Referendums in the United Kingdom - Constitution Committee Contents

Memorandum by Nigel Smith[34]


  UK democracy would benefit from extending its use of referendums.

  Referendums must first be removed from their existing plebiscitary context and placed on a new constitutional footing independent of government.

  Referendums on constitutional matters should become obligatory.

  The Lords should be given a qualified right to call a referendum.

  PPERA is unsatisfactory and needs to be revised or replaced.

  The Electoral Commission should be continued but its role clarified.

  Multi-option referendums on major issues should be avoided.

  Thresholds in referendums should be avoided as far as possible and when used be open and minimal.

  Major referendums should be stand alone political events held separate from General Elections.

  The minimum length of the referendum period should be increased to protect and improve the deliberative process prior to the vote.

  The introduction of Citizens Initiative should be encouraged and anticipated.

What are the strengths & weaknesses of the referendum as a democratic tool?

  1.  Absolutists argue that ceding any decisions to incompetent voters undercuts representative democracy leading to "wrong" decisions, that referendums are conservative devices incapable of dealing with reform and worse they can be used to oppress minorities. Much of their evidence lies in a plebiscitary past.

  2.  Having studied referendums in modern democracies, I believe the criticisms overstated. If absolutists were prepared to cede a little, they would find our democracy would gain a lot just at the time it needs to be revitalised.

  3.  Referendums cost money and take time, are not so good with new, unfamiliar issues or with multiple options for reform, are cautious rather conservative when considering change. Without fair and good procedure especially in initiative referendums, they can soon demoralise voters and distort political priorities.

  4.  On the other hand, they tiebreak major issues that have split politicians and parties, make decisions that voters are committed to, induce consensus and embrace reform. They educate and involve the voter in both the issues and the democratic process. They can introduce new issues or highlight ones that politicians would like to ignore.

  5.  Most of the procedural problems that critics highlight can be avoided by good design that sets out to give a proper role to referendums rather than with a wilful intent to cripple them. The UK is in the position to choose best practice from around the world.

  6.  Some argue that constitutional issues are too complicated for the voter yet they remain a frequent source of referendums usually with positive effect.

What assessment would you make of the UK's experience of referendums?

  7.  The UK has used referendums for a long time—my local authority decided by referendum in 1878, more than 130 years ago, to municipalise the gas supply—but much of this experience of minor referendums is lost. Even bigger events like the 1920 Scottish referendum rejecting prohibition are forgotten and we are left largely with the post war record from 1973 the Northern Ireland Border referendum onwards.

  8.  I would separate the 70s referendums from the 90s. Some features of these deserve recalling. The way the Government arranged the 1975 EU public debate (a plan now declassified) would probably not be acceptable to the Electoral Commission, the use of thresholds to hobble the first devolution referendums, the option of the Orkney & Shetland veto, the very use of a referendum and subsequent abstention of the catholic vote in 1973. It is arguable that the 70s referendums although providing valid results didn't entirely settle matters.

  9.  There have been big changes in society since the 1970s. Voters and the media are less deferential and both are much less inclined to take for granted the advice of politicians. Better educated voters served by plural sources of information, the replacement of class politics by managerial politics and the fragmenting of the big party duopoly created the more fluid political society desirable for referendums.

  10.  The devolution referendum of the 90s apart from making a historic decision cutting through years of political impasse also sustained public support for the Scottish Parliament through the first difficult years to a degree that would not have been possible without a referendum. In Northern Ireland, the result showed a large body of Unionists prepared to share power certainly enough to protect reforming politicians and sustain support for the long end stage. And in Wales, it showed how far public opinion had moved from 1979 to 1997 making clear that the narrow result was part of a movement of opinion.

  11.  Yet even the major referendums of the 90s remained very much a tool of government—closer to a plebiscite than a referendum. Governments used them to take controversial issues out of a General Election or consider putting one on PR into an election in order to appear progressive. Governments decide the issues for a referendum then change their mind. They choose the timing and even the length of the referendum debate then give themselves a privileged position in it.

  12.  In the absence of a culture of direct democracy, the referendum is conducted as much as possible on a party basis suppressing dissenters in their own party where they can and hoping that giving a party lead will be decisive with their voters. Even political commentators became lulled by the Westminster prism into seeing it as a tool of government. Journalists were surprised to learn that in the honeymoon after 1997 the majority of Labour voters in Wales ignored Blair's appeal and stayed at home rather than vote for the Assembly.

  13.  Although the UK experience may be described as mixed with more recent experience the more encouraging, it is clear that referendums are here to stay. The first reform must be to extricate the referendum from this plebiscitary context and to give it a separate role in our democracy, independent of government where all referendums would use fair and good procedure and some become obligatory.

How does the referendum relate to the UK's system of Parliamentary democracy?

  14.  The UK is a representative democracy and should remain so. Putting a few major issues and rather more minor ones to referendum will hardly undermine the principle of representative democracy. There would have to be a wholesale adoption of Citizens' Initiatives in a quite radical way to bring the degree of change some fear.

  15.  The more immediate danger for our democracy is that in 50 years the standing of Parliament has fallen from its post-war high to its present low to a degree that simply can't be explained by a loss of deference or the problems of the past year. Restoring trust needs radical reform on several fronts. Greater use of referendums should be one part of this wider reform but it is not a panacea on its own.

  16.  It is worth noting the loss of trust in politicians was a material factor in the referendum on the NE Assembly and the Edinburgh and Manchester transport referendums.

Should "constitutional issues" be subject to a referendum? If so how should it be defined?

  Rather than raising all the referendum options piecemeal, I list them together here as a scheme.

  17.  Constitutional issues should be subject to obligatory ratification referendums.

  All the recent and most of the proposed major referendums are prima facie constitutional in nature involving either a change in governance, method of election or the ceding of powers not easily retracted.

  The obvious difficulty is the lack of a written constitution. The Scottish Parliament must certify that a bill is within its devolved powers. The Speaker in the Commons could certify that a bill or treaty does not contain constitutional issues as defined in a prior set of tests. The creation of the tests is itself a political act but it would separate the principle from the issue and ought to reduce the area of future controversy.

  18.  Governments and Local authorities retain their existing right to hold optional referendums.

  19.  The Lords should be given the power to call an optional referendum on a contentious bill.

  This would rarely lead to a referendum because its value lies as a check on government leading to some referendum proofing by the government of the day. It would certainly be a very British use of direct democracy but not without variants elsewhere.

  20.  Optional veto referendum on new legislation invoked by valid Citizens' petition.

  This gives rise to perhaps one referendum a year in Switzerland. In more than half the instances, the voters support the Government. I include it for the sake of completeness but a reformed House of Lords might well anticipate at least some of the bills that would otherwise provoke its use.

  21.  Citizens Initiative can propose new law which may or may not lead to a referendum.

  There must be a process to produce both a legally viable initiative and sufficient democratic backing to validate it.

Is PPERA 2000 an effective piece of legislation? How if at all could it be improved

  22.  The short answer is that for referendums—it is not. It is both unfair and unenforceable in places

  23.  It is unfair that all participants except the Government must obey the Act for the whole of the referendum period (a maximum of six months) while the Government need only comply for the last 28 days. According it this privileged position which betrays once again the plebiscitary origins of referendums in the UK and is certainly inappropriate. The Government should be put at least on the same footing as everyone else or better still removed entirely from the campaign.

  24.  The Act was the first attempt to control spending in a UK referendum. Leaving aside the important principle involved, the controls would have proved unworkable and become a serious distraction from the referendum debate given journalists obsession with process.

  25.  The main loophole was the £10,000 limit that could be spent by each of an unlimited number of people provided they acted as individuals and not in concert. This class of "donor", a little analogous to 527 campaign groups in the US, would have been beyond the reach and influence of the designated campaigns. The Electoral Commission would not have been able to pin down any shadowy co-operation between them or the presence of richer donors sprinkling money surreptitiously.

  26.  Although our campaign had every intention of abiding by the financial requirements of the Act (we had to convince the Commission we had the means to do this before being designated) there was a substantial risk that despite our best efforts at compliance we would have failed. As MPs have had difficulties enough with returns from permanently established constituency offices, we should not have been surprised if the temporary nature of cross party campaigns, their instant dissolution on referendum day and the scale of a UK wide referendum meant that our "Responsible person" required by the Act would have been left unable to satisfy the Commission on all counts. The Commission would have been faced with prosecuting a non-existent campaign or invalidating the result.

  27.  The maximum referendum period is six months but the minimum could be as short as eight weeks amounting to a snap referendum and given that some of that time is taken up with designation process and early voting, the referendum debate could be very short. It should be increased to at least 12 weeks minimum for major referendums to protect the deliberative process.

Is the role of the EC in regard to referendums as set out in PPERA appropriate?

  28.  If state aid is not to be given automatically to every campaign group then some selection has to be made and an independent body such as the Commission would be required to do it. I don't think the option of no state aid is practical.

  29.  In the two years that I dealt with the Commission, I came to respect its intellectual approach to the wording of the question, the consultation on combining referendums and general elections and the way it tackled voter information all of which seemed appropriate.

  30.  However I had many doubts about its ability to handle the more practical issues of designating organisations and managing the permitted participants in the actual campaign. As there was to be no pre-qualification process (something that might be possible in obligatory referendums) I thought that the Commission would introduce a protracted bureaucratic diversion at just the moment both sides should be concentrating on the referendum debate. My worst fears were confirmed in the NE referendum, a fraction of the size of a UK referendum. Much of the complication and potential delay arises from the financial controls required by the Act.

Whether or not there should be any threshold requirements?

  31.  Over much of the referendum world such thresholds have been incorporated in the referendum process either in a covert way or at unreasonable levels with the (usually) wilful intent of crippling the referendum process. With the result that in many countries, neither the GLA nor the Welsh Assembly would exist.

  32.  Democracy seems to get along pretty well on a simple majority. So the supermajority of 60% of votes cast required to validate the "Yes" vote in British Columbian PR referendum seems unreasonably high if at least open and honest.

  33.  A common threshold is that 50% of the registered electorate must vote yes for the result to be valid. The arithmetic of this beguilingly democratic requirement is, given a 60% turnout, yes must win 83% of the votes cast. Just this threshold has neutered Italian referendums and the variant used in Scotland in 1979 produced a strong adverse reaction to referendums afterwards.

  34.  Using registered voters instead of votes cast makes non voters into no voters and thus encourages abstention campaigns. A democracy requires precisely the opposite effect.

  35.  Thresholds can also be uneven in effect. Threshold seen as reasonable for a major issue can prove impossible for a more specialist issue like "Dentures for old people."

  36.  If used at all, turnout quorums should be set at levels to avoid ridicule and not as additional democratic hurdles. Above all the thresholds must be obvious to all voters.

The wording of the referendum question including multi-option questions.

  37.  The question should not be leading nor combine decisions in a single question.

  38.  A single referendum is not an ideal device for a major multi-option issue. Yet often more than one reform of the status quo is possible. The central problem is not the ballot paper design but that nobody has yet found a way of conducting a multi-option referendum debate that enlightens the voter and doesn't make broadcasting balance impossible to achieve.

  39.  One reason the Scottish Referendum gave such an unequivocal result was the independence option was not only not on the ballot paper (obviously) but also largely removed from the debate by the consent of the SNP. Then as now, pitching Sovereign independence against Devolution is more likely to confuse than enlighten. A referendum on independence should be just that.

  40.  The Scottish Devolution referendum was a simple multi-option referendum, a feature largely ignored by the voters. But if held under PPERA, the Commission would have had to fund all possible outcomes on the ballot paper, including—a Parliament without a tax power—complicating the debate as every broadcaster would have had to follow suit.

  41.  In some multi-option referendums, authorities have tried to manipulate the result by deliberately excluding popular options leading to write-in campaigns.

  42.  Sweden held a multi-option referendum on nuclear power generation involving one principle and variations of degree. While the narrower focus helped, the interpretation of the result remains controversial.

  43.  The most satisfactory way of holding a major multi-option referendum is to hold a pair of referendums as New Zealand did on PR. In the first referendum the alternative PR systems were ranked by voters. The winning PR system then went head to head with the status quo (FPTP) 14 months later.

Should there be formal constitutional triggers for a referendum?

  44.  There is already such a trigger in the Belfast agreement which requires a referendum should NI wish to join the Irish Republic.

  45.  Less certainly, powers devolved to the Scottish Parliament were powers retained by Westminster. But it is hard to see that what came by referendum won't have to go by referendum.

Whether a referendum should be indicative or binding?

  46.  A binding referendum is much to be preferred. Voters know they are making a decision they will have to live with and the decision attracts the media earlier and more intensely thus invigorating the referendum debate, educating the voters and increasing turnout.

Whether a referendum should ask broad questions or refer to specific legislation?

  47.  A broad principle is generally not sufficient. The voters are perfectly capable of extracting a broad principle of what is at stake in specific legislation.

  48.  For example, I supported the broad principle of a regional assembly for the NE until I saw the legislative detail and then predicted its rejection. Sadly the result has been widely interpreted as the wholesale rejection of a broad principle rather than the sensible rejection of a weak proposal.

  49.  A broad principle may be sufficient dealing with "liberty" like the Independence of East Timor but as Scottish independence would be more about good governance than liberty the referendum ought to refer to a specific proposal post negotiation with the UK.

  50.  Mandate referendums may only have a broad question to deal with. De Klerk in SA sought a mandate to continue negotiating. The Conservatives have talked about getting a mandate to negotiate with the EU. The SNP could seek a mandate to negotiate independence by referendum rather than General Election raising the possibility of two referendums on independence.

  51.  Defensive referendums again using a broad principle have been suggested recently to counter Independence. It seems unwise as a general practice to call a referendum as soon as a "threat" appears on the horizon. The risk is that there is differential voting, one side seeing an opportunity, the other not sufficiently aroused by a distant and apparently premature issue. The outcome becomes a mandate to negotiate.

  52.  The only way round it is to hold the referendum at the same time as a General Election banking on the trend that the status quo vote increases as it often does. But this would breach UK good practice and lay the referendum open to charges of rigging. When the Swiss voted to join the UN it was a completely stand alone referendum with a turnout of 81% vindicating the policy of selective participation.

Whether a referendum should precede or follow statutory enactment?

  53.  There should be a firm proposal. Obviously an un-ratified Act is one way of achieving this but not the only way. A treaty, the existence of the euro and the Scottish Convention in their various ways crystallised the issue.

  54.  The Scotland Act might well have been improved if the referendum had followed enactment because the Government used the referendum result to ram it through Parliament paying little heed to the suggestions of friend or foe. So if the issue has emerged from Parliament, I would prefer ratification referendums.

Campaigning organisations and the funding of campaigns

  55.  The Electoral Commission may designate one organisation on either side as lead campaigns.

  56.  Because of the way the UK currently holds referendums, these campaigns are usually hastily formed working against the clock drawing activists, endorsements and donations wherever they can. Beside campaigning skills, they must now build compliance structures. Because of the indecision over the euro referendum, the euro campaign had the luxury of time to prepare. Nevertheless from the opening of the referendum period there would have been a great scramble to be selected as the designated organisation with a still half formed organisation. It is a selection process that could make enemies of allies usually on one side only and thus a considerable responsibility for the Electoral Commission.

  57.  Apart from immediately being elevated to lead campaign in the eyes of the media and public, the state aid for each designated organisation is a free mailing of a referendum address, UK funding of £600,000, free use of premises and referendum broadcasts.

  58.  Though PPERA didn't exist in 1997 it is worth considering how it would have worked in the Scottish Devolution referendum.

  59.  The UK funding pro-rated for Scotland would be £55,000 that is about 9% of what the Scottish Yes campaign actually spent and would have paid for the polling research. Free premises represent a further modest financial contribution.

  60.  There were no referendum broadcasts because of court rulings in prior referendums. So this provision is an important new contribution to the referendum debate though the format could be improved.

  61.  The most valuable aid is the free mailing to the electorate. This was far beyond the financial resources of either side in Scotland. At first the new Blair government tried hard to keep out of the referendum process anxious to be seen as "whiter than white" after its criticism of the Major government. Eventually convinced that the electorate had to be given some information but to avoid being seen as partisan, a rather milk-and-water leaflet was distributed by the Scottish Office.

  62.  Drawing on experience elsewhere, the format of the referendum address could be developed into a statement and rebuttal from both sides contributing more to the referendum debate.

  63.  There is never enough money in campaigns. The state aid in the Act might be raised for obligatory referendums but otherwise it is probably about right. My concerns are more about the regulatory process and compliance getting in the way of the deliberative debate.

Public information and media coverage

  64.  Though hardly ever an Athenian ideal, Switzerland gets close, the referendum debate is one of the defining features of a referendum. It is greatly helped by the UK tradition of 50:50 broadcasting balance in referendum coverage, something the UK shares with few other countries, notably Switzerland and Ireland.

  65.  But if this referendum debate is itself preceded by some kind of event which both crystallises the issue and educates many especially the media, politicians and opinion formers then the subsequent referendum debate will be enhanced. In Scotland obviously the Constitutional Convention, in Northern Ireland the long negotiation of the Belfast Agreement, on the euro its very existence alongside the "five tests", even on a small scale the Harris super quarry and its public inquiry, all these served to prepare the arguments, brief the media and move opinion.

  66.  The council tax referendums in the South of England and the movement opinion in the Harris super quarry referendums showed that public information works. But greater thought could be given to how this can be improved especially in those parts of the country not well served by regional newspapers and the broadcasters.

Party political activity

  67.  Politicians should take part in referendums as individuals but the role of the Party should be restricted to endorsing an issue.

Whether they should coincide with other elections or not?

  68.  Major referendums should not coincide with a General Election, minor referendums may or may not. No major UK referendum has yet coincided with a General Election. While this tradition has arisen more from low politics than high principle, the Electoral Commission have since given their independent view that this is a sound practice worth continuing.

  69.  Towards the end of 2002, there was talk of the euro referendum being held at the same time as the second Scottish General election. As a devolutionist, I had seen the way the Kosovo crisis had overshadowed the first SGE and knew the euro referendum would dominate the second. So I put the international experience to the Electoral Commission who in turn launched a consultation. The Commission subsequently recommended against holding a major referendum at the same time as a General Election.

  70.  The arguments briefly are General Elections are party events, referendums are not. Broadcasters find it difficult enough to maintain the 50:50 referendum balance without the simultaneous complication of reflecting party strength in a general election. So the referendum debate is damaged even obliterated.

How does the referendum relate to other tools such as citizens initiatives?

  71.  All referendums discussed so far represent responses to the legislative agenda. Citizens' Initiatives would for the first time in the UK allow voters to impose their agenda directly on the Legislatures raising issues ignored by politicians.

  72.  The US Federal Government for long opposed clean energy policy. At the same time citizen initiatives encouraged several States to introduce laws promoting the use of clean energy so change came from the bottom up. The recent initiative in Switzerland banning minarets articulated a problem being swept under the carpet. Not all successful initiatives end in a referendum because governments may simply adopt the initiative into their own agenda or make a counter proposal.

  73.  There is also a provision for an EU Citizens Initiative in the Lisbon treaty. However it is closer to transnational petition for a policy. The more likely development seems to be the Conservatives who will nominate policy areas within which initiatives can be used to set the agenda (on whom is unclear). If successful, one can see them self seeding into other policy areas but it is just this kind of unplanned growth which gives rise to later procedural difficulties.

  74.  The Committee is right to anticipate this and to set the innovation and procedure in proper constitutional context.

How would you assess the experience of other countries in relation to the use of referendums?

  75.  Many countries use referendums, the majority on an occasional basis. There are quite a number of countries with a referendum process in their constitution, rather fewer with an initiative process but still plenty of experience good and bad.

  76.  The ones the UK should look at are—Switzerland is in a class of its own but has many good practices especially the deliberative culture. California will this year hold a constitutional convention to reform it's over active initiative process. Despite the current difficulties, it is very striking how attached Californians are to Initiatives. US states have certainly something to teach us. Germany has greatly expanded its use of citizen initiatives in the last 10 years and Ireland has used referendums from the foundation of the State.

9 January 2010

34   I chaired the cross party campaign for a Yes vote in the Scottish Devolution referendum in 1997, advised the Yes campaign in the Northern Ireland referendum in 1998 and chaired the UK Euro No campaign from 2002 to 2004. In the latter role, I worked with the Electoral Commission for two years interpreting PPERA in preparation for the Euro referendum and then beyond its abandonment in June 2003 to the conduct of the NE Regional Assembly referendum in November 2004. Back

previous page contents next page

House of Lords home page Parliament home page House of Commons home page search page enquiries index

© Parliamentary copyright 2010