Different Process, Same Outcome? The Representative Problems of Within-Party Sortition

SUMMARY

• A recent article in Political Quarterly argues for a ‘sortition of candidature’, where parliamentary candidates would be chosen at random from the membership of political parties.
• However, this scheme would not result in a socially representative Parliament, because parties are not socially representative.
• Drawing on data from the Party Members Project, while some demographic groups would be better represented than at present under within-party sortition, in other ways, things would become less representative.
• Unless under-represented groups join political parties en masse, this seems likely to change.

Candidate selection was once described as the secret garden of politics. A recent article in Political Quarterly advocated pulling the garden up by the roots and replacing it with a version of the Athenian practice of sortition. Under the scheme proposed for ‘sortition of candidature’, parliamentary candidates would be chosen at random from within the membership of political parties, who would then battle it out at the ballot box against rivals from other parties, who would also be chosen by lot.

It is an intriguing idea, but it would likely undercut one of the traditional benefits of sortition: that it generates institutions that are – for good or ill – a mirror of the wider population.

The key problem is that political parties are not socially representative bodies. Their members tend to be older, whiter, more male, more highly educated, and more middle class than the electorate. Selecting by lot from within an unrepresentative population is unlikely to generate a representative end product.

We show this using the most recent surveys of British party members, carried out in July 2024 by the Party Members Project.

The table below shows estimates for the composition of a parliament elected under modified sortition, assuming the electoral outcome and party membership were as they were in 2024.

The first three columns of data show the current situation: the Commons is on average older than the public, has a much higher proportion of graduates, but a lower proportion of those in social grade DE (the lowest socio-economic classifications in the UK, combining semi-skilled/unskilled manual workers with pensioners, casual workers, and those relying on state benefits), women and those from minority ethnic groups. These differences will come as no surprise to anyone who knows anything about the institution – and it is this that sortition is supposed to remove.

The next columns show the estimates for a sortition by candidature parliament, given the same party comp osition as the Commons elected in 2024, but estimating what MPs’ social composition would have been if they were drawn at random from parties’ memberships.

In some cases, things do get more representative under within-party sortition – fewer graduates, more people from social grade DE, slightly more women – although in all three cases these groups would still be underrepresented compared to the wider public. But in two cases, things get less representative: this alternate parliament would be even older and whiter than the current one.

We can show that this outcome is not a one-off, driven by the election result in 2024; we have modelled a variety of different electoral outcomes; they all mostly produce similar results.

So modified sortition would not produce a representative parliament; it would produce an unrepresentative parliament that was unrepresentative in exactly the same ways as the current one, albeit in slightly different proportions.

For within-party sortition to generate a descriptively representative House of Commons we would need to see a significant influx of new party members – of currently underrepresented groups – attracted by the possibility of elected office. This seems to be very unlikely, based on what we know about the public’s appetite for political engagement. There are plenty of reasons why individuals might not want to be an MP: loss of privacy; career disruption; the damage to family life or security; or a myriad of other objections. Sortition within parties could give people a strong reason not to join parties.

Out of 2003 people in a national representative sample surveyed at the same time as the party members, who were asked how keen they were to stand for election on a 0-10 scale, 71% gave a score of zero. Just 35 respondents (1.7%) gave the most enthusiastic response of 10. And asked if they had actually sought to be a candidate for elective office, there was just one positive response. One.

Even many of those currently in political parties aren’t keen. Even at election time, a considerable proportion do no work at all for the party – and only small minorities say they would be interested in being a candidate. When asked if they would like to become a candidate, the percentages giving the most enthusiastic response of 10 varied between 7% (Labour) and 12% (Lib Dems). The percentage of members who said that they had tried to become a candidate at the 2024 election was even smaller, ranging from 0.5% (Conservative) to 3% (Green).

And party members have an option not traditionally open to participants of sortition: they can exit. They can cease to be members of the party. It seems to us that any such sortition scheme would have to include some form of opt-out, allowing members not to be chosen as candidates, else it would lead to a further decline – and probably quite a dramatic decline – in party membership. Yet any opt-out scheme reduces further the benefits of sortition. Rather than everyone, the pool of candidates would now just be the most politically ambitious members of unrepresentative political parties, similar to now.

Indeed, there is every chance that the scheme might lead to parties haemorrhaging members, as the less active members jump ship – unless some form of opt-out provision was included, but that would largely negate the whole point of the exercise.