Electoral ‘reform’ is back on the agenda: but do we need it?
In a major Commons statement today (10 June 2009) introducing a national debate on constitutional reform, the prime minister included the system of elections to the House of Commons in his five constitutional topics for debate and possible reform:
[L]ast year we published our review of the electoral system and there is a long-standing debate on this issue. I still believe the link between the MP and constituency is essential and that it is the constituency that is best able to hold MPs to account. We should only be prepared to propose change if there is a broad consensus in the country that it would strengthen our democracy and our politics by improving the effectiveness and legitimacy of both Government and Parliament; and by enhancing the level and quality of public representation and engagement. Mr Speaker, we will set out proposals for taking this debate forward.
The Alternative Vote
The new system beginning to emerge as front runner for Commons elections is AV — the Alternative Vote. Under AV the voter ranks all the candidates in a single-member constituency — or as many or as few of them as he wishes — in order of preference: 1, 2, 3, etc. If no candidate receives 51% of the first preference votes, the candidate with the fewest first preference votes is eliminated and his second preference votes are distributed to the rest. This process continues until one candidate has received 51% or more of the votes — his own first preference votes plus eliminated candidates’ second and third etc preference votes allocated to him. That candidate is then declared elected. It’s important to remember that AV is not a form of proportional representation (PR). There’s some dispute among the experts over whether AV is likely to produce a more proportional result than First Past the Post (FPTP) — i.e. whether the proportion of seats won under AV will be closer to the proportion of votes cast for each party. In some circumstances it’s argued that AV might actually produce a less proportional result.
In a post today (10 June 09) on Labour List, the General Secretary of the Fabian Society, Sunder Katwala, has written a persuasive case for a combination of AV for the House of Commons and some kind of PR for the House of Lords, or whatever an eventually reformed second chamber turns out to be called. Katwala includes extremely useful links to a number of other online documents describing and assessing the various systems available, including Katwala’s own Fabian Essay of autumn 2007 which argues the case for AV in greater detail. He regards it as less than ideal but the change most likely to command wide acceptance.
I can’t myself see a lot of point in AV. A candidate who fails to win 51% of the first preference votes in his constituency can’t credibly pretend that after transfers of some (not all) preference votes from other candidates he has somehow magically acquired more than 50% support when in sober fact he hasn’t — the majority of the voters actually voted for someone else. Second preference votes are manifestly not the same as first preference votes and there’s something slightly absurd about pretending that they are, by adding both kinds together in order to produce a desired result. Anyway, why transfer to the candidate with the most first preferences only the second preferences of other candidates who have been eliminated, who by definition will always be those with the least support of all? Why give special treatment and effectiveness to all the least supported candidates’ second preferences and ignore the second preferences of, say, the candidates who came second and third in the first count, were never eliminated, but failed to get elected? (At the time of the London mayoral elections, held under a simplified form of AV, I had great difficulty in convincing several friends, none of them politically illiterate, that if they were casting their first preference votes for either Livingstone or Boris Johnson, it was a waste of time recording a second preference vote for anyone else, such as the LibDem policeman, because obviously neither Ken’s nor Boris’s second preferences would ever get redistributed.)
There’s a lot to be said, at both constituency and national level, for a straightforward system under which most of the time the winner is the candidate (or the party) which has won more votes (or seats) than any other — as FPTP does. There have only been two elections since the second world war when a party winning the most votes nationally has not also won a majority of the seats in the house of commons, once benefiting the Tories and the other Labour (so swings and roundabouts…). There’s nothing especially significant about 50% in either context. Is 48% unfair but 51% somehow fine? Since 1935, not a single party has ever won 50% or more of the national vote. It can be said of every single government since the war that more people voted against it than for it. So what? Almost invariably the government won more votes than any other party and that should be good enough — and vastly preferable to any form of PR, under which the LibDems (or Labour, if it becomes the third party!) become permanent ex officio king-makers with the power to decide, by demands and threats in a private horse-trading session, which of the two biggest parties gets the keys to No. 10, along with a messy compromise policy programme forced on it by the minority party in negotiations after the polls have closed as the price for their support, a programme for which not a single voter can have voted. Katwala’s Fabian essay puts the case against PR in a nutshell: the advantage of FPTP or AV “is accountability: voters choose governments, rather than minority party leaders in some smoke-filled room having disproportionate power to decide who governs.” [Emphasis added]
The Jenkins Commission recommendations for PR
In October 1998 Lord (Roy) Jenkins, former Labour home secretary and later breakaway Social Democrat, and his colleagues, published their report and recommendations on possible changes to the electoral system for elections to the House of Commons. They proposed a complex scheme under which voters would have two votes: one under AV for a constituency MP, and another for a party top-up list that would produce an overall result of seats more nearly in proportion to votes cast. You can read the full Jenkins report here (if you have the time and the intestinal fortitude; it’s quite hard work).
Among the voices currently calling for “electoral reform” as the centrepiece of a series of mostly disconnected proposals for constitutional change, there have been suggestions that we should go back to the Jenkins report and adopt his suggested system. In my view this would be a mistake. I spelled out my reasons for this view a month after the report was published, in a commentary on my website (here). Among the principal objections to Jenkins are two, each in my view decisive: first, that it would almost always produce unstable coalition or minority governments with unpredictable compromise policy programmes for which no-one could have voted in the election; and second, that it would entail two classes of MP, one class elected in and accountable to individual constituencies and the other in effect appointed by the party machines on the basis of wider areas, different from the AV constituencies.
As I said in my commentary on Jenkins, all electoral systems have their defects and injustices: none is perfect. Any system involves a trade-off between ‘fairness‘ — ensuring that the distribution of votes is accurately (or reasonably accurately) reflected in the number of Commons seats: and pragmatism — ensuring as far as possible that most elections produce a stable government with a sufficient overall majority to put its election manifesto into effect without having to negotiate deals with other parties in order to get into and stay in office. Whether you attach more weight to stable and accountable government than to theoretical fairness in the make-up of the House of Commons, or vice versa, will always be a subjective decision. It’s common ground, even with Jenkins, that the principal function of the Commons, that which most heavily influences most individual votes at election time, is to generate and sustain a government. Personally I dislike the idea of endless coalitions in which a government which has received more electoral support than any other party is permanently at the mercy of potentially fickle and wayward minority parties (look at Israel, to quote an extreme example): not a recipe for sound (or radical) government. So I remain convinced that the alternatives to FPTP (or AV) for a government-generating chamber are worse. PR is fine for a debating chamber that doesn’t create governments, like the House of Lords.
One member of the Jenkins team, Lord Alexander, disagreed with the majority recommendation for AV, giving his reasons in a minority report, still available here. His arguments against AV seem to me a lot more persuasive than those in the majority report in favour of it. Among other things he makes the cogent point that under FPTP, whether or not a candidate gets 50% + of the votes, once elected he serves all his constituents, regardless of party — a valuable convention that would be destroyed by an STV system involving multi-member constituencies, inevitably resulting in MPs regarding themselves as serving only their own party supporters and being answerable only to them. That would surely be very retrograde. Unfortunately Lord Alexander does, however, support the Jenkins proposals for a second top-up vote to produce a form of PR.
Incidentally, we should all be campaigning against the vicious system of party lists foisted on us for European elections — more control freakery from Westminster.
It’s going to be a long and difficult national conversation!
Brian
While I too have concerns about PR (I rather like the idea of an non-PR Commons plus PR Lords) I do think that the FPTP system is fatally flawed in the way that vote splitting can mean that a candidate can get in with more than 50% of the voters preferring a specific other candidate and the way it encourages “tactical” voting. Supporters of minority parties have to make a decision whether to “waste” their vote on the party they actually prefer or to vote for the main party they loath least.
Anybody who supports a system with these problems for anybody who is not behind one of the entrenched parties doesn’t get to moan about voter apathy, I’m afraid.
Brian writes: I’m afraid that there will always be constituencies where no one party is supported by 50% or more of the voters. No British government since 1935 has won as much as 50% of the votes cast. It just means that public opinion is divided into penny packets, none of them reaching anything like 50% supporting any one party. That, after all, is also the situation nation-wide: for three-quarters of a century and counting, not a single British political party has commanded the support of as much as half the electorate. This is bound to be reflected, however unevenly, in individual constituencies, and no electoral system, however complex, can wish it away. As PR would encourage big parties to split up into smaller ones, and enable very small one-issue parties to get some representation in parliament, the chances of any party even approaching 50% support would be significantly further reduced compared with FPTP. Why does it matter?
Whatever fancy voting systems anyone can devise, you can’t in the end get away from the unpalatable fact that the choice for each voter at a general election is a binary one: there are at present only two parties likely to be capable of forming a government, and the vote simply enables you to choose between those two. There’s no way of making your vote do other things as well, such as registering your support for some items in a party’s programme but not for others, or indicating regret that the Greens or the LibDems have no chance of forming a government, or wishing for a hung parliament, or hoping that two specific parties will form a coalition. The only practical effect of voting for any third party is either to damage or improve the chances of either Labour or the Conservatives forming a government, depending on the situation in the relevant constituency. Your vote might have a marginal effect on a third party getting an extra seat in parliament, but the minority parties’ influence on legislation and government decisions is not really affected by the size of its representation (except just possibly in the case of a hung parliament or very tight parliamentary votes, both rare events in normal times). The most significant effect of a vote for a third or other minority party is to damage or improve the chances of either Labour or the Conservatives forming a government. Of course in some constituencies and in some political circumstances no individual vote is likely to affect the outcome at all, but the only way to get round that is to turn the whole country into a single multi-member constituency — not a remedy likely to command widespread support.
Most political decisions, including voting, require a choice between rival evils: the trick is to identify the least evil of them, hold your nose, and take the plunge. Faffing around in search of a crock of perfect gold is not just a waste of time: it actually obscures the nature of the choice that has to be made.
I don’t follow your (and many other commentators’) objection to ‘tactical voting’. All that it means is making a rational assessment, in the specific circumstances of the constituency in which you vote, of the likely effects of the various voting options available on the outcome of the election, and choosing whichever option produces the least objectionable likely consequences. This is a process which we have to undertake many times a day throughout our lives, and since we have to take responsibility for the probable effects of our actions, there seems no reason to object to it when it comes to deciding how to vote. It’s a far less complicated and difficult calculation than the one every voter would have to try to make under the STV system of PR: you only have to look at the way the votes are counted and allocated here [pdf] to realise how fiendishly difficult it would be to decide how best to vote under that system.
I’m afraid that there will always be constituencies where no one party is supported by 50% or more of the voters.
Of course. However, that doesn’t mean that it’s a good thing for candidate A to get in when, given the choice between just A and B, more than 50% of the voters would prefer B.
Whatever fancy voting systems anyone can devise, you can’t in the end get away from the unpalatable fact that the choice for each voter at a general election is a binary one: there are at present only two parties likely to be capable of forming a government, and the vote simply enables you to choose between those two.
I think you’re probably right that there will always be two dominant parties. However, perhaps the system should be more open to replacing one or both of them if they become too indistinguishable and obnoxious or if the issues change: from a left/right to an authoritarian/liberal axis, for example.
Brian writes: Thanks once again, Ed. I entirely agree about the desirability of a system that makes it possible for one of two dominant parties to be replaced if the political situation changes fundamentally in one of the ways you suggest: and I would accept that AV (or any PR system) would facilitate that, although whether this is enough to outweigh the accompanying drawbacks is a matter of judgement.
On your first point, though, I can’t think offhand of any electoral system that would prevent a candidate preferred by more than 50% of the voters from being elected, if there are only two candidates. If there are more than two candidates, any of them winning more than 50% of the votes would automatically be elected. But under AV, if no candidate gets more than 50% of the first preferences, and if candidate A gets more first preference votes than candidate B, B might nevertheless get elected if he receives enough second preference votes redistributed to him from candidates who have come bottom of the first and subsequent counts (and are therefore eliminated). But you can’t infer from this that a majority of all the voters, if they had been given a straight choice between A and B, would have preferred B, since you have no way of knowing which of A or B would have got the greater number of second preferences redistributed from candidates who were not eliminated, and whose second preferences were therefore never redistributed or counted. In this scenario, B owes his victory over A purely to the accident that he got more second preference votes from the candidates who received the fewest first preference votes (and were thus eliminated), and not from anyone else. A random selection of second preferences has overridden the choice of all those casting first preference votes. This is pure lottery: you might just as well stick pins in the list to choose the winner. On any reckoning, FPTP would be fairer.
I’ve been following with interest the dialogue on the subject of electoral reform. I don’t know what worries me more: the fact that half the electorate has (have?) unyielding loyalties to one party or another, and would vote for a pig with the right rosette in its buttonhole; or the tendency of much of the remainder to switch preference with every gaffe, ‘revelation’, or snide statement, or to refrain from voting at all out of mental laziness, cynicism, or despair and general fed-up-ness. There seems to be no simple system – emphasis on simple – whereby the composition of the House of Commons approximates the electorate’s distribution of voting preferences, and at the same time the present system is retained of single-constituency members who reside in constituencies small enough for them to be well familiar with. More complex systems than FPTP either contain illogicalities or are ridiculous – I can (just) visualise a candidate being returned under the transferable vote scheme as the result of the second choices of those voting for the eliminated Fascist or Monster Raving Loony candidates – and moreover their true effectiveness is limited (I say this with regret) by the fact that the average tabloid-educated Briton has a low reading age and a pretty short attention span, which limits both the complexity of any ballot form and the length of a voter’s pamphlet explaining how to fill in the form and how the system works.
A pity, though: I would like to have a formal means of saying ‘I would like to see party A in government but not if X continues to lead it’, and ‘Candidate P is a good man and I want him to continue to represent me, but I don’t want his party in power’, and ‘I like party B’s policy on banking reform and C’s on the Health Service, and the Ds have interesting ideas about education’; the latter being advisory comments, indicating what the next government might ignore at its peril or where it may need to make a special effort to get public support. Meanwhile, we continue to vote for one package deal or another. Returning to the possible, I would be greatly interested to read comments on other electoral systems, not of a theoretical nature on the likely merits and demerits of suggested changes in the British system, but of a more substantial type from citizens of those democracies where something other than FPTP is already in place: reports, in other words, from satisfied and dissatisfied users of the alternatives.
What I really meant to ask you was, what is your opinion on making the vote mandatory? I believe it is in several countries, and not only those where the One True Party is voted back by 99.8% of those voting in a 99.9% turnout. Australia (or some of its states?) comes to mind, though I may be wrong. For me, voting in a parliamentary election should be considered a legal duty like paying taxes, serving on a jury if drafted, reporting for duty when called up, and obeying the laws generally. The word to the electorate has got to be: “This is your country and you have a small say in its destiny. Make the most of it, take an interest, study the issues and form an opinion for pity’s sake, and use your vote and know you’ve done your duty.”
I would make one concession to those who are sincerely repelled by having to hold their noses and chose a candidate; a box to tick labelled “I am not satisfied with any of the above candidates.” If Not Satisfied gains most votes in any constituency, all the (live) candidates will be bent over on the steps of the Town Hall and given six of the best, regardless of age, sex, or religion, while thousands cheer. And if Non Satisfied wins the most constituencies…? Just joking, but only just.
Brian writes: Tim, thanks for these interesting comments. I don’t share your concern about the large number of voters who are reasonably firmly committed to a particular party and will usually vote for it regardless of the qualities (or lack of them) of its candidate. In my view a vote in a general election must in the nature of things be a broad-brush choice between wide-ranging philosophies and values as represented by each of the two parties capable of forming a government. That choice can’t by definition be nuancé in the way you would like it to be. No party can reproduce in fine detail all the shades of opinion on every political issue held by each individual voter. Every thinking person will have strong reservations about many of each party’s policies, leaders and candidates. You just have to decide which comes closest to sharing your fundamental political values, and vote accordingly.
Similarly, your wish for a system that reflects, even approximately, the distribution of opinion and party allegiances in the country, retains the good features of fairly small single-member constituencies, and (I would add) simultaneously produces a government with a sufficient majority to be able to carry out its election promises without having to adopt compromises forced on it in unavoidable horse-trading with a minority party or parties after the election, seems to me incapable of fulfilment. It’s a fact of life, not a feature of the electoral system, that in normal current circumstances no one party ever enjoys the support of the majority of the electorate. Ergo the sum total of elected members of parliament representing the party with a plurality of the vote can’t have enjoyed, in total, the support of a majority of the electorate. AV tries to fudge this reality but can do so only by pretending that the second preference votes of minor parties, often those from the lunatic fringe as you rightly suggest, are equal in value to the first preference votes cast for the other candidates — and that the second preference votes of minority parties which have never been eliminated from recounts are without any value at all! That this intellectual conjuring trick should be winning such mounting public support is really pretty depressing.
Finally, I entirely agree with you about compulsory voting — as indeed practised in Australia and elsewhere. Not only would it reflect the clear obligation resting on every eligible adult to play his or her part in the democratic process at least once every four or five years, but it might also focus more minds on the need to make broad political choices and to acquire the information required to exercise them. I’m not sure about your suggestion, often made, of offering a vote for “None of the above”. I fear that would actively encourage a lazy attitude of “a plague on both, or all, your houses”, and “they’re all as bad as each other”, which is just a feeble excuse for not taking the trouble to examine the issues and the records of the parties that seek your vote. Such indolents already have the option of spoiling their ballot papers, as quite a large number of people do. Compulsory voting would inevitably increase the number of spoiled papers (as an expression of resentment by the ignorant or congenitally lazy at being obliged by law to vote). Personally I think this would be a price worth paying.