The Plebiscite Problem: Could another referendum solve Brexit?

Dr Simon Kaye

In a recent article and discussion on BBC News I argued that the intractable political situation created by Brexit—with a seeming majority in the Commons against all three tabled outcomes, explicitly framed as a three-way choice by the government—plausibly reflects a deep paradox of preferences in the wider public as well.  With three choices now in serious contention there is a huge literature available to explain the various perverse democratic outcomes that could emerge from any vote.  With a possible majority opposed to each mooted outcome in both the Commons and the country we are also in a situation where more people will be disappointed than pleased, no matter the final decision.

Under these circumstances the government may be tempted over the coming days to follow the line of the 'people's vote' campaign and set up a new referendum to resolve the deadlock. After all, the public will expressed in an act of direct democracy can only plausibly be revised or undone by a similar moment of direct, plebiscitary participation.  In this piece, I set out the structural challenges entailed in the referendum escape route and the risks these would entail in presenting its results as democratically legitimate.

A lot of critical responses to this idea have been thoroughly rehearsed: that a second referendum would involve a difficult and lengthy set-up, necessitating an extension of Article 50; that it would in a sense be undemocratic; that it could deepen rather than resolve the political dilemma over Brexit if it results in a close outcome or has lower turnout than the 2016 vote; that it might so incense leave supporters from 2016 that they are alienated from political participation for good, or even cause some of them to turn to direct action.

Setting all these claims aside, I here expand upon the purely structural difficulties that would have to be overcome for a new referendum to produce the powerful political mandate that it would need to, if necessary, supersede that created by the 2016 poll. It is difficult to define the authority of such a mandate in formal terms, but it nevertheless exists: a second referendum will not resolve anything if it does not enjoy the endorsement of, or at least seem legitimate and non-arbitrary to, the lion's share of those who are asked to cast votes in it.

Unfortunately, there may not be a single approach to a new referendum that would not result in some loss of legitimacy.

A second referendum that only presents two options – say, one between the Government’s deal and remaining in the EU, or between the Government’s deal and leaving with no deal/on WTO terms, will immediately appear to be delegitimised in the eyes of some significant section of the public. The decision over which of the three currently extant alternatives to exclude will always be to some extent an arbitrary one. While a two-option vote sidesteps the risk of a formal Condorcet paradox, it comes with the genuine risk of some voters publicly boycotting the poll. All democratic procedures require some degree of abstraction, of boiling-down, so that complex issues can finally be captured by explicit alternatives to be chosen between. Yet a two-option referendum would probably be an abstraction too far to retain democratic authority.

This suggests that all three options – assuming no additional alternatives emerge into serious consideration in the coming days – should be present on the ballot in some form.  A referendum that presents three options on a simple, single-preference ballot, however, will inevitably result in a ‘split’ in one side of the vote, with the final results likely dismissed as arbitrary. The government’s Withdrawal agreement and political declaration do not sit equidistant from the polarised alternatives of ‘no deal’ and ‘remain’. It is a species of Brexit, and has been understood as such by all but the most ardent leave campaigners. This suggests that leave-supporting voters will be divided between deal and no-deal Brexit, allowing remain to assemble a plurality of votes. This is a fairly ordinary outcome in British electoral politics, but the outcome would be vulnerable to powerful accusations of unfairness, especially if Remain does not achieve an outright majority of support (50 per cent of all votes, plus one).

The easiest way of avoiding the seeming illegitimacy of such an outcome would be a referendum that uses some kind of preference listing system to arrive at a decision, such as the Alternative Vote model that is used for some elections already, but was rejected as the electoral method for general elections in another referendum in 2011. If such a referendum were to be called, we could expect all of the criticisms levelled against this kind of voting to once again get an airing. In addition, such a referendum would risk producing a paradoxical result, if it revealed that the British public’s preferences resemble a Condorcet paradox, or – more likely - deliver an outcome based upon second preferences, which the losing side could find very difficult to accept. Third, this approach could produce an outcome where the ‘Condorcet winner’ – which, according to three recent polls, would be the Government’s deal – would not win, so that some majority’s least preferred outcome ends up selected instead.

The key here is not that there would not be a way of counting the votes to produce a clear outcome. Rather, it is that different, similarly reasonable approaches to casting and counting the votes could easily yield different outcomes: by any definition, any outcome settled upon under such circumstances would be arbitrary.

What’s left? A referendum that breaks the decision down into two separate questions stands the best chance of being as legitimate as the 2016 vote. All options would be presented, and some of the risks and arguments raised by use of a preference-ordering voting system could be avoided. Part one of the ballot could pose the leave vs. remain question, while a second part asks a conditional question: “If the UK opts to leave the EU, should it do so under the terms of the government’s negotiated withdrawal agreement and political declaration, or should it leave with no deal at all?”

But even here there is a risk of a contradictory result – where the public simultaneously votes for ‘remain’ in part one and a hard ‘no-deal’ Brexit in part two, of lower turnout overall or lower engagement with one question than with another, not to mention the difficulty of incorporating any other alternatives that are yet to strongly emerge (a ‘managed no deal’? A withdrawal based on some as-yet-unnegotiated variant of the Norway model?).

On the issue of Brexit, the public is as deadlocked as the Commons – more so than ever with three possible outcomes in contention. A new referendum may seem to be an appealing escape route for MPs, but there is a serious risk that it could simply deepen our political logjam instead.

Dr Simon Kaye is the Research Director for the Morality & Markets and GovernUp Projects at P4MD.