Referendums in the United Kingdom - Constitution Committee Contents

Memorandum by Dr Matt Qvortrup

  Thank you for asking me for advice on the issue of referendums. It is an honour to be able to help you. I have carefully considered the questions, and I have endeavoured to answer them as thoroughly as the limited space permitted. Please do not hesitate to contact me if you require further elaboration on the questions.

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

  The strengths are that it allows the people to have a final say on fundamental changes affecting the constitution. It has become standard in most democratic countries to require referendums before constitutional changes. The problem though is that the voters are sometimes uninterested in the often technical constitutional issues, eg as has been the case in several recent Irish referendums (the turnout was below 35% in 2001).

What assessment would you make of the UK's experience of referendums? What positive or negative features of this experience would you highlight?

  The positive experience is that it allowed the voters to have a final say. However, as the low turnout rates in Wales 1997 and the North East 2004 shows, the interest in the issue was low—possibly because the voters did not like the options on offer. I would suggest that referendums be made optional, so that the voters can decide when or if to have a referendum. This would ensure that referendums are only held when there is a public demand. This system has worked tolerably well for the provision that there must be a citizen initiated referendum on whether to have directly elected mayors in England.

How does, and how should, the referendum relate to the UK's system of parliamentary democracy?

  I think this question has been answered by history. The debate as to whether referendums are compatible with a Diceyan constitution was raised by Wilson in the House of Commons in 1967—but when it became opportune he changed his mind. Referendums are now—to use a somewhat technical term—"a convention of the constitution". Further the argument that they are not compatible with the doctrine of the Sovereignty of Parliament has been eroded by the practice of referendums in this country—as well as by extensive usage of referendums in New Zealand (a country without a written constitution and explicit adherence to the principle of the sovereignty of Parliament).[17]

Is it possible or desirable to define which issues should be subject to a referendum?

  I would argue that it is up to the people to decide the issues. In some countries, eg Denmark there is a ban against referendums on taxation issues—but there can be referendums on foreign treaties. In Italy the reverse is true. So the evidence is mixed, but—needless to say—referendums should be avoided on issues affecting human rights and named individuals.

Is the Political Parties, Elections and Referendums Act 2000 (PPERA) an effective piece of legislation? How, if at all, could it be improved?

  Of course, one should never be complacent, but in international perspective, PPERA is seen almost as the gold standard in referendum regulation.

Is the role of the Electoral Commission in regard to referendums, as set out in PPERA, appropriate? What assessment would you make of the Electoral Commission's work in relation to referendums?

  This is difficult to say as the only experience we have is the 2004 referendum. I would say that the jury is still out as regards the role of the Electoral Commission in referendums.

What comment would you make on key components of a referendum campaign, such as:

  1.  Whether or not there should be any threshold requirements, for instance in terms of the percentage of the vote required, or the level of turnout required, for a vote to be carried. Thresholds are rare but they do exist. It is common but inaccurate to cite the Canadian Clarity Act as a precedent for supermajority requirements. As the example is often mentioned, it is instructive to give a brief description on this model. After the 50.58% to 49.42% result in Quebec in 1995, the House of Commons in Canada sought to establish that, a future referendum was not won by a narrow margin. The act, however, stops short of recommending a specific majority. See Clarity Act, 2000, c. 26 [Assented to 29 June 2000] 2. (1) The House of Commons shall consider "whether, in the circumstances, there has been a clear expression of a will by a clear majority of the population of that province that the province ceases to be part of Canada". Factors for House of Commons to take into account include (2) (a) the size of the majority of valid votes cast in favour of the secessionist option; (b) the percentage of eligible voters voting in the referendum; and (c) any other matters or circumstances it considers to be relevant." A better example of a Supermajority, albeit a small one, was used in 2006 in Montenegro (a referendum that has become the gold standard of best practice); the law stipulated that independence be approved if supported by 55% eligible to vote. (The total turnout of the referendum was 86.5%. 55.5% voted in favour and 44.5 were against breaking the state union with Serbia. Another example is St Kitts and Nevis in the Caribbean. Under the constitution, Nevis has considerable autonomy and has an island assembly, a premier, and a deputy governor general. Under certain specified conditions, it may secede from the federation. In June 1996, the Nevis Island Administration under the concerned citizens' movement of Premier Vance Amory announced its intention to do so. Secession requires approval by two-thirds of the assembly's five elected members and by two-thirds of voters in a referendum (Art 38.1 (b).[18] After the Nevis Reformation Party blocked the bill of secession, the premier called for elections for 24 February 1997. Although the elections produced no change in the composition of the assembly, Premier Amory pledged to continue his efforts toward Nevis' independence. In August 1998, a referendum on the question of independence for Nevis failed and Nevis presently remains in the Federation. The March 2000 election results placed Vance Armory, as head of the CCM, the leader of the country's opposition party. In most other referendums (e.g. East Timor 1999, Malta 1955, the referendums on independence for former Soviet States 1990) there were no special majority requirements. Special majorities and turnout requirements were used in Scotland in 1979. The outcome was a rejection of the proposal for self-government, although a majority had voted in favour. The result exacerbated antagonisms. The method of voting must be as simple as possible, to allow maximum effective participation. In conclusion, a result endorsed by 50% of those voting plus one should be accepted, as long as a majority of those eligible (and registered) to vote have cast a ballot.

  2.  The wording of the referendum question (including the appropriateness of multi-option questions). The experience tells us that the wording of the question is of little importance. E.g. in 1980—despite the employment of pollsters and rudimentary focus-groups—the secessionists were not able to win despite having chosen a very loaded question. As for multi-option referendums, they have been used in Sweden in 1955 and in 1980, on the first occasion, when there was a clear majority, the result was positive (the same was true in New Zealand in the early 1990s and in Puerto Rico in 1993). But more often than not it results in competing options—with no one receiving support (eg as in the case in Sweden in 1980). So generally, multi-option referendums confuse matters and detract from the major attraction of the referendum, namely that it provides a clear choice.

  3.  Whether there should be formal, constitutional triggers for referendums and whether a referendum should be indicative or binding. As a matter of constitutional law a referendum cannot be binding in this country. If we were to have a written constitution (and I might personally have reservations about this) there would normally be formal constitutional triggers for a referendum. It is possible to have constitutional triggers for a referendum under our system. In New Zealand changes to the electoral laws are entrenched and require a referendum before they are changed. A.V. Dicey the constitutional lawyer suggested that referendums should be held before changes to fundamental parts of the constitution. The problem—as he admitted was—that this Act could be repealed.[19]

  4.  Whether a referendum should ask broad questions of principle, or refer to specific legislation. There are examples of both. In 1998 there was reference to a special agreement (the Good Friday Agreement).

  5.  Whether a referendum should precede or follow statutory enactment. There are—once again—examples of both. In 1997 the legislation on devolution was pre-legislative, whereas in the referendum on the Good Friday Agreement the referendum was held to ratify a decision. Post legislative referendums are the most common (by a factor of 8 to 1). This is possibly because referendums are seen as final decisions. From a theoretical perspective—ie one that stresses that the referendum is a constitutional safe-guard—it is preferable that referendums are held after the third reading of a Bill.

  6.  Campaigning organisations and the funding of campaigns. Public information campaigns and media coverage and Party political activity; This is a vast topic worthy of a doctoral thesis. But the general principle is that referendums are not party political contests but (as was the case in 1975) a public vote that goes beyond party lines. As a result of this it is necessary to secure that both sides get a fair hearing—and it might be necessary to have public funding. I have divided my answer to this in two:

Government Spending

  In referendums around the world, concerns are often raised about the government using public funds to support a favoured position, when the perceived goal is for people to decide without politicians biasing their thought process. In 1994, the Austrian government spent considerable sums on a pro-EU campaign, but without violating Austrian election and referendum laws. The same has been true, more recently, in Spain where the government is reported to have spent considerable amounts of public monies on a (successful) campaign in support of the European Constitution.[20] In other countries—most notably in Ireland—similar examples of government spending in support of a proposition have been ruled illegal by the courts. In an often-cited case from Ireland in 1995, an Irish MEP to the European Parliament argued that the government had breached the Irish Constitution by spending public funds on aspects other than the impartial organization of the process. While the Supreme Court held that: "The Government is clearly entitled to spend money in providing information…[and that] the Government, as such, is entitled to campaign for the change and the individual members of the Government are entitled either in their personal, party or ministerial capacities to advocate the proposed change", it ruled that "the Government must stop short of spending public money in favour of one side which has the consequence of being to the detriment of those opposed to the constitutional amendment".[21]

  Although legally non-binding, this judgment has inspired legislation in both Ireland and elsewhere. There is an emerging consensus that it is illegitimate for governments to spend taxpayers' money on partisan information, or other partisan activities using state apparatus.

Campaign Spending

  The issue of whether there ought to be a ceiling on campaign expenditure is contentious. Some argue that expenditure ceilings keep costs within manageable limits, ensure that referendums cannot be "bought" by the richer side, and increase public confidence in the result. Others contend that ceilings prevent a truly effective information campaign.

  This is not a conclusive debate. Many argue that the outcome of the referendum seems to be driven by other structural factors, such as the economy, the length of tenure of the respective governments and other factors.[22] Some doubt the importance of money in ballot campaigns, though it has been reported that "negative" spending in many cases has been successful.[23] Still, restrictions on expenditures in ballot campaigns are common: In the 1970s, in the run-up to the first Qu

bec referendum on "sovereignty association", the provincial Parliament restricted campaign expenditure, and mandated that two campaigns be established representing each side of the argument.[24] Quebec's Minister of State for Electoral and Parliamentary Reform, in a 1977 paper, noted that the regulations it had passed were inspired by Great Britain's experience with a referendum in 1975, which it held up as an "invaluable guide", reflecting a "deep-rooted sense of fair play."[25]

  In the more recent past the UK Labour government has in turn enacted legislation based on the Qu

bec Act, namely The Political Parties, Elections and Referendum Act 2001 (PPERA). PPERA also introduced limits on campaign spending, and due to its comprehensiveness, this Act is often cited as a key reference point in debates about referendum regulation, internationally.[26] The restrictions on campaign spending are as follows (Sections 117-118):v Similar provisions exist in New Zealand under the Citizen Initiated Referenda Act 1993. Under this act, it is an offence to spend more than $50,000 promoting the petition (at the qualification stage), and to spend more than $50,000 promoting an answer to the referendum. As in the UK and Australia, what an organisation is spending on advertising in relation to the petition or referendum must be reported to the Chief Electoral Officer.

Whether referendums should coincide with other elections or not

  The practice is varied. A case can be made for the view that referendums on the same day as elections muddies the political waters—and make voters focus on something else. This is a valid objection, which can be raised against referendums in American states (where there are often several referendums on the same day as elections—and hence "a drop-off" as voters get tired of voting).

  However, Britain cannot be compared to California or Switzerland. The referendum is unlikely ever to become part of the political woodwork in the way that it is in the aforementioned cases. If referendums are held on the same day as elections in countries where referendums are relatively rare, there is very little that suggests that the voters get tired of voting—or even let their decision be influenced by their decision to vote for a particular party. For example, in 1959 Eamon de Valera was elected president—at the same time he campaigned for the abolition of Proportional Representation (and the introduction of First-Past-the-Post). While he won the presidency handsomely, he lost the latter by a clear margin. The voters, in other words, are capable of distinguishing between measures and men. Sure, the turnout in referendums tend to be marginally lower than in candidate elections (more people vote for candidates than for or against propositions), but turnout in referendums held on the same dates as elections tend to get higher turnout than other referendums. This strengthens the case for referendums held on the same days as elections.

4 January 2010

bec Referendum Act 1978.) The 1998 Amendment states "The total of contributions to each national committee by the same elector in the same referendum shall not exceed the amount of $3,000" (Section 91).

bec Ministry of State for Electoral and Parliamentary Reform Consulting the People of Qu

bec, (1977), p. 7

17   On this issue, see my book M. Qvortrup (2005 second Edition) A Comparative Study of Referendums. Manchester University Press, chapter 2. Back

18   The 1983 Constitution of St Kitts and Nevis states: 38.-(2) b) [a constitutional amendment must be] approved on a referendum by not less than two-thirds of all the votes validity cast on that referendum in the island of Saint Christopher and two-thirds of all the votes validly cast on that referendum in the island of Nevis. Back

19   See M. Qvortrup (1999) A.V. Dicey: "The Referendum as the People's Veto", in History of Political Thought, Vol.20, No.3, pp. 531-546. Back

20   El Pais 5 January 2005, "Periodistas, futbolistas y actors abren el viernes la campa?a del referendum europeoBack

21   McKenna v An Taoiseach, an Tanaiste and ors. 1995 Back

22   M.H. Qvortrup, "How to Lose a Referendum", The Political Quarterly Vol. 72, No. 1, 2001 Back

23   Elizabeth Gerber, an American political scientist, has found that campaign spending in support of a proposition was ineffectual. However, negative campaign spending, ie spending against a proposition was often effective. See E. Gerber, The Populist Paradox. Group Influence and the Promise of Direct Legislation, Princeton, Princeton University Press, 1999. Back

24   In a 1998 amendment, contributions were limited to $3,000 per donor to each campaign. (1978 Qu Back

25   Qu Back

26   See also: The Local Authorities (Conduct of Referendums) (Wales) Regulations 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