Wednesday, 12 August 2009

All about primaries

Progress have launched a campaign Prime Time to try to turn the talk about using primaries to select Labour candidates into action.

As part of this, Will Straw has written a new paper, extending the argument for primaries which he and Nick Anstead made in their Fabian pamphlet 'The Change We Need'.

The paper seeks to address criticisms including that made by Stuart White here on Next Left.

Jessica Asato writes on LabourList, addressing the most common objections to primaries. Long-standing advocate of primaries Anthony Painter has blogged on the campaign too.

I have also written a short commentary for Progress. At present, there are hopes and fears about what primaries would mean. I think we should hold some pilots and experiments to test these, so we can have a more informed debate after the next election.

It would make sense to use the opportunity of the increased number of late retirements to experiment with some late selections in this Parliament.


If Labour is considering primaries, then there is an especially strong case for a primary to select the next London Mayoral candidate. David Lammy, having argued for primaries in a Fabian lecture over a year ago, wrote in the Standard on Friday advocating primaries for the London Mayoral selections ahead of 2012.

Lammy wrote:

Imagine if the parties held a primary election in every London borough. Candidates would have to campaign across the capital, building momentum, borough by borough, community by community. No longer could politicians get away with just campaigning in their heartlands. No longer would Labour voters in Hillingdon or Havering feel ignored by those at the top, since their votes would count for just as much as a Labour voter in Hackney. And yes, it would work the other way round, too, with Tory voters of Newham having their say in their selection.

The Guardian's Dave Hill saw this as a Lammy bid to run but also wrote on the merits of London primaries as perhaps the best way to engage in and resolve the party's debate about its London strategy.

As open a selection as possible in the 2012 Mayoral cycle could both maximise Labour's chances of retaining City Hall and be one of the most important ways Labour could start to engage with the emerging 'movement politics'.

In selecting a candidate for a directly elected position - where there is preferential and so cross-party voting - a broad range of London Labour opinion ought to be able to agree to the principle of an open contest. Indeed, if Ken Livingstone does confirm that he will run again, he would begin with considerable advantages in profile and name recognition. But I would imagine that those from the other wing of the party who are advocating primaries would have to say 'let the best man or woman win', rather than adopting an instrumental approach depending on whether they expected to do well.

But I think it would be a particularly good idea to build support on this now before the debate is dominated by which specific possible candidates might gain from such an approach - and to avoid the shambles and stitch-ups of Labour's London selection in 2000.


But perhaps primary momentum would carry us even further than that. Tom Harris says the 'inevitable consequence' of primaries would lead political parties to use a primary to elect their leaders too. He intended this partly as a warning as to whether the argument for primaries stacks up, while noting parties might find the attractions in the idea too.

Andrew Sparrow has written about this for the Guardian. He picks up on a comment which I made on Harris' thread, I expect the Conservatives would have little difficulty in replacing the current members' run-off of the final two candidates with an open vote among those who register to participate. I predict they will hold a leadership election on that model within the next 10 years. Labour would find that rather more difficult, because of its electoral college and federal structure.

There are certainly barriers to change - and questions which the primaries campaign needs to answer. But they are asking some important questions of their own too.


Mike said...

I support closed primaries, where party members can vote for a PPC.

But open primaries would just make elections into such a long and arduous process where government is seen as this far away and dangerous place, that we could end up with an American style anti-government culture, which would make ending child poverty and funding public services much harder.

Stuart White said...

I've had a quick read of Will's new pamphlet. I'm still unconvinced that Will has taken the full measure of the criticisms of 'open primaries' as illiberal and bad for democracy.

In the section of his paper addressing criticisms, he doesn't directly address the criticism that open primaries threaten freedom of association by opening up candidate selection to the general public (who may or may not share the values of the party). This concern is addressed implicitly later in the paper where Will suggests a 'semi-open primary' as a possible model: would-be participants are required to sign up to a statement of party values before they can vote in the primary. This would go some way to address the concern. But Will also leaves open the possibility of a purely open primary which takes us back to the original problem - we would just have to hope that those voting in the primary share enough of a party's values not to undermine it as an effective community of shared belief.

Will does directly address the criticism that open primaries risk being bad for democracy by pushing candidate selection towards the preferences of the median voter. His argument is that it only pushes selection closer to the median voter among supporters of the party.

Two points on that. First, if the primary is purely open, then it isn't true that it will only push selection closer to the median voter among a given party's supporters. The selection, by definition, is open to all voters.

Second, if party supporters tend to have views closer to the median voter in the population at large than party members, then it will still be the case that opening up selection to party supporters, even under a relatively closed primary system, will tend to push candidate selection closer to the views of the median voter in the population as a whole.

This, of course, has an upside. But it has a downside, too. In the long-run it will tend to reduce the scope and depth of democratic deliberation as candidates all get drawn from, and reflect, the centre ground.

Take one example: inheritance tax. This is an issue where we need a social democratic party to get out there and make the case against the current grain of public opinion. It may well be - I stand to be corrected - that a majority of Labour supporters currently share the widespread public opposition to inheritance tax. It should be the job of party members to work to win over party supporters, as part of the job of winning over the wider public in general. But wouldn't even a relatively closed primary system work against this? Wouldn't the pressure be on to reflect back existing public attitudes rather than to challenge them where they need challenging?

Will Straw said...


Thanks very much for engaging on the substance of my paper. Both your points about freedom of association and democracy are interesting but only valid, I believe, in the unique case of a free-for-all open primary where anyone can vote. This is only practiced in the US state of Wisconsin and I oppose it.

There is a danger of getting bogged down in the definitions since my preferred option can be referred to as either semi-open or semi-closed and these terms mean different things to different people. Instead, let me set out what I see as the problem with free-for-all open primaries and the rules that I would put in place to address them.

In a primary system where anyone can vote regardless of their values, the result could become diluted (ie approach the median voter) and become open to sabotage. I think you can avoid this with (i) a primary election day for all parties where voters have to choose which party to vote for OR (ii) a Labour-only primary election day where people who want to vote have to register first. The leap of logic I make here is that members of the public just won't bother to register if they don't share Labour's values. The feared "entryism" has not happened in the US nor did it happen in Totnes.

Your point on IHT is fascinating and raises a good question about how we avoid candidates playing to the gallery in primary elections. Progress are advocating Labour members' choosing the short list as the Tories did in Totnes. I didn't cover this in my paper but can see that it could be used to weed out candidates who oppose IHT and other key Labour policies.

Thanks again,


donpaskini said...

I am really unconvinced by the worry about the median voters.

If there were a Labour primary in Hayes and Harlington, John McDonnell would walk it - if there were a London wide primary Ken would win, if there were a Birkenhead primary, Frank Field might well get beat by a local leftie.

Swing State Project has some excellent work comparing representatives on a left-right scale to the electoral make-up of their district, and it certainly isn't the case that you only get progressives winning primaries in progressive districts - organisation seems to matter a lot more than ideology.

Stuart White said...

Thanks for these comments, Will, which do go further to allay some of my anxieties (and maybe I would have spared you the trouble if I'd read the paper less quickly). I think the idea of party members controlling the short list takes some sting out of both criticisms.

Also, to engage in a bit of self-criticism, it occurs to me that in one sense the freedom of association criticism is misplaced - or, rather, it might not be quite right to state the criticism in these terms.

Are we talking about the state requiring parties to hold primaries, or parties volunteering to do so? If the latter, then, strictly speaking, the primary is an exercise of freedom of association, not a curtailment of it. It might still be criticised, if the design is wrong, as a strange, self-denying exrecise of freedom of association ('Come and pick our candidate for us - we are not worthy!'), but its not a restriction of freedom of association by the state.

I still worry that in an open primary - even when all parties hold them on the same day and you can only participate in one - there is a possibility of candidate selection being affected by non-supporters. But you are right that the extent of this problem is an empirical question and if the empirics suggest its not usually a big deal, then maybe...its not a big deal.

Worth keeping an eye on, though.

Sunder Katwala said...

Will, Stuart, Don

Thanks for comments and interesting discussion.

Stuart, In replying to yr initial post agin primaries, I thought that the freedom of association argument could work that way around, ie for the opportunity to choose to open the process up, if chosen by the party rather than mandated by the state.

It seems to me that the question of whether primaries would lead to a greater diversity or rather more homogenity of candidates (in terms of their views, and in terms of the diversity of candidates in other ways) is an unknown, as we have very little evidence as to how voter preferences in Britain would operate if a wide electorate had a choice between candidates of the same party.

Stuart White said...

Sunder: I'm sure your valid point about freedom of association filtered into my head, got filed away, and finally came back into my consciousness this lunchtime!

On your last point (and also Don's point), yes, I agree it is an unknown. I should have stated my criticism in terms of what could happen rather than in terms of what will happen.

I am persuaded that its worth experimenting with primaries and seeing what happens.