1.  This memorandum is limited to one issue only. It arises from the first of the four questions posed by the Constitution Committee in its Call for Evidence. It is the extent to which the adoption of the proposed Constitution of the European Union illustrates the need for new rules to govern whether particular constitutional questions should be subject to an advisory referendum. I appreciate that that question may not be uppermost in the Committee's mind in this present inquiry, and in addressing it I may resemble students who answer exam questions that they wish to see rather than those which are actually set. But I feel strongly that the adoption of the proposed European Constitution provides a compelling example of a fundamental defect in the United Kingdom constitution.

2.  The Constitution Committee has already said that the circumstances in which referendums should be held and what conditions should attach to them deserve a full inquiry.[114] I hope that one result of the Committee's present investigation will be that it will conduct such a full inquiry as soon as circumstances permit.

3.  I expressed my concern to the Committee in an earlier inquiry about the absence of rules in the United Kingdom which might govern the process of constitutional change.[115] In essence, there is no formal, legal mechanism in the United Kingdom constitution which prescribes how changes may be made to the constitution. No special majorities, or compulsory referendums, or other formal procedures are required, merely legislation passed in the ordinary way. So if a Government can get its legislation through Parliament, that is all that is needed in order to make constitutional changes, regardless of how important they may be. It is entirely for the Government of the day to decide whether to obtain any independent evaluation of possible constitutional reforms, whether and how to consult about them, and whether to hold a referendum. This freedom of manoeuvre has produced odd results. Under the present Government, House of Lords reform was preceded by a Royal Commission; possible change to the Westminster voting system was considered by an independent commission; and the public was comprehensively consulted on both issues. By contrast, there was no pre-legislative inquiry nor, therefore, any systematic consultation before the Human Rights Bill was introduced. Moreover, referendums were held on possible devolution to Scotland, Wales, Northern Ireland, London, and on whether local electors wanted an elected mayor, but not (for instance) on Lords reform, or on the incorporation of the European Convention on Human Rights. Is that complete freedom of action defensible in constitutional terms?

4.  My view remains that all major constitutional changes - what the Committee would call "principal" changes - should be preceded by the maximum consultation, by independent analysis, and where possible by cross-party agreement about them. And before being implemented the most important constitutional changes should be put to a national advisory referendum. (Indeed, approval at a referendum would be the next best thing to cross-party agreement if a Government were unable to achieve political consensus.)

5.  The Political Parties, Elections and Referendums Act 2000, Part VII, sets out a detailed legal framework to ensure that referendums above local level are conducted fairly. But it is of course notorious that the Act is silent on the circumstances which would trigger a referendum. Any national referendum must be authorised by fresh primary legislation passed for the purpose. The only time that a Government has thus far caused such legislation to be passed was for the 1975 referendum on whether the United Kingdom should remain in the European Community. Such legislation is also required for sub-national referendums, of which there have been several (including the devolution polls in 1979 and 1998), and of which there will be more (on possible English Regional Assemblies).

6.  Historically, resistance to national referendums in this country was based in part on the abuse of such polls in the 1930s by the dictators, and in part by parliamentarians' desire to maintain parliamentary supremacy and the representative function of MPs. But now that referendums have been held for a national purpose and for devolution purposes - even down to the level of whether local government electors would support the creations of elected mayors - that historical resistance must, I think, be taken to have reduced, although plainly the constitutional position of Parliament and of MPs remain significant factors in the future of possible referendums at national level. Indeed, the present Government is committed to further national referendums, of its own choosing, on the adoption of the euro, and on any possible change to the Westminster voting system.

7.  Currently, a debate is going on about whether there should be a referendum about the proposed European Constitution. The Government opposes one on the grounds that it is an insufficiently important question, given the nature of the document itself, given that the main issue about the United Kingdom's relationship with the then European Community was settled in 1975, and given the promise of a future referendum on the single currency issue. Others argue for a referendum, on grounds which include the importance of the treaty as a fresh statement of the fundamental rules for the European Union, and the nature of the changes which would be made by it. That these arguments can be made on either side demonstrate the inadequacy of the present ad hoc situation concerning referendums. The drawbacks of that situation include the following.

  (i)  Political - sometimes, as currently, party-political - arguments can break out about whether a particular constitutional question should or should not be put to a referendum. While political debate is to be encouraged, it should not have to rage around the working rules of the constitution in particular cases.

  (ii)  The traditional, ad hoc, approach to referendums enhances the control of the Government of the day over constitutional development. Ministers have the final say about whether there will be a referendum: the Government decides purely at its own discretion whether a poll should take place. Constitutional arrangements, in the main, should ideally not be at the disposal of the Government of the day, and should not allow it to pick and choose which method should be used for a particular change.

  (iii)  A Government which is free to decide whether to hold a referendum may be tempted to do so for what, in purely constitutional terms, is an illegitimate reason. The 1975 referendum was designed as much to hold the then Government together as to get an answer to a constitutional question - although a referendum at some stage (and ideally before entry to the then EC) was, in my opinion, constitutionally desirable. And a Government may be tempted to resist a referendum on an equally illegitimate - though entirely understandable - ground, namely, that it might produce the "wrong" result for Ministers. It has been claimed that worry about the outcome is a crucial part of the present Government's current resistance to a referendum on the European Constitution. Standing rules could largely remove the Government from the decision whether to hold a referendum on a given issue.

  (iv)  The ad hoc approach produces inconsistent decisions about whether to hold a referendum. Examples were given above in paragraph 3. If there are to be any referendums at all, matters that are broadly similar in constitutional importance should be subject to broadly the same mechanism for resolution: questions of broadly similar importance above a certain threshold should be subject to a referendum. How that threshold should be defined could be decided by new rules, or possibly could be left to the judgement of a body independent of government.

  (v)  The ad hoc approach to referendums, requiring an ordinary statute to authorise a particular referendum, contributes to the view that constitutional change is qualitatively the same as change in any other area of public policy. It is not: the rules under which a state is governed are (or should be) of a different quality to the rules that are produced to implement particular areas of public policy. Constitutional rules are (or should be) different from other rules, and they should be established in ways which underline that difference. In other words, constitutional rules should be established and changed in ways which are unique to them, including for the most important by the use of a referendum.

8.  Those drawbacks would be reduced if agreement could be reached about what should trigger a national referendum. Rules could be incorporated into legislation which would provide standing authority for any future national referendums. I do not say here what those rules might be. There are several approaches to writing them. The rules might require a national advisory referendum before existing and listed constitutional legislation was amended or repealed. Or the rules might require a referendum if a proposed change would affect specified areas of the constitution, such as the monarchy, or the composition or powers of either House of Parliament, and so on. Or the rules might require a specified body, perhaps using criteria listed in the rules, to judge whether a constitutional referendum should be held. The Constitution Committee itself could be one candidate for such a role. What I am clear about is that the current debate about whether, and if so how, the United Kingdom should ratify the proposed European Constitution will intensify rather than abate. It is, I believe, a perfect illustration of why more work should be done on the place of national referendums in the United Kingdom constitution.

29 August 2003

115  Ibid, pp 50-52^. Back