Written evidence from The Electoral Reform Society[1] (EUR 93)

 

Public Administration and Constitutional Affairs Committee

Lessons Learned from the EU Referendum inquiry

 

 

Introduction

 

 

 

 

 

The role and purpose of referendums

 

What is the relationship between direct democracy and the UK's tradition of representative democracy?

 

  1. The relationship between direct democracy and the UK’s parliamentary traditions is still evolving.

 

  1. Examples from elsewhere, however, can show us the potential pitfalls to avoid. Used too frequently and referendums may begin to undermine parliamentary democracy. Switzerland has the lowest national turnouts in Western Europe. No parliamentary election there has seen a turnout above 50% since 1975. Turnout in the last election, in 2015, was 48.4%. Most of the time, Switzerland’s four largest parties are represented in government, as any sizeable opposition force could deluge the government with referendums. A political system like Switzerland’s may well reduce engagement, electoral fatigue can set in and representative democracy may seem less important.

 

  1. However, it cannot be said that referendums are without purpose. In the right conditions they may even be helpful to representative democracy. In the UK, referendums have generally been used to debate issues of major constitutional importance that cross-cut parties – e.g. devolution, Europe and the electoral system. In this way they may act as a forum for teasing out issues which are too big or divisive for the traditional mechanisms of parliamentary democracy, and indeed, act to enhance representative democracy.

 

  1. It is also worth noting that, unlike other Western democracies, referendums are still very much structured by representative democratic institutions. While the Electoral Commission and parliamentary acts such as PPERA have a role, these institutions and legislation can be changed and amended by simple majority parliamentary vote. Only parliament can start referendums, and is ultimately responsible for their rules as well.

 

  1. Yet even when a referendum is advisory, as this year’s on the EU was (and even a binding referendum could simply be overturned by parliament), it is clear that referendums are viewed as morally binding on the political system. Citizens’ expectations are that referendum results will be acknowledged and fulfilled.

 

  1. While referendums are becoming increasingly popular, so are new forms of participative democracy, such as citizens’ juries and assemblies, which allow smaller groups of citizens the time and space to conduct informed deliberation on certain topics.

 

  1. If referendums include or are preceded by forums for deliberation where people can access balanced information alongside diverse perspectives on the topic in question, they can play a capacity-building role, especially for those not usually involved in politics. This could help to enhance the democratic process.

 

  1. In a typically British way, use of direct democracy is increasingly being used alongside traditional representative democracy, just as other, newer forms of participative or deliberative democracy are also taking hold. It is important to capture and learn how these experiences are working as a whole, to help nurture a healthy democracy fit for the 21st century. 

 

What is the legal status of referendums and what questions are appropriate to be determined by referendums?

 

  1. The traditional core principle of Britain’s uncodified constitution is often said to be parliamentary sovereignty. Referendums then, might seem counter to that principle. But, in reality, they are not, as it is parliament that ultimately retains the capability to call and regulate referendums and which has the capability to overrule them unilaterally.
  2. There are problems around referendums being events decided entirely by governments, with governments and parliaments involved in deciding their framing. As in the UK, this can mean referendums used as tools for managing internal or intra-party (the referendums of 1975, 1979 and 2016, in particular, can be pointed to as being to manage internal disputes) disputes rather than as a means of gaining a real popular mandate for an issue of major constitutional importance.

 

  1. This is in sharp contradiction to other states with codified constitutions where other actors can call referendums (for example, by petition), where referendums can sometimes not be reversed except by constitutional amendment, and where the constitution may lay out the role and regulation of referendums.

 

  1. Of course, the Political Parties and Referendums Act 2000 does provide a regulatory framework for referendums, but this could be repealed or amended by the decision of parliament at any moment.

 

  1. Appropriate referendum topics would seem to be ones where great national questions can be answered that struggle to be answered by the representative democratic system, hence why referendums are often on issues that crosscut parties.

 

  1. It is also worth noting that different referendums have also had their own legal bases, with different rules, even different franchises, for each. The failure to give 16 and 17 year olds the vote in this referendum, as in Scotland, was an immense missed opportunity to invigorate the debate, for instance. Evidence from researchers such as Jan Eichhorn of Edinburgh University shows that 16 and 17 year olds engaged with the Independence referendum in a sophisticated way. They accessed information from more sources than voters from other age groups. Turnout was higher among 16- and 17-year-olds than among either the 18- to 24-year-old and 25- to 24-year-old age groups. This allowed for a greater engagement of young people in the referendum process. By learning from best practice we can make sure that future referendums more closely match the ideal of a national deliberative event.

 

  1. Currently there are no clear criteria for what triggers a referendum, and a more coherent sense of what is and isn’t appropriate for referendums would be advantageous.
     

The regulatory system for referendums

 

How effective, and comprehensible, is the existing system of regulation for referendums in the United Kingdom?

 

  1. The regulatory system for referendums falls down in several regards.

 

  1. Firstly, a huge issue in the referendum was the question of misleading statements. Statements from both campaigns were criticised as misleading during the course of the campaign.

 

  1. In New Zealand both the Electoral Commission and the Advertising Standards Authority have powers to critique political campaigns, while in Ireland a Referendum Commission is created for each referendum – and part of its mandate is to make for a more informed debate.

 

  1. It is worth recognising that the nature of political campaigns means that very few statements can be determined as out-and-out lies. Usually there are questions of interpretation or speculation that can raise enough doubt to make a clear ruling difficult. This makes such rules politically sensitive and difficult, as well as a potential freedom of speech issue. But given the scale of concern about misleading claims, particularly in referendum campaigns, it may be appropriate to invest an official body, (existing or specially convened) with powers of intervention. Appropriate sanctions might include immediate withdrawal of the offending material. There may also be a deterrent effect deriving from the public reaction to any campaign reprimanded by an official body tasked with this role.

 

  1. The campaign period was noticeably shorter than previous referendums. While it may have seemed like the EU debate had been going on forever for political anoraks, most voters do not engage with the detail of politics and needed time to engage. The four-month campaign period, with local and devolved elections in the middle of them, did not provide that time.

 

  1. Polling commissioned by the ERS showed that 46% felt that they were poorly informed or very poorly informed about the referendum on the 24th of February. This fell to 38% by 26th of April, directly in the run-up to the local elections, and fell again to 34% by the 25th of May, a couple of weeks after the locals – and then again to 28% by the 17th of June, the final week of the referendum campaign.  But given more time, it’s possible this figure could have been much lower. In our final poll, 33% described themselves as informed, whereas 56% described themselves as informed in Scotland in December 2013, ten months before the referendum.

 

  1. There should be mandatory pre-legislative scrutiny for any parliamentary Bill introducing a referendum, lasting at least three months. This should include real citizen involvement, through a randomly selected Citizens’ Select Committee and/or a wider consultation process. This would give citizens and all parties an opportunity to question the desirability of a referendum, and to shape the process and rules in order to maximise the chances for an informed and engaged campaign.

 

  1. All referendums should have a minimum six-month regulated campaign period. This will allow the public enough time to get to grips with the issues and make a low-information, low-deliberation referendum campaign less likely. This would also give campaigns more time to make their points.

 

  1. As soon as possible after a referendum Bill has been passed, the Electoral Commission should publish an official ‘rulebook’ setting out the timetable, rules for campaigners and all other technical aspects of the vote. This should be the ‘bible’ for the referendum, to minimise controversy around the administration of the campaign.

 

  1. There should be an official, publicly funded resource for stimulating deliberative discussion and debate about the referendum. Initiatives which equip people with the information and platforms needed to deliberate the issues around the referendum should receive official support.

 

  1. The Electoral Commission or a specially appointed Referendum Commission should provide a toolkit for members of the public to host their own deliberative discussions about the referendum. Our ‘Better Referendum’ intervention demonstrated widespread appetite for members of the public to get together in a high-information environment to discuss the issues. Similar deliberative tools should be rolled out as part of any public engagement initiative.

 

Should referendums on constitutional questions include a minimum percentage threshold?

 

  1. It is fully understandable why this demand is made. The sense that a referendum might be won on a low turnout, by a tight margin, causes many to question the possible legitimacy of results. Historically many have warned of the potential ‘tyranny of the majority’ that democracy could produce, hence such complaints attempt to protect the rights of minorities. There are broadly three types of win threshold, ones that attempt to guarantee a strong turnout, ones which seek to guarantee a geographic representation and (less common) ones which require a supermajority.

 

  1. The problem with minimum turnout threshold is that they can turn abstention into a political tactic. Referendum turnout thresholds of 50% are fairly common stipulations in Eastern Europe and in other countries such as Italy. In 2012 Romania held a referendum to impeach the controversial President, Traian Basescu, 89% of voters supported his impeachment, but with a turnout of 46%, it failed to reach the turnout threshold. Had turnout been 50%, and every new vote been for no, Basescu would have been impeached. Basescu specifically advised his supporters to boycott the referendum. Closer to home the turnout threshold in the 1979 Scottish referendum: that 40% of registered supporters had to vote for the proposal as well as a simple majority of voters only served to deepen discontent with Westminster when the latter was achieved but the former was not. Given that 37% of registered voters voted Leave, had such a threshold been applied to the EU referendum, Leave would have failed, but in the context of a result in which a million more people voted Leave than Remain and the highest turnout for a UK-wide ballot since 1992 this would have its own legitimacy issues.

 

  1. Such results create questions around legitimacy and turn abstention into a political tactic, making it unclear whether voters abstained due to disagreement with the proposal or because of genuine disinterest. There are key differences between abstention and opposition and we should avoid conflating them.

 

  1. In Switzerland and Australia, for instance, a ‘double majority’ is required in a referendum amending the constitution, in which a referendum proposal must receive majority support from a majority of Swiss cantons or Australian states as well as from an absolute majority of voters. Before the EU referendum there were calls from Scotland for the country’s voters to have a veto. The double majority model may be superior, but given that there are four UK countries, a majority would mean only one nation voting against. Additionally, there are difficulties with the territorial imbalances in the nature of the UK – with England representing 84% of the population. Additionally, such a geographic stipulation could lead to too much attention being focused on ‘swing nations’ in referendums rather than on every vote counting.

 

  1. Rarer are referendums which require a supermajority of voters. The 2006 Montenegrin Independence referendum required 55% of voters to vote for independence. It passed with 55.5%. Similarly, the 2005 British Columbian referendum on changing the electoral system required a double majority – a majority of parliamentary constituencies, which it easily achieved, and a supermajority of 60% of voters voting in favour. 57.7% of voters voted in favour. These two examples illustrate the legitimacy issues with such supermajorities. In British Columbia the referendum was eventually re-run in 2009, with a lower result in favour. In Montenegro it is unclear what would have happened had the result been between 50 and 55%. This may have potentially led to a crisis.

 

  1. In an advisory referendum, such as the 2016 EU referendum, it is arguably unnecessary to stipulate such a minimum threshold as, ultimately, it is up to government and parliament to interpret the result. If the result had been much narrower on a far lower turnout, the response would perhaps be more open to interpretation.

 

  1. All these are reasons to have a sufficient campaign period and higher quality of engagement to encourage high turnout across society.

 

The campaign

 

How sufficient was the purdah period provided during the EU Referendum campaign and should a longer purdah period apply in future referendums?

 

  1. The ERS is a strong believer in a public debate with the most level possible playing field. Purdah’s existence serves to restrict government from using its resources to advantage one side in an election or referendum – an important principle of democratic governance.

 

  1. The four weeks of Purdah, however, did not stop the government leaflet in April which generated some controversy. While the ERS’s polling suggests that the leaflet may not have been all that effective, this controversy is understandable as the government spent £9m of taxpayers’ money outside the government spending limits for the campaign. It may be worth considering a set of guidelines other than Purdah for the regulated campaign. Purdah itself is fairly strong, and, rightly, prevents large amounts of day to day government action, such as signing contracts. A lighter set of guidelines around government advertising may prevent similar actions in the future, or at least provide a clearer set of rules for a similar exercise. It may also be worth considering lengthening Purdah when compared to a general election campaign given the different dynamics of a referendum.

 

How effective was the designation process for lead campaigners, and will future referendums require further reform of the designation process?

 

  1. The designation process currently works against a plurality of debate on either side of a referendum campaign. Referendums are sometimes criticised for forcing debate on an issue into crude binaries, but if this is the case it is only exacerbated by a system which forces campaigners into that binary even more so. There are left-wing and right-wing reasons to be in favour of and against the EU, and cosmopolitan and nationalist reasons for both too. But the lead designation system encourages a monolithic approach to campaigning. This also has the unfortunate side-effect of making a referendum look more like a general election campaign, with policy pledges on each side which neither side may necessarily be in a position to implement.

 

  1. Alternative models may be to allow a system which recognises multiple lead campaigns, or a system which treats all campaigns equally. However, it is worth noting that there are potential difficulties with systems such as these in which false campaigns could be registered with the intention of harming a particular side, and it could be harmful to one side if one side has a plurality of views and the other has a single united campaign. Care needs to be taken in this area, but it may be worth reviewing the lead campaign designation process to see whether the system can be modified to allow more of a spectrum of views than those that British referendums have so far been known for.

 

What role did the machinery of Government play during the referendum campaign and what were the consequences for Civil Service impartiality?

 

  1. The government’s mail-out to households in April was a subject of some controversy, with accusations that the government was abusing its position to get the result it wanted.

 

  1. Our April poll, the first after the UK government’s national pro-EU mailout which went to 27 million households, the number saying that they received a leaflet rose from just 25% in March to 63% in the weeks following the mailout.

 

  1. However, the percentage of people who said that the Government were the most important source of information when making up their mind on the EU referendum rose by just two percentage points, from 8% to 10%, between the end of March and the end of April after the leaflet was sent out. Our polling also found that the percentage of people who said they felt well or very well informed about the referendum actually fell from 23% at the end of March to 21% at the end of April, two weeks after the mailout sent on the 11th April. This suggests that static information sources – and in this case a one-sided publication – had little effect on people’s levels of informedness.

 

Electoral administration

 

What impact did individual electoral registration have on the referendum?

 

  1. The online registration aspect of the IER system allowed for mass publicity around the need to register, with much discussion of the large number of late registrations, which eventually crashed the website shortly before deadline. At its peak, the online registration system received 515,256 applications to register to vote in one day – June 7th. However, it is not necessarily the case that those applications were new registrations, with some EROs reporting that the majority of applications were duplicates. This illustrates a difficulty with the current system for registration – it is not clear to voters if they are registered, nor do easy mechanisms exist to rectify being left off the register, such as the same-day registration used in many US states and Canada.

 

  1. While it is hard to know how many voters were turned away from the polls due to either not being registered or having an inaccurate registration it is worth noting the Electoral Commission’s report from earlier this year on the state of the electoral registers which found that the parliamentary register (used in the 2016 referendum) was 84% complete (ie 16% of the eligible population is unregistered) and 91% accurate. Registration differs demographically. Only 65% of 18-19 year olds were registered, for instance. Non-white voters are also less likely to be registered as are those in lower social grades. Those who move frequently are also a big gap in the register. Those who had moved into their home less than a year before the survey were only 27% registered. 


 

  1. IER is still bedding in, and registration tends to go up around elections, but IER provides a strong basis for future innovations with the electoral register such as same-day registration, or ‘motor voting’. Experience elsewhere has shown that these innovations would be likely to raise registration levels and perhaps even turnout, with an evidence base from the US which suggests that turnout rises between 3-8% in States with same-day registration.

 

What mechanisms are in place to ensure the integrity and robustness of the electoral register?

 

  1. The nature of fraud means that it is difficult to measure, however, the evidence suggests that it is not especially common. The Electoral Commission records only 38 allegations of fraud relating to the register during 2015.

 

Planning

 

What level of contingency planning should be provided by the Civil Service as to possible referendum outcomes? For example, should it adopt a model similar to that found in General Elections (whereby the Civil Service holds meetings with the Opposition before the result is known to prepare to implement its programme if necessary)?

 

  1. Holding such meetings with the referendum campaigns would seem likely to improve public faith in the referendum process and aid smooth transition afterwards. Whereas a single core plan for the eventuality of a Leave vote is perhaps questionable, substantial work could have been carried out on the various options that could have been pursued in the event of leaving. In addition, with the particular nature of referendums like the EU one, it may be worth considering even wider consultations with academia, civil society and senior politicians.


Appendix

 

Nine recommendations the Electoral Reform Society would like to see considered.

 

  1. There should be mandatory pre-legislative scrutiny for any parliamentary Bill introducing a referendum, lasting at least three months. This should include real citizen involvement, through a randomly selected Citizens’ Select Committee and/or a wider consultation process.


This would give citizens and all parties an opportunity to question the desirability of a referendum, and to shape the process and rules in order to maximise the chances for an informed and engaged campaign.

 

  1. All referendums should have a minimum six-month regulated campaign period.
     

This will allow the public enough time to get to grips with the issues and make a low-information, low-deliberation referendum campaign less likely.

 

  1. As soon as possible after a referendum Bill has been passed, the Electoral Commission should publish an official ‘rulebook’ setting out the timetable, rules for campaigners and all other technical aspects of the vote.
     

This should be the ‘bible’ for the referendum, to minimise controversy around the administration of the campaign.


 


This would lay the groundwork for a more informed and engaged electorate better equipped to deliberate on the issues around a referendum.


 

    1. At the start of the regulated period the Electoral Commission, or a specially commissioned independent body, should publish a website with a ‘minimum data set’ containing the basic data relevant to the vote in one convenient place[2].

      A major source of complaint about the conduct of the referendum was the supposed lack of independent information available about the vote. While there are real difficulties in separating out fact from political argument in these cases, a minimum data set ought to be possible. 


 

    1. An official body – either the Electoral Commission or a specially appointed Referendum Commission – should be empowered to intervene when overtly misleading information is disseminated by the official campaigns.


Misleading claims by the official campaigns in the EU referendum were widely seen as disrupting people’s ability to make informed and deliberate choices. Other countries including New Zealand have successfully regulated campaign claims – the UK should follow suit.

 

 

  1. There should be an official, publicly funded resource for stimulating deliberative discussion and debate about the referendum.


Initiatives which equip people with the information and platforms needed to deliberate the issues around the referendum should receive official support.

 

  1. The Electoral Commission or a specially appointed Referendum Commission should provide a toolkit for members of the public to host their own deliberative discussions about the referendum.


Our ‘Better Referendum’ intervention demonstrated widespread appetite for members of the public to get together in a high-information environment to discuss the issues. Similar deliberative tools should be rolled out as part of any public engagement initiative.

 

  1. Public broadcasters should consider more deliberative – rather than combative – formats for referendum-related programming. And Ofcom should conduct a review into an appropriate role for broadcasters to play in referendums.


While there is clearly a place in our politics for TV debates and traditional political journalism, the binary nature of referendums demands more space to be made for more reflective and discursive formats.

 

 

September 2016

 

 

 

[1] The Electoral Reform Society is the UK’s leading voice for democratic reform. Our vision is of a democracy fit for the 21st century, where every voice is heard, every vote is valued and every citizen is empowered to take part.

 

[2] See Albert Weale, Is there a future for referendums? Constitution Unit blog, 25th July 2016. Available at: https://constitution-unit.com/2016/07/25/is-there-a-future-for-referendums/