Sir: Peter Jones (25 May) is right to draw an unfavourable comparison between ancient and modern democracy, but he is focusing on the wrong institution. The Athenian council was merely the secretariat for the general assembly, and the legislation passed by the assembly was often as erratic as modern referenda. After the restoration of democracy in 403 bc, legislation was entrusted to nomothetai — large randomly selected juries that, unlike modern parliamentarians, were obliged to listen to the arguments of well-informed advocates for and against the proposed law before deciding the outcome by secret vote
If David Cameron wants the people do decide. . .
This proposal, written in response to André Sauzeau’s proposal for minimal reforms, was submitted as an article (see below) and originally accepted for publication by the Spectator, but ended up cut down into a short letter. The Spectator website has a comments section, so suggest we use that as an opportunity to kick-start the conversation on sortition there, rather than commenting on this forum.
Put the EU on Trial
By Keith Sutherland
The answer to Britain’s EU problem is not a public referendum, it’s an adversarial judicial inquiry in front of a large citizen jury, selected by lot
The success of UKIP in the recent elections has led to unprecedented soul searching within the political class in general and the Conservative Party in particular, with no fewer than three former cabinet ministers arguing that Britain should leave the EU. David Cameron has committed the party to a referendum on EU membership, but the public often just use referenda as an excuse to put two fingers up to the government. There is an urgent need to find a more reliable mechanism to allow the people to make a well-informed decision on what is arguably the most important issue in contemporary politics.
Judicial inquiries have, on the whole, a good procedural record – the Hutton Inquiry into the role of the BBC in the death of David Kelly was widely praised for its well-balanced and highly public proceedings. The problem was the lack of democratic participation – there being no jury to decide the outcome. The inquiry verdict (guilty) was entirely down to m’lud – in this case a lord justice who had spent most of his time in Northern Ireland’s Diplock courts and whose conclusions were coloured by his own thinly-concealed contempt for the media. The Leveson Inquiry could be criticised along similar lines, and its additional failing was the adoption of the continental system of the examining magistrate, who summons his own witnesses and permits no cross examination. Although the press were on trial (again), there were no lawyers for the defence. Leveson was unimpressed by Private Eye editor Ian Hislop’s suggestion that the inquiry might like to call some members of the public and ask them why they bought the News of the World. “I don’t think”, retorted the noble lord, “we’ll necessarily do it in quite that way”. Lord Justice Leveson wrote his own report (albeit with a little help from Hacked Off) – how different the outcome of both inquiries might have been if they had followed normal trial procedure, with adversarial exchanges followed by a decision by twelve randomly-selected ‘good men and true’ (in fact the jury would have to be much larger in order to be considered representative, but more on that later).
Why not adopt a similar approach – an adversarial judicial inquiry before a randomly-selected citizen jury – as an alternative to a referendum? There is nothing remotely new about the juridical approach to policy-making. Athenian democracy – invented by Cleisthenes in 508 BC – is normally categorised as a form of ‘direct’ government, as every (male) citizen was entitled to attend and vote in the general assembly (ecclesia). But this meant poorly-informed snap decisions under the influence of charismatic demagogues, which led to the emptying of the public coffers and a disastrous defeat in the Peloponnesian wars. So in 403 BC the Athenians established a system of legislative courts (nomothetai) and every new law had to run the gauntlet of adversarial debate in front of a jury comprised of several hundred citizens selected by lot (they even devised a special machine called a kleroterion to make the draw). The proposer(s) of the new law would argue for it and the Assembly would appoint advocates to oppose it, but the outcome was decided by the randomly-selected jury voting in secret.
Although Aristotle was hostile to government by popular assembly he praised this ‘wisdom of crowds’, concluding that, under the right conditions, ‘the many’ (hoi polloi) judge certain matters better than individuals or small groups. This is because of the sheer diversity of perspectives involved: “for each of them may possess some part of goodness and wisdom; and when they get together, as the mass may be a single man with many feet and many hands and many senses, so it may be with their character and thought.”
The truth of Aristotle’s conjecture was proved in 1785 by the Marquis of Condorcet’s Jury Theorem, according to which a jury is increasingly likely to converge on the ‘right’ answer as its numbers increase (assuming independence of judgment and a minimal threshold of competence). And there is a wealth of modern research evidence to suggest that the ‘cognitive diversity’ produced by large randomly-selected juries is the best way of deciding important issues. Election, by contrast, tends to select people of similar backgrounds (lawyers, Oxbridge PPE graduates, policy wonks etc) who are often prone to the malign ‘groupthink’ effects that are characteristic of homogeneous assemblies. The psychologist Richard Tetlock’s demonstration that ‘expert’ political judgment is little better than average has now been widely replicated.
Stratified random selection is also the best way of establishing an assembly that represents the entire population ‘descriptively’ – a ‘portrait in miniature’ of the entire citizen body, as US founding father John Adams put it. A randomly selected group of a few hundred would be almost exactly 50/50 male/female and would be an accurate reflection of age, socio-economic category, political leanings etc of the whole citizen body to a level of statistical confidence of over 90%.
And this is not just a matter for political theorists and statisticians. Professor James Fishkin’s Centre for Deliberative Democracy at Stanford University has been conducting social science research experiments along these lines for over twenty years in countries all over the world. Fishkin’s Deliberative Polling (DP) experiments demonstrate that ordinary citizens are perfectly capable of deciding complex issues when presented with balanced information and advocacy and empowered to deliberate together over a period of a couple of days. (Fishkin’s notion of deliberation is derived from the Latin liber (weight) so the role of the citzen jury is to silently ‘weigh’ the competing arguments before registering their vote.) The success of the experiments has led him to claim that the deliberative verdict of the microcosm in the DP indicates what everyone would think under ideal circumstances: “the microcosm offers a proxy for the much more ambitious scenario of what would happen if everyone discussed the issues and weighed competing arguments under similarly favourable conditions.”
But then why not provide balanced information and advocacy to everybody prior to a referendum? The problem that plebiscites share with electoral democracy is ‘rational ignorance’ – it makes no sense for voters to take the considerable time and effort required to inform themselves properly on election or referendum manifestos as each individual vote carries negligible causal weight. As NYU political scientist Russell Hardin memorably put it, “having the liberty to cast my vote is roughly as valuable as having the liberty to cast a vote on whether the sun will shine tomorrow.” The other problem with referendums is that voters often decide simply not to answer the question, and merely return an overall verdict on the government
Fishkin’s DPs have mostly had an advisory role; the only time that the results were automatically adopted was in the People’s Republic of China. In 2005 the Communist Party leadership in Zeguo province commissioned a Deliberative Poll of a random sample of 235 citizens to determine public priorities in infrastructure spending. Even though the popular preferences indicated after the information and deliberation sessions were entirely contrary to their own, the party leadership duly implemented them, leading Fishkin to salute the Chinese for developing an entirely new model of democracy that “may set an example for public consultation in many settings around the world”.
If we don’t want to be outdone by the Chinese in democracy as well as everything else, we would do well to look seriously at these experiments, and where better to start than with an issue that has caused so much heartache to the Conservative Party for so many years. Who would the advocates be for an adversarial inquiry on Britain’s membership of the EU? The arguments for and against in the 1975 referendum were coordinated by cross-party umbrella groups and the same would be the case for a new referendum, public inquiry or Deliberative Poll. There would be no shortage of those seeking to advise and inform.
But however good the advocacy, referendums indicate, putting it charitably, poorly-informed preferences. Much better to assemble a representative microcosm, or ‘mini-public’ as veteran Yale political scientist Robert Dahl terms it, and allow them to weigh the competing arguments as a proxy for the rest of us. No doubt political jury service would be just as tedious as its judicial namesake, so those of us who fail to draw the lottery ticket would be the lucky ones.
Keith Sutherland’s books The Rape of the Constitution? and A People’s Parliament are published by Imprint Academic.