Recent referendums in Ireland and the UK highlight stark differences regarding the place of the referendum as a political institution within the respective overall political systems, and also raise some questions about how a referendum outcome should be interpreted.

The most obvious difference is the highly specified role of the referendum in the Irish political system. Leaving aside the provision for a rejective referendum on a bill passed by the Houses of the Oireachtas outlined in Article 47.2 of the constitution, referendums in Ireland are firmly tied to questions of constitutional change. Every change to the constitution, no matter how minor, requires the assent of the electorate as expressed in a referendum (Article 46). The obverse, while not spelled out in writing anywhere, has pretty much become part of the unwritten constitution: that if a step does not entail a change to the constitution, there should not be a referendum. Thus, while parliament has discretion as to whether or not to put forward a proposal (such as, in 2018, to make abortion more widely available), once it has decided to do so it does not have any discretion as to whether or not to put the matter to a referendum.

In the UK, in contrast, the absence of a definitive codified constitution means that parliament has limitless discretion in invoking the referendum as a means of reaching a decision. The UK did not hold a referendum before joining the EU in 1973 and there is nothing in writing to prevent its leaving without one. Likewise, parliament could have changed the electoral system on its own authority in 2011 rather than holding the alternative vote referendum. As in France, a government that controls parliament is free to decide to put a matter to a referendum for purely partisan reasons: to try to resolve an internal party dispute, for example, or to expose divisions among the opposition.

The meaning of a referendum outcome is also different. In Ireland, the vote is on a precise proposal: that the wording of the constitution be amended in a particular way. Sometimes the constitutional amendment is such that the consequential change is precisely specified and the outcome of a positive vote is clear. On other occasions, as in the ‘repeal the 8th’ referendum in May 2018, the amendment simply frees up the legislature to act as it sees fit, but on that occasion the prior publication of the abortion legislation that the government undertook to introduce if the eighth amendment were deleted ensured that everyone was fully aware of the effect of their vote.

In the UK, in contrast, the absence of rules means that the referendum is in effect an informal political institution. The principle of the sovereignty of parliament means that no referendum can be more than advisory, whereas Irish referendums are binding. At UK referendums the options can be ill-defined; the EU referendum is an archetypal example. There was no definitive statement as to what ‘Leave’ actually meant, leaving scope for bitter disagreement over whether any particular exit agreement, such as what has become known in the UK as ‘Theresa May’s deal’, or some arrangement that entails remaining within a customs union, amounts to ‘really’ leaving.
David Cameron’s model for the 2016 referendum was evidently Labour’s pledge in the second 1974 election to hold a referendum on leaving the EU if Labour were returned to power, something that was also in part a response to internal party divisions, and in that referendum too the details of the ‘Leave’ option were not spelled out. They did not need to be given that in the referendum the following year Remain was endorsed by a 2–1 margin by the electorate, but a variety of factors ensured that this comfortable margin in favour of the status quo was not repeated in June 2016. The 1975 result reduced the salience of the issue in British politics for a while, but Cameron, motivated at least in part by a desire to protect his party’s fortunes against the threat from UKIP, and unable to foresee that by the time the referendum took place Labour would be led by a lifelong Eurosceptic, saw his initiative trigger a number of unintended consequences. With the benefit of hindsight, the 2016 referendum should have specified that, in the event of a majority in favour of the principle of ‘Leave’, the terms of the final departure should go back to the people for approval or rejection, with the latter meaning the continuation of the status quo.
The model would have been the New Zealand referendums on adopting a new electoral system in the early 1990s. In 1992 New Zealanders voted heavily (85–15) in a consultative referendum in favour of the principle of changing from SMP (first-past-the-post) to a PR system, but did not commit themselves to any specific version of PR. The following year, after further deliberation, there was a binding referendum with a straight choice between the status quo and one specific alternative, namely mixed-member proportional, as used in Germany. The majority voted in favour of change but by a much narrower margin, 54–46. This two-stage approach to constitutional reform avoids the risk of endless and unresolvable debate as to which precise options are mandated by a vote in favour of a broad principle such as ‘adopt a PR electoral system’ or ‘leave the European Union’.

Two further aspects of recent Irish and UK referendums raise interesting normative issues. One is what policy outcome is mandated by the result: the policy endorsed by the majority, or the policy favoured by the median voter? In the UK, commentators (not necessarily only Remain supporters) have observed that given the narrow winning 52–48 margin, proponents of a no-deal or ‘hard’ Brexit cannot credibly claim that their views have been endorsed by the electorate. The median voter was only just on the ‘Leave’ side of the line, and can be taken to be in favour of the softest possible Brexit. Should this place an obligation on MPs to support such an arrangement, or are they entitled to press for no deal or a hard Brexit?
The same issue arose after Ireland’s May 2018 referendum on removing the constitutional restriction on abortion provision; those on the losing side, such as Peadar Tóibín, have argued that instead of simple implementation of the wishes of the majority, account should be taken of those who voted against the proposal. While the terminology of the median voter is not invoked, the implication is clearly that in some sense it would be more ‘just’ to implement the policy favoured by the median voter than that chosen by the majority. While the median voter after a 66–34 result, as in the abortion referendum, is clearly well on the majority side of the line rather than close to the ‘undecided’ point as in the Brexit referendum, the underlying normative question is the same: should a referendum outcome be interpreted as a mandate to implement the majority outcome, or the position favoured by the median voter?

The second normative issue raised by the Brexit referendum in particular enters the controversial territory of whether all votes should be equal or whether votes should in some way be weighted according to how much of a stake each voter has in the outcome. One striking if not indeed disturbing aspect of the result was that attitudes were strongly related to age: younger voters favoured Remain, older ones Leave. According to the YouGov exit poll 75% of those in the 18–24 age bracket, and 56% of those aged 25–49, favoured Remain, while only 44% of those aged 50–64 and 39% of those aged 65 or more did so. It would hardly be an exaggeration to say that older voters, whose lives were not going to be much affected either way by the outcome, took away from younger voters choices about their future. The pollster Peter Kellner generated some controversy earlier this year when he reported that by his calculations, given the death rates among older voters and the number of younger voters newly on the electoral register, the Leave majority was shrinking by about half a million per year and that 19 January 2019 was the pivotal day when, if every 2016 voter voted the same way again, a fresh referendum would produce a Remain majority.
In this the Brexit referendum was reminiscent of Ireland’s 1983 ‘pro-life’ referendum, in which the more conservative sections of society, aware that younger voters were much more likely to be liberal, mobilised to pre-empt any moves to legalise abortion. In both cases an older generation was conscious that if there was any delay in copper-fastening the decision, it might soon become ‘too late’, as the votes of younger citizens would loom steadily larger in the electorate. The victory of the pro-life movement in 1983 succeeded in preventing any significant change in abortion law for over 30 years; though, as argued in a previous post, its prohibitionist absolutism, by closing off the possibility of incremental reformist measures, probably meant that as of 2019 Ireland’s abortion regime is more liberal than if the 8th amendment had never been inserted. The 1983 decision of the Irish electorate could be and eventually was overturned, whereas the 2016 Brexit decision was intended to be final; if the UK does leave the EU on foot of the referendum, it is hardly likely ever to be welcomed back, certainly not on the same terms as those it enjoyed until departure. The issue of whether extra weight should be given to the votes of younger citizens, at least on issues where the consequences of a decision are long-term, is one that merits consideration.