The Independent Commission on Referendums has published its final report. This sets out almost 70 conclusions and recommendations, all agreed unanimously by the 12 distinguished Commissioners, who span the major divides in recent referendums.
The report is the product of eight months of discussion and deliberation amongst the Commissioners, backed by comprehensive Constitution Unit research into referendums in the UK and other democracies. The Commission has also consulted widely with experts and the public, including seminars in each of the four constituent countries of the UK. We hope that, like the work of the Constitution Unit’s previous commission on referendums, this report will set the agenda for debate about the future use and conduct of referendums.
The use of referendums internationally has increased dramatically over the past three decades. This has been driven partly by changing public expectations of democracy: deference has declined and public desire for input in decision-making has grown. The UK experience has mirrored this trend. Following the first non-local referendum in 1973, there were three further such polls in the 1970s. A further nine non-local referendums have been held since the late 1990s – two of which were UK-wide.
Unlike many countries, the UK has no formal rules regarding when or on what a referendum should happen. As explored in an earlier blogpost, decisions to hold such votes have been driven by a mixture of principle and pragmatism. Nonetheless, conventions have emerged for holding referendums on fundamental questions to do with devolution and the European Union; in some cases, these conventions have even been codified in law. Referendums provide a mechanism for entrenchment in the absence of a codified constitution: decisions explicitly endorsed by the electorate are hard to reverse without further reference to the people.
The role of referendums in democracy
Referendums can enhance democracy: they can answer fundamental questions about who ‘the people’ are, strengthen the legitimacy of major decisions, and allow the public a direct say on major issues.
But referendums can also in some ways inhibit democracy. Voting is central to democracy, but so are processes such as deliberation, compromise and scrutiny. Binary referendum campaigns don’t necessarily create space for these: rather, they can encourage polarisation and division. Badly designed referendum processes can also risk undermining the institutions of representative democracy, which are essential for democratic governance across the board. There are also some topics, such as those affecting minority rights, where using such a majoritarian device may be inappropriate.
Thus, the Commission recommends that referendums be used with caution. Engaging the public in policy-making processes is essential, but there are often better ways of doing so.
The Commission considered whether politicians’ discretion to call referendums should be limited. In the absence of a codified constitution, however, it would be difficult to restrict the topics on which a referendum could be held, and the lack of consensus on the topics on which a referendum should be required would make mandatory referendums difficult too. Given its concerns about referendums in general, the Commission does not recommend extending the power to call a referendum to ordinary voters. International evidence suggests that referendums of this kind are often used by small but well-resourced interest groups, sometimes for illiberal goals.
Instead of formal rules for the use of referendums, the Commission urges politicians, political parties, and campaigners to think carefully before committing to hold a referendum. Above all, that means ensuring that any proposal for change is prepared and scrutinised in detail. Referendums work best when they come at the end of a thorough policy development process, as occurred before the 1997 Scottish devolution referendum and the 1998 Good Friday Agreement referendum. The recent Irish Citizens’ Assembly, which paved the way for that country’s referendum on abortion, offers a promising model for the future. The Commission urges that a range of mechanisms for preparing for referendums should be explored in the UK.
The need for clarity
The proposals put to a referendum should, so far as possible, be clear and immediately actionable. If they are not, voters may be unable to make an informed choice and tensions between popular and parliamentary sovereignty may arise – as they did during the Article 50 notification debates in parliament. The best way to ensure clarity is to hold referendums after the legislation to implement the change has been passed by the relevant parliament or assembly: the Commission recommends this wherever possible.
There may, however, be some circumstances where this is not possible: for example, when a referendum is required to begin a process, such as the negotiation of Scottish independence. In such cases, the government calling the referendum should provide as much clarity as possible for voters, producing a detailed White Paper on what it intends to happen in the event of a vote for change. The legislation enabling the referendum should set out the whole decision-making process. If the plan set out in the White Paper is delivered, the change can go through. If not, a further referendum would be required. It would be for the parliament or assembly that passed the legislation to decide whether that further referendum is needed. Readers should not, however, read into this any recommendation for a second referendum on Brexit: any double-referendum process should be set out before the first referendum; so, if the Commission’s recommendations were followed, we would not be where we now are. The Commission’s recommendations are for future referendums, so it does not take a view on what should happen in the current situation.
Questions and thresholds
Another important aspect of preparation for referendums is deciding the question to be asked. By international standards, the UK has a rigorous process for ensuring that questions are unbiased and are easy to understand, and the Commission commends this. Although not appropriate in all circumstances, multi-option referendum questions could sometimes help to reduce polarisation, and their use in future should be considered. The Commission’s report considers the voting systems that might best be used.
Some countries require special thresholds to be met for a referendum result to be valid. Such a threshold has been used only once in the UK, in the 1979 devolution referendums, when 40% of the eligible electorate had to back the change for it to go through. In Scotland this prevented the referendum from passing, even though a narrow majority (51.6%) voted yes, and similar thresholds have been seen as illegitimate ever since. The Commission also notes that a simple majority is sufficient for almost all other processes of democratic decision-making in the UK and concludes that applying thresholds to referendums only would be inconsistent and potentially harmful to the legitimacy of decision-making.
The regulation of referendum campaigns
The current legislation regulating referendums called by the UK parliament – the Political Parties, Elections and Referendums Act (PPERA) – was enacted in 2000. Since then, we have held five referendums under its terms, and the nature of campaigning has changed considerably. Learning lessons from past UK referendums and taking inspiration from international practice, the Commission has considered all aspects of the conduct of campaigns and makes recommendations on how this framework could be improved.
In the run-up to the EU referendum, the UK government’s decision to spend £9.3 million on a leaflet advocating a vote to remain created significant controversy. PPERA bans governments from publishing anything relating to a referendum within 28 days of the poll. But this allows for unlimited spending in the early weeks of the campaign, potentially tilting the playing field. The Commission recommends that restrictions should be extended to the whole of the campaign period, but narrowed to apply only to campaigning activity so that the business of governing can go on.
The formal designation of a lead campaigner on each side of the debate is unique to the UK. The Commission believes that it works well, but that the process of designation could be improved. First, it should take place earlier: at present, it could take place as late as four weeks before polling day, leaving campaigns with little time to prepare. Further, if there are two applications for one side and only one for the other, as happened prior to the EU referendum, the latter can be confident of designation and therefore gain an advantage in terms of planning. Second, as referendum campaigners receive public money, their leading figures should be subject to a ‘fit and proper person’ test.
The UK has a more comprehensive framework of financial regulation of campaigns than most other democracies, even those that hold more referendums. But improvements could still be made. Data has become a valuable ‘alternative currency’ for campaigners, so the Information Commissioner’s Office and the Electoral Commission should consider how they can best work together to accurately capture the true cost of campaigns. A challenge particular to the financial regulation of referendums is ensuring accountability when campaign groups are temporary. At present, large campaigns have six months to submit their spending returns. Reducing this to three months could allow faster action where necessary.
The quality of campaign discourse was a concern after the EU referendum: the Electoral Commission’s post-referendum survey found that over half of the public disagreed that ‘the conduct of the campaign was fair and balanced’, a third of these citing ‘inaccurate/misleading information’ as the reason. The Commission explored a range of interventions designed to promote good information in other democracies. It concluded that any mechanisms for promoting good discourse should be ‘bottom-up’ rather than ‘top-down. The US state of Oregon, where citizen panels produce considered information prior to ballot initiatives, provides an interesting example. The Commission suggests that similar exercises should be piloted here.
Since 2000 the nature of campaigning has changed dramatically, and has increasingly moved online. As explained in an earlier blogpost, this poses a number of challenges for referendum regulation. Unlike in the broadcast media, there are no restrictions on the use of online paid political advertising. This is consistent with print media but there is a question as to whether such disparities remain justified – the Commission urges an inquiry into political advertising across all forms of media. A further problem is that many of the rules to promote transparency are not fit for purpose in regulating online campaigning. For example, imprint rules requiring campaign materials to declare their source only apply to printed materials. Campaign spending categories make it almost impossible to tell what campaigners are spending on social media. The Commission recommends that these gaps in transparency be closed and that a repository of online political advertising should be created.
Implementing the Commission’s recommendations
Implementing the Commission’s recommendations will require changes of different kinds. First, the norms and expectations surrounding the use of referendums will need to shift. The Commission urges anyone proposing a referendum to think carefully about whether a referendum is the best mechanism for that decision, whether adequate preparation has taken place, and whether the proposals are sufficiently clear. Second, some recommendations will require changes in legislation. This should be done for referendums in general: at the UK level, PPERA should be amended; the Commission urges the devolved administrations to consider introducing generic legislation on the conduct of referendums in their jurisdictions. Finally, the Commission identifies a number of areas that deserve further inquiry, and it hopes these will be taken forward.
This post originally appeared on Constitution Unit blog and is reposted with the editor’s permission
The full report is available here