Dual Candidacy Again (and Again?)


It’s back – if not exactly by popular demand. Debate on the Wales Bill in the House of Commons last week occasioned yet more acrimony over ‘dual candidacy’ – whether individuals should be allowed to stand as both constituency and regional list candidates under the AMS electoral system used for National Assembly elections.

A particularly notable contribution to the debate came from former Secretary of State Peter Hain; his impassioned Commons speech was followed by this Western Mail article. When Hain introduced the dual candidacy ban, in the 2006 Government of Wales Act, one of his political opponents suggested to me that the issue was simply a political tool: a means to unite Labour MPs behind the Act and distract their attention away from the significant advances in devolution contained within it. (A suggestion, I might add, made not in anger but in admiration for Hain’s political skill.) But it’s clear that Hain, and Welsh Labour in general, continue to find dual candidacy strongly objectionable.

I don’t agree with them. Regular readers of the blog will be wholly unsurprised that I find Peter Hain’s arguments unconvincing. I don’t propose to rehearse my own views again (those interested can find them here, here and here.) But without retreating from those views on the substance of the issue, I want here to discuss a wider question raised by last week’s debate.

After two Assembly elections (1999 and 2003) in which dual candidacy was permitted, the 2006 Act meant that the elections of 2007 and 2011 occurred with it banned. The new Wales Bill proposes a further change, restoring dual candidacy for 2016. However, Labour’s attitude offers the prospect of yet another change, with the ban being re-instated.

We therefore face the prospect of the parties repeatedly squabbling over the issue (one that remains, for most voters, deeply obscure), and it generating some form of legislative ping-pong. This prospect is, to put it mildly, unappealing. General U.S. Grant once stated his willingness to “fight it out on this line if it takes all summer”. Do we really want to be fighting it out over dual candidacy for many, many summers?

How can this be avoided? I don’t think it’s really adequate for each side simply to insist that the other should give way. We could be a long time waiting for that to happen. Those on different sides of the argument have suggested either that the ban on dual candidacy damaged democratic politics in Wales; or that its repeal will do so. There seems little prospect of reaching consensus on this. But I think nearly everyone can understand that endless squabbling over the issue will be damaging.

Is there a way out of the impasse? While the AMS electoral system is in place, this seems unlikely – the only way out is for one side to give way. Only with a more fundamental change to the electoral system, to a system that does not have two (constituency and list) types of representative, might the issue be wholly resolved.

There is at least one other compelling reason for considering electoral system reform. There is a gradually emerging political consensus that the National Assembly requires an expanded membership. In a well-argued, evidence-based paper, the UK Changing Union project and the Electoral Reform Society have argued for 100 AMs. In practice, an increase from 60 to 80 might be more politically feasible. But there is no way of increasing the number of AMs (to 80, 100, or whatever number), without having a broader impact on the electoral system.

For instance: the easiest way to change from 60 to 80 AMs would be to raise the number of list AMs in each region (from 4 to 8). But with list AMs now comprising half of the Assembly’s membership, rather than one-third, the proportionality of the electoral system would be changed substantially. An 80-seat Assembly where 40 members came each from the constituency and list ballots would be more-or-less a fully proportional system, rather than the semi-proportional system we have at present. Of course many people (me included, actually) would view a more proportional electoral system as a very good thing. But it would change the party balance substantially; make it much more difficult for Labour ever to win a majority in the Assembly; and be rather different from what the Welsh public approved in the 1997 referendum.

To maintain the balance between constituency and list AMs in the same proportions as now, within an 80-seat National Assembly, would require increasing the number of Assembly constituencies to 53. This would require the drawing of new constituency boundaries, and losing ‘co-terminosity’ between Westminster and Assembly constituencies. Co-terminosity, of course, is not everything: it has already been abandoned in Scotland, and would be lost anyway in Wales if the number of MPs is reduced. But any move in that direction would probably increase the scope for public confusion, in the context of a Welsh devolution settlement, and an Assembly electoral system, that are already little understood by many people. Moreover, whatever adjustments might be made to the numbers of people elected under the AMS system, it still leaves in place the basic problem – the two types of representative – that has underpinned all arguments over dual candidacy.

Is there a way out of this? I believe there is. A decade ago the Richard Commission recommended that an 80-seat National Assembly should be elected by the Single Transferable Vote (STV) system. The detailed case made for STV by Richard was never really dealt with seriously by the Labour party at the time; it was simply dismissed out of hand. But STV has two potentially important advantages.

First, STV gets rid entirely of the issue of two types of AM. The whole dual candidacy issue immediately disappears.

Second, STV would likely produce results about as proportional as those delivered by the current electoral system. We can’t be certain about this: we don’t know the political circumstances of future elections, while estimating how elections conducted under AMS would translate into elections conducted under STV is inherently difficult. However, past Welsh Election Studies have helped generate reasonable estimates of the implications of STV, by asking people how they would have voted in a preferential (STV-style) ballot. These results can then be mapped onto plausible ways of electing 80 AMs via STV. (Two main ways have been suggested: grouping the current 40 constituencies into 20 pairs that each elect 4 AMs, or basing NAW elections around local authority boundaries.) These exercises have indicated that STV would be likely to produce very similar levels of overall proportionality to the current, 60-seat, version of AMS. The most detailed simulation of STV, conducted after the 2011 National Assembly election, projected Labour to win 40 seats in an 80-seat Assembly – the exact same proportion as they actually won.

At the moment, although STV is supported by Plaid Cymru and the Liberal Democrats, it does not seem to be something that the Labour party in Wales is willing to give serious consideration to. (Many in Labour seem curiously oblivious to the fact that they would, in nearly all political scenarios, do very well out of STV in Wales: winning not only lots of first preferences votes, but also plentiful transfers from supports of Plaid Cymru, the Greens and even the Liberal Democrats. See details here). This is a pity. STV may be the best means of avoiding dual candidacy becoming a running sore in Welsh politics.


  • J Parker

    I don;t see as a lecturer you don;t set up a campaign for STV, your in a perfect position to lobby, youve talked about this for years! Since 2004! It needs a more direct approach than writing in a blog!

  • Roger Scully

    Well, on the campaigning front I think the Electoral Reform Society are doing a decent job, and I have no wish to replace them.

  • Paul Kerslake

    How can it be right or fair if for instance Leanne Wood fails to win her seat and then creeps in by the back door on Plaid Cymrus` say so ?

  • Roger Scully

    I’m sorry, Paul, but I think that’s a rather daft way to put it.

    The voting system we use for Assembly elections, AMS, is explicitly designed to compensate parties that do not win constituency seats. The list seats that they are allocated are not ‘creeping in by the back door’; they are an integral part of an electoral system that seeks to combine constituency representation with at least some proportionality. If parties that are defeated at the constituency level can still win representation, then why should that not apply to individuals?

    There is also a particular concern in Wales. We have a wholly inadequately sized National Assembly. If it is to have any chance of doing its job remotely properly, we need to do everything possible to maximise the quality of people represented in the Assembly. In that context, it is particularly stupid to seek to make strong candidates have to choose between standing on the list or in a constituency. The most detailed review of AMS conducted anywhere in the world in recent years, that in New Zealand, argued strongly that dual candidacy was a positive virtue, helping parties give the best chance of election to some of their best people. And they were absolutely right to argue that.

Leave a Reply

Your email address will not be published. Required fields are marked *