Why the EU Referendum Provides No Genuine Democratic Mandate to Leave

Leaving aside all substantive issues to do with the social, economic and cultural background to the UK’s EU Referendum of 23 June 2016 and its result and future implications, this article explains why this referendum could and should be considered to be a deeply contested event on purely formal political, legal and constitutional grounds.

The article sets out a number of observations on appropriate conduct of referendums in representative democracies in general, as well as a number of more specific observations on the UK’s EU Referendum in particular.

The article has three sections:

  1. Direct and representative democracy
  2. Legal status of referendums
  3. Thresholds for referendums



  1. Direct and representative democracy. Most modern democracies are almost always representative democracies, even when they occasionally opt to use a direct democratic instrument such as a referendum. When they do, extreme caution must be taken. If a referendum is advisory (with no legal power), parliamentary discretion over interpretation and implementation of a result remains supreme. If binding (with legal power), precise conditions under which Parliament consents to renounce discretion must be specified, most especially in relation to the valid threshold for a majority vote.
  2. Legal status of referendums. Referendums should be presumed advisory unless stated as binding and criteria for implementation specified. Only statute passed by Parliament for a Referendum has legal primacy, not government or governing party literature or ministerial spoken word. In both advisory and binding referendums, franchises for participation must extend to all citizens likely to be affected by a result on a long-term basis, inasmuch as the decision is for a lifetime. Advisory referendums cannot overrule an existing legal state of affairs, unless with the agreement of Parliament to all key steps involved in the process. Binding referendums can do so if, and only if, the power authorising implementation of a result is sovereign.
  3. Thresholds. Binding referendums should specify a threshold and a minimum turnout requirement. For major constitutional change, many modern democracies require a super-majority of votes cast and minimum turnout or, alternatively, set an overall percentage threshold for votes required by eligible voters. Consent of a super-majority of representatives and consent of devolved administrations is also typically required. Long-term change to a long-term status quo places the onus of a result heavily on the vote for change. A status quo is already directly known, whereas change to it is not. Therefore expression of the strength of a desire for change in a vote must be that much more powerful and decisive.



Direct and representative. When a representative democracy opts to deploy a direct democratic instrument, such as a referendum, extreme care must be taken. If not carefully constructed, and especially when binding, referendums are liable to problems long referred to as problems of the “tyranny of the majority”, where minority interests are disregarded.[1]  This may occur particularly when the majority in question is a marginal one or is no more than a majority of votes cast rather than a majority of eligible voters (i.e. when diminished turnout is ignored).  Modern representative democracies have devised systems to counter such problems, often known as parliamentary “checks and balances”, designed to preserve political pluralism and to counter elements of transient or incoherent emotion that may occur in a popular vote with long-term consequences and without possibility of considered redress at a periodic general election. These checks and balances are expressed further in the competence of representatives not only to represent voter opinion but also to judge in the general national interest, including sometimes with the assistance of an advisory upper chamber (the House of Lords in the UK).  Representatives do not automatically implement popular opinion.  They retain the right and the duty to take decisions in the general national interest, even if unpopular with some voters.

Parliamentary sovereignty. In representative democracies, sovereignty of the people is represented in Parliament. It is never delegated to the people directly and absolutely, other than in a binding referendum.  In its sovereign representation of the people, Parliament is supreme above (i) the Crown or the executive state and (ii) the people themselves, other than in a binding referendum.  Abrogation of Parliament’s sovereign representation of the people, in either of these two regards, may be considered, in any degree of severity, despotism or dictatorship.

Failures of direct democracy: specific. When a representative democracy decides to pass sovereignty temporarily to the people on a direct and absolute basis, as in a binding referendum, it must specify, very clearly, a vote threshold and a turnout threshold for a result to be implemented in law (see further below). If a representative democracy conducts a merely advisory referendum, Parliament must meet speedily after the poll in order (i) to interpret the result and (ii) to decide on its advisability for implementation in law, with due regard to the general national interest.  Any further implications of a result should not be considered before this primary parliamentary meeting has taken place.  If a significant length of time is allowed to pass before such a meeting occurs, the status of the result in the mind of the people becomes unclear.  For some, it may appear increasingly set in stone; for others it may appear increasingly doubtful.  An advisory referendum cannot be considered a completed event until such a debate has taken place and a decision has been agreed.

Failures of direct democracy: general. In modern democracies since the 20th century, representatives typically face intense pressure to appear popular in the public mind.  Popularity, however, is not always responsibility to the nation, and should not be mistaken for it.  In competition with one another and with other parties, representatives can feel tempted to make frequent appearances in the mass media that garner popularity, sometimes even to the point of making promises not supported in legislation or not likely to be supported, raising false expectations in the minds of the people.  Confronted sometimes with widespread opprobrium cast on them as members of an allegedly out-of-touch “elite” or “establishment”, they face pressure not to exercise independent discretionary judgment but to execute popular opinion directly.  They must resist this pressure; they are political professionals, and their capacity for skilled and informed judgement is one key ingredient of the quality, maturity and intelligence of political life in a nation.



Advisory and binding. Referendums must be presumed advisory unless stated as binding and criteria for implementation are specified.

The AV and EU Referendums compared: in Briefing Paper 7212. Section 5 of this document, published by the House of Commons, describes the 2011 Referendum on AV as a binding one and the 2015 Referendum on EU Membership, by contrast, as a merely “pre-legislative or consultative”, i.e. advisory, one.  It states: “This Bill … does not contain any requirement for the UK Government to implement the results of the referendum, nor set a time limit by which a vote to leave the EU should be implemented. Instead, this is a type of referendum known as pre-legislative or consultative, which enables the electorate to voice an opinion which then influences the Government in its policy decisions. The referendums held in Scotland, Wales and Northern Ireland in 1997 and 1998 are examples of this type, where opinion was tested before legislation was introduced. The UK does not have constitutional provisions which would require the results of a referendum to be implemented, unlike, for example, the Republic of Ireland, where the circumstances in which a binding referendum should be held are set out in its constitution.  In contrast, the legislation which provided for the referendum held on AV in May 2011 would have implemented the new system of voting without further legislation […]”[2]

The AV and EU referendums compared: in statute. The above distinction is directly reflected in the text of the two Acts. Section 8 in Part 1 of the Act for AV is headed: “Result of the Referendum – Commencement or Repeal of Amending Provision”.  Paragraph 1 states: “the Minister must make an order bringing into force” the system of AV if a simple majority vote is received in favour of it (commencement).  Paragraph 2 states: the Minister “must make an order” not for AV if the result received is below simple majority (repeal); and paragraph 3 states that if the result is above simple majority, the Minister “must … bring into force [the order] on the same day” of receipt.[3]  By contrast, the 2015 EU Referendum Act contains no section on result; implementation of result; non-implementation of result; or timetable for implementation of result.[4]

Legal status of government word concerning referendums. Government statements about a referendum have no legal status superior to statute passed by Parliament. Statute for a referendum is legally supreme, even when a referendum is merely advisory and ascribes no legislative authority to a result received.  Therefore, governments should refrain from making statements or promises to the public about the legal status of a referendum until there is a statute for the referendum, and then only in accordance with the terms of that statute. If statute for a referendum is, or likely to be, only of the advisory rather than binding type, governments should not speak of it to the public as binding. They should be explicit that it is, or is likely to be, only of an advisory nature. If the referendum is not binding, governments have no legal power to issue promises such as those expressed in the words “We will implement what you decide” (as with UK Government leaflet, delivered to all UK households, 11-13 April 2016, concerning the EU Referendum), or vow orally to implement a result by a certain date.[5]  Such statements are at best misleading, at worst deceitful, and possibly illegal.  Nor should Parliament as a whole collaborate in broadcasting such statements.  Parliament cannot consent to anything more than that which it has passed on behalf of the people, and should not behave as if it has done so.

Franchise. A referendum is potentially a decision for an entire lifetime, not just for 5 years, as with a general election. Therefore, in all referendums, a franchise should extend to all citizens likely to be affected by a result on a long-term basis.  It should include: 16/17-year olds (only 1 or 2 years away from formal maturity); ex-pats, irrespective of number of years already resident abroad, who may wish to return at some point within their lifetime; and other sectors of the population such as long-term non-national residents with rights of residence such as EU-27 citizens, who should have the right and necessary time to take appropriate measures before a result.  In such cases, it would be reasonable to apply criteria such as a minimum of 5 years residence, or marriage to, or parenthood of, a UK national.  In the case of the EU Referendum, EU-27 citizens resident in the UK on a long-term basis before the General Election of 2015 could not have known that the Conservative Party’s promise of a referendum would become government policy and therefore could not have been obliged to adopt UK citizenship after this point merely as a condition for participation.  It is inconsistent that the franchise extended to resident Commonwealth and Irish citizens but not to EU-27 citizens.  Irish citizens are EU-27 citizens; therefore, all other long-term EU-26 citizens should have been granted the same right.[6]

Legal scope of referendums. A result of an advisory referendum cannot abrogate an existing legal state of affairs, unless agreed by sovereign Parliament. A result of a binding referendum, or a result of an advisory referendum accepted by Parliament for implementation, can do so if, and only if, the power authorising implementation is sovereign over all other bodies of law affected by it.  In the case of the result delivered for the Referendum on EU Membership, at least three bodies of law will be affected if implemented, or deemed to merit implementation; and the question remains open at present as to whether UK law-making power in Westminster is indeed sovereign over them:

EU citizenship rights. Through Article 50 of the Treaty of Lisbon, UK law has the power to abrogate the sovereignty it currently shares with the EU.  However, it is not clear that it has the power to abrogate individual citizenship rights enjoyed by UK EU-citizens (resident in the UK or in the EU-27) and by non-UK EU-citizens (resident in the UK) in virtue of the UK’s current membership of the EU.

Devolved administrations. The question also remains open as to whether UK law-making power in Westminster has the power to abrogate the limited sovereignty it has devolved to Scotland and Northern Ireland since 1998.



Thresholds. Binding referendums must specify a threshold and a minimum turnout.

High thresholds. For major constitutional change, most representative democracies require not a simple 1/2 majority but a 2/3 super-majority of votes cast and minimum turnout. Consent of a super-majority of representatives and of devolved administrations is also typically required.  Alternatively, a high percentage threshold is specified for the overall electorate, covering both votes given and not given by eligible voters.  If a decision is of enormous consequence, potentially affecting an entire population for many years to come and removing rights of citizenship previously held, reason should exist for specifying not just a high but a very high one.

Thresholds in the UK and other countries. Government Standard Note PC02809, entitled “Thresholds in Referendums”, published on 29 June 2011, explains the use and non-use of thresholds in previous UK referendums and in other national systems in Europe and further afield.  This document is presumably well-known to MPs and required reading in the drafting and debating of any piece of legislation concerning the holding of a referendum.  The document explains why thresholds can be difficult to set and have been often resisted in previous UK referendums; and it sets out the most basic reason for this in table form on p. 13, namely that, in principle, referendums in the UK are, in contrast to some other countries, “consultative” rather than binding.  Nevertheless, the document opens with the following key statement of rationale: “Discussions of the need for, or advantage of, some form of threshold usually arises in the context of ensuring the legitimacy and acceptance of the outcome of a referendum exercise. This incorporates the idea that major constitutional change is something more important than the result of ordinary elections, and therefore should be the result of something more than a simple plurality of the votes”.[7] Briefing Paper 7212 adds the further sentence: “The UK does not have a comprehensive written constitution and so any requirement for a threshold has to be included in the individual referendum legislation”.[8]

Onus of the vote for change. High thresholds follow from a principle of the onus of a result on the vote for change, not on the vote for status quo.  If a long-term constitutional status quo is to change on a long-term basis, the onus must fall exclusively on those who want change to convince a significantly large number of other people that it is worth having.  No onus to vote falls on those either content with the status quo or lacking in any great objection to it.  Satisfaction or lack of objection can quite properly be presumed from the very existence of the status quo as a part of everyday life.  It is illegitimate of a democratic government to coerce voters content with the status quo or less discontent with it than with the proposed change to vote for it actively, especially if they had no control over the appearance of such a proposal.  The very occurrence of a referendum may not have been willed by a majority of the electorate in a preceding general election, or even by a majority of popular votes cast in that election, but only by voters for the party promising the holding of a referendum in an election manifesto and subsequently elected to government (on a first-past-the-post system).  Moreover, a fundamental asymmetry obtains between a status quo and a proposal for change, which is that the former is already directly known and practically experienced by everyone, while the other is not – by anyone.  The one is a reality, while the other is nothing more than a proposal.  The one has all the authority of everyday relative certainty and predictability behind it, while the other is at best something known theoretically.  Any projected change must, at the very least, be carefully researched, planned and transparently presented to voters; and even then, it cannot, in the nature of things, carry the same weight of certainty as the status quo.  If an entire country is to risk substituting a state of affairs it does not and cannot know directly for a state of affairs it does know directly already, the onus of the risk must fall exclusively – and heavily – on the vote for change.[9]

Turnout. The principle just described also governs the appropriate treatment of below-threshold turnout in a referendum. If diminished turnout occurs, no consent of the people to change can be presumed. If a government cannot know the minds of eligible voters who do not turn out to vote, it is less presumptuous of a government to treat votes not given as consent a state of affairs directly known than it is to treat them as consent to a state of affairs not directly known.  A government cannot ignore non-votes, i.e. votes not given by eligible voters: if it recognises turnout thresholds as meaningful at all, it must, in principle, count non-votes rather than not count them at all; and, by least presumption, it is less illegitimate of it to treat them as indicative of consent to status quo than it is to treat them as indicative of consent to change.



  1. The EU Referendum Act specified neither a majority threshold (not even a simple one) nor a minimum turnout. It specified no criteria and no timetable for implementation of a result.  Therefore it must be assumed to be a merely advisory referendum, its result still pending parliamentary interpretation and judgement as to advisability for national policy and implementation as law – exactly as stated in Commons Briefing Paper 7212, section 5, and reflected in legislation.  However, thus far (date of writing: 4 September), no meeting of Parliament has occurred – nearly two and a half months since the referendum took place.
  2. If this referendum should in some way be taken to be a binding one, it is impossible to see how its result by itself could be a valid mandate for implementation in law. The result is widely thought to be defined by a 51.9% vote for Leave and a 48.1% vote for Remain.  However, these percentages express only the proportions of actual votes cast.  If votes not cast by eligible voters are counted – as they rightly and properly should be in a referendum of this type – the true result of this referendum is the following: 37.4% for Leave, 34.7% for Remain, and 27.9% non-turnout.  Therefore, whether by voting or by not voting, 62.6% of the electorate indicated no wish to Leave the European Union on 23 June 2016.  This does not mean that a majority of the electorate voted to Remain.  But it does mean that only a minority of the electorate (37.4%) voted to Leave.
  3. The figure of 37.4% of the electorate for Leave may be seen as further diminished when percentages for wrongfully non-enfranchised classes of voters are included in the distribution.  The figure falls to approx. 26% when both 16-17 year olds and EU-23 citizens are included (25% + 5% respectively), and still further when a percentage for over-15 year non-resident ex-pats is calculated.  (EU-23 refers to citizens of all 28 member states of the EU except the recently acceded states of Croatia, Romania and Bulgaria, + Ireland + UK.)  Nor could even this resultant overall percentage for Leave be seen as taking due account of the fact that only a minority of the electorate consented to the holding of a referendum in the first place, by voting for the Conservative Party in 2015.
  4. Lack of a turnout threshold, depriving voters of any means of expressing principled abstention. may have unduly shifted the onus that should have fallen on the vote for change back onto the vote for status quo instead. Some may have reluctantly voted to Remain in order to forestall Leave, but others may have felt such reluctance as not to vote at all – enabling Leave to “win” on a mere simple majority of votes cast, the 27.9% non-turnout not being accorded any mitigating weight at all.  Further, as widely noted, no concrete details or plans underpinned the option for Leave on the ballot paper.  All that appeared was a blank negation, almost certainly encouraging wishful thinking, projection or disengagement (a “two fingers to the establishment” attitude).  The wording of the two options on the ballot paper may also have predisposed voters toward Leave, the Remain formulation containing 7 words, including 2 genitives, a largely redundant noun (“member”) and a verb laden with negative connotations (“remain” – as in “remains”: vestiges, ruins or “left-overs” of something).  Leave, on the other hand, involved no more than 4 simple words – a formulation that so easily could have been mirrored on the other side by: “Stay in the European Union”.[10]
  5. One final question remains: why has this apparently merely advisory referendum nevertheless been treated publicly by the Government so far as a binding one, its result ready for implementation in law, without any need for prior authorization by Parliament? And, it must be asked: how could its result – a mere 37.4% minority vote of the electorate – be treated as binding, without fundamental violation of constitutional principle, indeed of the basic concept of democracy, representative or direct?


[1] Cf. James Madison, The Federalist Papers (1787), Article 10: “The Same Subject Continued: The Union as a Safeguard Against Domestic Faction and Insurrection”.


[2] http://researchbriefings.files.parliament.uk/documents/CBP-7212/CBP-7212.pdf

[3] http://www.legislation.gov.uk/ukpga/2011/1/part/1/crossheading/result-of-the-referendum/enacted

[4] http://www.legislation.gov.uk/ukpga/2015/36/contents/enacted

[5] Full text: “The referendum on Thursday, 23 June is your chance to decide if we should remain in or leave the European Union. The government believes it is in the best interests of the UK to remain in the EU.  This is the way to protect jobs, provide security, and strengthen the UK’s economy for every family in this country – a clear path into the future, in contrast to the uncertainty of leaving.  This is your decision. The government will implement what you decide.” https://www.gov.uk/government/publications/why-the-government-believes-that-voting-to-remain-in-the-european-union-is-the-best-decision-for-the-uk/why-the-government-believes-that-voting-to-remain-in-the-european-union-is-the-best-decision-for-the-uk

On 9 June 2015, in the House of Commons, Mr. Hammond also stated: “This is a simple, but vital, piece of legislation. It has one clear purpose: to deliver on our promise to give the British people the final say on our EU membership in an in/out referendum by the end of 2017”. https://hansard.parliament.uk/Commons/2015-06-09/debates/15060939000001/EuropeanUnionReferendumBill.

[6] These issues were raised in the House of Commons on 9 June 2015. Mr Hammond replied:

“The Bill also sets out the entitlement to vote in the referendum. Since this is an issue of national importance, the parliamentary franchise is the right starting point. It means that British citizens in the UK or resident abroad for less than 15 years and resident Commonwealth and Irish citizens can take part. […]

Some will argue that we should extend the franchise further to 16 and 17-year-olds, perhaps, or even to citizens of other EU countries resident here. We do not agree. This is an issue of national importance about Britain’s relationship with the European Union and it is right that the Westminster parliamentary franchise should be the basis for consulting the British people. I concede that there are those in the House who will wish to debate whether that franchise itself should be extended to 16 and 17-year-olds, but the Government are not persuaded and that is a debate for another day. It would be wrong to include 16 and 17-year-olds in this referendum as an addition to the Westminster franchise. / I reject, too, the suggestion that EU citizens living in the UK should be included. The referendum is about delivering a pledge to the British people to consult them about the future of their country. It would be a travesty to seek to include EU nationals whose interests might be very different from those of the British people”. https://hansard.parliament.uk/Commons/2015-06-09/debates/15060939000001/EuropeanUnionReferendumBill

[7] http://researchbriefings.files.parliament.uk/documents/SN02809/SN02809.pdf (p. 13).

[8] http://researchbriefings.files.parliament.uk/documents/CBP-7212/CBP-7212.pdf (p. 26).

[9] Cf. Jon Elster, Ulysses Unbound: Studies in Rationality, Precommitment, and Constraints, Cambridge University Press 2000, pp. 1-174.

[10] This point is made in a trenchant and widely read article of 18 July 2016 by the digital and social marketing consultant Mike Hind. His criticisms are directed to the pro-Remain campaign.  But ultimate responsibility for this lies with Government, in the drafting of the ballot paper.  https://www.linkedin.com/pulse/epicfail-how-britain-stronger-europe-blew-basics-pr-marketing-hind

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s