A People's Vote without a referendum

There is an alternative way of "letting the people decide" on an issue where MPs seem incapable of agreeing a coherent policy. The Greeks had a word for it: democracy.

Richard S. Forsyth
4 December 2018

View of Athens from Pnyx, at dusk, 2015. To the left is Lycabettus, to the right Acropolis. Wikicommons/ C Messier. Some rights reserved.

The recent publication of Theresa May's Brexit deal incited a number of zealots in Parliament to throw various toys out of their prams so fast that it is hard to believe they had time to read it. It has also provoked a great clamour for a "people's vote" in another EU referendum.

One of the arguments offered in favour of a second referendum is that the people didn't know what they were voting for in June 2016. Now, it is claimed, we have the details, so voters will be better informed. The idea that more than a handful of voters would read and digest the 585-page withdrawal document, along with supplementary papers on the proposed future trading arrangements between Britain and the EU, shows that this line of argument is essentially spurious – promoted by those who simply want a different result from last time. They might get a different result from last time, but that brings dangers of its own, as many commentators have pointed out, for example Richard Shrimsley in the FT, (October 8, 2018).

Those Greeks

There is, however, an alternative way of "letting the people decide" on an issue where Members of Parliament seem incapable of agreeing a coherent policy. The Greeks had a word for it.

Democracy as implemented by deliberative assemblies of randomly chosen citizens rather than by elected chambers of representatives is very much a minority activity in the modern world, but it has begun to impress a growing number of political scientists with its effectiveness wherever it has been tried (see the writings of Tin Gazivoda, Carole Pateman, and Matthew Taylor, cited below). The idea is to go back to something more like the classical Greek model, in which ordinary citizens had a direct input into political decision-making. The key to dealing with larger populations than in an ancient Greek polis goes by the name of sortition, namely random selection from a pool of eligible voters.

Below I propose an EEC (Extraordinary Electoral College) with a step-by-step procedure that is likely to deliver a verdict that would command more widespread assent than yet another referendum. (For clarity, a number of implementation details, though important, are omitted from this outline – discussed separately in the next section.)

1.  650 people, chosen purely at random, one from each of the electoral registers of the UK's 650 parliamentary constituencies, are selected to decide upon the issue. (What to do about alternatives for seriously unwell or heavily pregnant people or others with good reasons for being unable to attend is discussed later on.)

2.  These people are given 14 days' notice to gather for 8 days (Saturday to Sunday) in a conference centre somewhere in the UK, situated north of Bedford and south of Berwick upon Tweed. In a time of national crisis it is impressed on them that it is their duty as citizens to take this task at least as seriously as jury service. (What to do about payment, accommodation arrangements, finding replacements for them in their workplaces and so on is discussed below.)

3.  When they arrive at the conference centre on Saturday morning they are given the withdrawal document to study and a hotline is made available to a panel of civil servants who have been involved in the negotiations to call on for explanations of difficult passages.

4.  They are left at liberty in their accommodation (with regular meals provided, of course, preferably in a communal dining area) to read and try to understand the document until Monday lunchtime. After lunch on Monday they are each given a multi-choice quiz, previously compiled by the civil servants, to test their comprehension of what they were asked to read and understand. (This test must later be published, with its expected correct answers.) The 50 people with the lowest scores on the test take no further part in the process, with tiebreaks to be decided at random, if necessary. (What to do about anonymity and whether they should depart is discussed below.)

5.  The remaining 600 participants are divided into fifty groups of 12, again entirely randomly, to spend the next four days discussing in their groups the advantages and disadvantages of the proposed deal. Voters are asked to avoid discussion with members of other groups as far as possible. During this period mobile phones and other internet devices must be deposited in a secure holding area during the hours of 0900 to 1900 each day, but the hotline to civil servants should remain open at certain times, e.g. 1200 to 1800 for technical queries.

6.  On the second Saturday morning, all voters are asked to retire individually to their rooms for 24 hours to ponder their choice of all the viable options (in this case let us say the 3 options: May's deal, No Deal & Remaining in the EU) without any internet device and without speaking to other voters.

7.  On the second Sunday morning these 600 citizens vote by secret ballot on the three options available. They will be required to rank them in order from most to least preferred. The overall result should be calculated by adding 1 to the count of each option ranked top and subtracting 1 from to the count of the least preferred option, with zero for the middle option, giving each option an overall positive or negative score. (What to be done about spoilt ballots and suchlike is considered below.)

8.  On the Sunday evening the result is announced and the participants depart to resume their lives.

Why should this command more respect from the wider population than a vote in the House of Commons or a full-scale referendum? There are four main reasons. Firstly, the citizens taking part better represent the diverse range of people in this country than do Members of Parliament. Secondly, again unlike MPs, they have no personal ambitions for prestige or wealth that depend on their decisions. They are simply citizens who have been asked to take their responsibilities seriously to the best of their abilities. Thirdly, lobbyists from large corporations and other vested interests would not have time to exert pressure on the decision makers, who in any case would have nothing to gain or lose from such pressure. Finally, and most importantly, they will have studied and discussed and thought seriously about the choices at stake – something that is simply impractical for voters in a full referendum. (It could also be arranged more cheaply and quickly than a national referendum, though that advantage should not be decisive.)

Devils lurking in the details 

Naturally there are ways that such a procedure can be compromised, so it is important to guard against apparently minor elements of the process than can undermine its integrity. Nevertheless, it should not be beyond the wit of homo sapiens to devise a trustworthy procedure. Some suggestions in this regard are listed below in the same order as the steps outlined above.

1.  Some individuals selected from the electoral registers may have extremely strong reasons for not taking part. They may need urgent medical treatment; they may be in the late stages of pregnancy, or they may even be dead. Thus each constituency must devise a justifiable randomization process along with clear rules about what grounds for self-exclusion are acceptable, as well as a rapid way of selecting an alternative elector if needed. These must be open to scrutiny so that they can be seen to be unbiased. It is vital to keep screening to a minimum. The temptation to stratify by age, gender, socioeconomic or other criteria should be resisted: the nearer to a pure random selection of eligible voters the better.

2.  Participants should be well rewarded (by normal standards, even if not by MPs' standards) for their participation; and their employers, if adversely affected, should be compensated on relatively generous terms. Thus the exercise will cost public money, although a trivial amount compared to the renovation of the Houses of Parliament, for example. In addition, details of the venue, such as eating and sleeping arrangements, can make a significant difference, so expense should not be skimped on this aspect either. Above all, the participants should not find the event stressful. It may be that centres capable of holding events of such a size should bid for what will be a taxpayer-funded operation. It is important that the choice (presumably by civil servants) is transparent, i.e. that the reasons for the choice of location can be laid open to public scrutiny. The geographic limits stated in the previous section, above, are designed to keep the participants away from the delirium of the "Westminster bubble" and arrive at a venue not too far distant from the centre of the UK population.

3.  The Brexit withdrawal document does not contain a summary. It would be helpful for ordinary members of the public to have some kind of executive summary to guide them, but of course the danger then is that the summary will be used instead of the full document. Perhaps a small team of independent scholars could prepare in advance a synopsis that does not tend to bias towards one conclusion or another.

4.  Excluding a small number of people who haven't read or have very poorly understood the central document is controversial. The wider public will accept something of the kind as fair so long as it is restricted only to those who, for one reason or another, aren't able to make an informed contribution. However, the proportion excluded should be small, definitely less than 10 percent of the total; and if a tiebreak among equal scores is indicated, it would probably be better to proceed with slightly more than 600 participants rather than fewer. (Presuming that the official document under scrutiny is in English creates potential disadvantages those for whom English is a second language, but time is not sufficient to provide for speakers of other languages.) It is also very important to guard against groupthink. It would be unnatural to expect a group of strangers with a specific topic at the forefront of their minds not to talk about it at all with their fellows, but arrangements should be in place to emphasize that what is being sought is 600 separate informed decisions not 600 repetitions of some influential person's opinion. Hence fraternization with other participants, other than at meal times, will have to be discouraged.

5.  Hosting fifty groups in surroundings conducive to free-ranging discussion isn't an easy task. Something like a campus with separate meeting rooms will be required. Each group should share thoughts and have the chance to seek clarifications from the civil servants, but too much discussion with other groups might undermine the independence of each discussion, so should be discouraged as far as possible. In western so-called democracies we have become accustomed to, but also disenchanted with, adversarial debate on divisive party lines. The present arrangements are explicitly designed to minimize the risk of polarization into 2 or 3 factions who then devote their energies to disparaging each other rather than seeking solutions to a problem. To encourage freshness, another option might be to reallocate participants to new groups, again randomly, half-way through the four-day discussion phase.

6.  Again, the idea of a period of private reflection is to allow debate to lead on to deliberation, thereby reducing the chances of groupthink.

7.  The Alternative Vote (AV) might be thought appropriate here for a 3-way choice. The reason for advocating a tallying system, such as that described above, is that AV, in effect, stops as soon as one alternative reaches 50% support. The present method results in an overall order of preference, not just a 'winner'. This gives more information about the decision process to politicians and to the wider public. Spoilt ballot papers, as agreed by a panel of trusted returning officers, would be left out of the computations.

8.  Questions of anonymity will need to be considered. It would be hard to guarantee all participants anonymity, but it will probably be necessary to impose strict limitations on whether they should be allowed to give interviews to media organizations and suchlike. Possibly a four or five-year moratorium on revealing their deliberations should be imposed on participants, long enough for the information to pass from news (from which political points can be scored) to history (from which lessons can be learned).

Problem-solving politics 

Even more important than finding an acceptable route through the tangled thicket of Brexit, successful adoption of such an approach would demonstrate direct democracy in action as a more effective way of doing politics. This is something that will be sorely needed if we are to confront the coming climate crisis, (which will make Brexit look like a mere hiccup), without turning it into a catastrophe.


Gazivoda, T. (2017). Poles are making democracy work again in Gdansk.

Pateman, C. (2012). Participatory Democracy Revisited. Perspectives on Politics, 10(1), 7-19. doi:10.1017/S1537592711004877

Shrimsley, R. (2018). A second Brexit poll is a bigger risk than leaving. Financial Times, 8 October 2018.

Taylor, M. (2018). Is Deliberative Democracy Key to a 21st Century Social Contract? RSA Publications.

Who is bankrolling Britain's democracy? Which groups shape the stories we see in the press; which voices are silenced, and why? Sign up here to find out.


We encourage anyone to comment, please consult the oD commenting guidelines if you have any questions.
Audio available Bookmark Check Language Close Comments Download Facebook Link Email Newsletter Newsletter Play Print Share Twitter Youtube Search Instagram WhatsApp yourData