FPTP, after the votes are counted, we usually know what government we have. That is what empowers the voters.

For more on this topic, see: Anne McElvoy on why not to vote for AV and Peter Kellner on why electoral reform won’t just change the way we choose MPs, but the way we do politics