The Brexit referendum represents a constitutional turning point for many reasons and it has shown how referendums can be dangerous for the British constitution. According to Claudio Martinelli (University of Milan-Bicocca) British institutions should accept a fundamental distinction between two types of referendums: those that allow the electoral body to choose between two clear and defined results, and those that instead involve an asymmetry between the two options or, worse, two indeterminate results.

Since the beginning of the 1970s, referendums have been not only accepted as a constitutional tool to make specific political decisions, but have been widely practised, at any level of government (national, regional and local), and always on sensitive issues. The Westminster model itself is changing because of the frequent use of these popular consultations.

The 2016 referendum is therefore only the last link in a long chain, but it also represents a turning point in the use of referendums in the UK: because of its assumptions, the outcome, the political clash it has caused, and the constitutional cleavages which it has opened. The statute implementing the Brexit referendum clearly shows the political dimension of the whole operation. During the legislative process, the Conservative government rejected all amendments that could have weakened David Cameron’s bargaining power at the negotiating table with the EU. The Prime Minister’s political calculation was that if the EU institutions really feared a negative outcome, they would be persuaded to accept most of the requests made by the UK in order to redefine its membership. His strategy was correct: the Cameron-Tusk Agreement was largely receptive to the wishes expressed by the PM. The EU feared a domino effect following a negative referendum in the UK. On the basis of political strategy alone, voters were faced with a referendum founded exclusively on the principle of majority.

But the proposed alternative was basically asymmetrical: in the case of a Remain vote, the Cameron-Tusk Agreement would come into force according to a precise and pre-established legal agreement; in the case of a Leave vote, there were no predictable effects and proceedings to comply with the voters’ will, particularly when it came to the procedures and terms of the withdrawal. A vote to Leave had been a strong choice of direction, that had to be respected by the constitutional bodies, but with undetermined political and legal consequences. Ironically, only after the referendum, it was realized that the Leave victory, rather than providing an answer, had opened up several disturbing questions – many of which are still open three years later.

The reference to the legal requirements of a referendum, established in every statute, is important to understand the characteristics of its consequences. Indeed, the Supreme Court of the United Kingdom in a crucial passage of the Miller judgment stated that the effects of each referendum depend on the provisions contained in its founding statute [par. 118]. This statute can regulate the legal consequences of popular consultation or can refrain from doing so. In the case of the 2016 referendum, as in that of the 1975 EEC referendum, the law did not stipulate the consequences of the outcome [par. 119]. On the contrary, in the case of the 2011 referendum on the electoral system, the mandatory profile of the popular vote had previously been regulated by Parliament, which had taken steps to write a law establishing the electoral rules of the new AV system. But in the two consultations on Europe, no such specific arrangments could have taken place, since the concrete consequences of those decisions were not up to Westminster alone, but depended on bargaining with the European institutions. Similarly, the Scottish independence referendum of 2014 certainly could not determine the nature of the new relations between Scotland and the United Kingdom. The statute of December 2015 at least could have established the concrete ways of invoking Article 50 of the Lisbon Treaty (TUE) according to the British constitutional law, but this was avoided in order not to add any details to the substance of the choice between Leave and Remain.

It would be appropriate if British institutions would accept this fundamental distinction between two types of referendums: those that allow the electoral body to choose between two clear and defined results, and those that instead involve an asymmetry between the two options or, worse, two indeterminate results.

The British constitution and, within it, the Westminster model, has always been in constant evolution. The acceptance of plebiscitarianism in constitutional practice is an aspect of this change. The vision expressed by Dicey in 1890, which considered the referendum compatible with parliamentary sovereignty, has been confirmed in recent decades. Many celebrated referendums have provided a further boost to the evolution of the British constitution, making it more modern and articulated. In particular, those concerning devolution and local autonomy: popular involvement in these political decisions has also strengthened the subsequent parliamentary legislation. However, the Brexit referendum has shown that the peculiar characteristics of the British constitution, unwritten and not codified, require that Parliament uses this instrument accurately and within limits defined in a statute. To prevent the danger of destabilising questions, it would be advisable to exclude the possibility of implementing referendums of a merely political nature, where the implementation of the result may be hard to determine.

We must not forget that the British system already provided important examples of referendums that are removed from pure political discretion and aim to protect constitutional structures and, therefore, ultimately, to safeguard citizens’ rights. An important example is the Scotland Act 2016. This law makes the devolved institutions permanent, thereby erasing the dominus role of the Westminster Parliament vis-a-vis the Scottish parliament and government. It states that ‘the Scottish Parliament and the Scottish Government should be abolished on the basis of the decision of the people of Scotland voting in a referendum’. We find similar rules in Wales Act 2017. Other positive examples of the electorate’s involvement can be found in the Local Government Act 2000, in the Localism Act 2011 and in the European Union Act 2011.

In short, the referendum as such is not only compatible with the modern British constitution, and the recent evolutions of the Westminster model, but it can enrich British democracy if it is used correctly, that is, not only as a political strategy but in the framework of constitutional procedures aimed at consolidating institutions and their relationship with the public.

This post represents the views of the author and not those of the Brexit blog, nor the LSE. Image is licensed under the Creative Commons Attribution-Share Alike 4.0 International

Claudio Martinelli is Professor in Comparative Public Law and in Parliamentary Law at University of Milan-Bicocca. He is the author of “Diritto e diritti oltre la Manica” (il Mulino, 2014) and “Il referendum Brexit e le sue ricadute costituzionali” (ed.) (Maggioli, 2017).