Wales Says Yes: Devolution and the 2011 Welsh Referendum

Wales Says Yes

Wales Says Yes

Devolution and the 2011

Welsh Referendum

Richard Wyn Jones and Roger Scully

UNIVERSITY OF WALES

CARDIFF

2012

© Richard Wyn Jones and Roger Scully, 2012

All rights reserved. No part of this book may be reproduced in any material form (including photocopying or storing it in any medium by electronic means and whether or not transiently or incidentally to some other use of this publication) without the written permission of the copyright owner. Applications for the copyright owner’s written permission to reproduce any part of this publication should be addressed to the University of Wales Press, 10 Columbus Walk, Brigantine Place, Cardiff CF10 4UP.

www.uwp.co.uk

British Library CIP Data

A catalogue record for this book is available from the British Library.

ISBN     978-0-7083-2485-1

e-ISBN  978-0-7083-2642-8

The right of Richard Wyn Jones and Roger Scully to be identified as authors of their contributions has been asserted in accordance with sections 77 and 79 of the Copyright, Designs and Patents Act 1988.

i Gwenan Creunant

Cyfaill, cefn, Cymraes

Contents

Preface

1. The road to the referendum

2. The unlikely survival of the platypus: constitution building in Wales

3. The evolution of public attitudes

4. From coalition agreement to polling day

5. The referendum result

6. The people’s choice: explaining voting in the referendum

7. The implications

Appendices

Notes

Bibliography

Figures

6.1 Effects of significant variables on predicted probabilities of voting in referendum

6.2 Intended turnout (rolling three-day average): % ‘certain to vote’

6.3 Referendum voting intention (rolling three-day average) %

6.4 Effects of significant variables on predicted probabilities of voting yes in referendum

Tables

3.1 Constitutional preferences (%) in Wales and Scotland, 1970: Crowther/Kilbrandon Royal Commission Survey

3.2 Opinion polls on Welsh devolution in the 1970s

3.3 The 1979 Welsh devolution referendum

3.4 Constitutional preferences (%) in Wales, 1979

3.5 Constitutional preferences (%) in Wales, 1983–96

3.6 Opinion polls on the 1997 referendum

3.7 The 1997 Welsh devolution referendum

3.8 Constitutional preferences (%) Wales, 1997–2007

3.9 Constitutional preferences Wales, 2010

3.10 Most influence over ‘the way Wales is run’ (%), 2001–10

3.11 Referendum voting intention by constitutional preference, May 2010

3.12 Opinion polls on the 2011 referendum

5.1 The 2011 Welsh devolution referendum

5.2 Turnout in major UK referendums

5.3 Correlations with local authority 2011 referendum turnout

5.4 Multivariate analyses of local authority 2011 referendum turnout

5.5 The final polls compared

5.6 1997–2011 ‘swing’ by area

5.7 Correlations with % Yes in local authorities, 2011

5.8 Multivariate analyses of % Yes in local authorities, 2011

6.1 Reported referendum participation by social characteristics

6.2 Reported referendum participation by political attitudes

6.3 Rival models of 2011 referendum participation

6.4 Pre-referendum vote intention, by likelihood to vote, among non-voters

6.5 Referendum campaign contact

6.6 Forms of referendum campaign contact

6.7 Campaign perceptions

6.8 Party cues in the referendum (%)

6.9 Devolution attitudes (%) before and after the referendum

6.10 Referendum vote by social characteristics

6.11 Referendum vote by political attitudes

6.12 Rival models of 2011 referendum voting

Preface

The referendum that took place in Wales on 3 March 2011 was a curious event. As we discuss here, there are good reasons to argue that it should never have happened. And when the referendum did occur, the clear majority of the Welsh electorate chose not to bother participating. Yet, underwhelming as it was in many respects, the referendum was also an event of signal importance to Wales.

The referendum was important because the process by which this rather bizarre vote came to occur encapsulates so much about char­acteristic pathologies of Welsh political life. The referendum was also important because the way in which the campaign was fought, and the result that the vote produced, crystallized deeper and broader changes in Welsh politics over the previous decade. The referendum mattered because of the direct consequences that the result would bring over subsequent years for the policies of the Welsh Government and the responsibilities of the National Assembly for Wales. The referendum was also important because of the further consequences, for how Wales and the other nations of these islands are governed, that may well follow from the result.

For all these reasons, the 2011 Welsh referendum deserves serious, sustained analysis. In this book, we attempt to provide that: focusing not only on the vote itself, and the immediate campaign preceding it, but also trying to place this vote in its broader histor­ical, institutional and political context. The referendum was an important event, but as we have come to understand all too well in Wales, devolution is a process.

The research on which this study was based has received support from a number of quarters. We gratefully acknowledge financial support from the Economic and Social Research Council (grant RES-000-22-4496) to conduct the 2011 Welsh Referendum Study (WRS), which provides the bulk of the data used in chapter six. WRS was conducted in collaboration with the survey company YouGov: at various stages of their work on this project, Laurence Janta-Lipinski, Joe Twyman, Coralie Pring and Kate Davies at YouGov all distinguished themselves for their professionalism and their determination to help us deliver the best surveys possible. We are also very grateful for the helpful suggestions on the content of the surveys made by Robert Johns of the University of Essex, and by participants in a seminar at Cardiff University in January 2011.

We are also happy to thank the McDougall Trust for a small grant which supported the conduct and transcription of the interviews with those involved in the referendum campaign that provide much of the evidence discussed in chapter four. We are very grateful to our interviewees: Leighton Andrews, Rachel Banner, John Broughton, Nigel Bull, Meilyr Ceredig, Cynog Dafis, James Davies, Nigel Dix, Mark Drakeford, Lisa Francis, Len Gibbs, Daran Hill, Rob Humphreys, Roger Lewis, Cathy Owens, Lee Waters and Rebecca Williams, as well as some who spoke to us on condition of anonymity. They were, without exception, generous with their time and their insights. Given that some of the events discussed in the book are so recent and politically contentious, there have inevitably been some instances in which we had to rely on the testimony of our anonymous sources. Where we have been unable for this reason to supply a named source we have in each case sought to verify the claims made with two different (if anonymous) sources. We also thank Einion Dafydd and Catrin Wyn Edwards for their assistance in conducting the interviews, and Laura Considine and Lorrae Wright for help with transcription. Such is the richness of the material supplied in many of these interviews that we will be depos­iting edited copies with the Welsh Political Archive at the National Library of Wales in Aberystwyth, in order that they be available to future researchers. Some of the archival and interview material utilized in chapters one and two was collected as part of an earlier research project undertaken by Richard Wyn Jones on behalf of the University of Wales Board of Celtic Studies. We would like to thank both the Board and project researcher Bethan Lewis for their support. A final note on sources: all the website URLs referred to were current as of August 2011.

Aspects of the research reported here were presented in draft form at the annual conference of the Political Studies Association in London, April 2011; at breakfast seminars in Cardiff and Aberystwyth in June 2011; and at the annual conference of the Elections, Public Opinion and Parties specialist group in Exeter, September 2011. We thank participants in all those events for helpful comments and criticisms. We also thank Dafydd Trystan Davies, Lee Waters, Joanne Foster and an anonymous reviewer for their detailed comments on individual chapters. We are wholly responsible for any errors or inadequacies that remain.

We also gratefully acknowledge support and assistance from our home institutions, Aberystwyth University and Cardiff Universy, and particularly our colleagues in the Institute of Welsh Politics at Aberystwyth and Cardiff’s Wales Governance Centre. And we thank Sarah Lewis and the rest of the team at the University of Wales Press for all their assistance.

Last, but certainly not least, we thank our families for their consistent support, encouragement and love.

Roger Scully

Aberystwyth

Richard Wyn Jones

Treganna, Caerdydd

1

The road to the referendum

In April 2009, the House of Lords Select Committee on the Constitution published a report entitled Referendums in the United Kingdom.¹ A key issue for the committee – a group of genuine distinction – was the question of when it is appropriate for referen­dums to be held.² This is a question of relevance to many political systems, and one worthy of serious consideration.

Referendums have traditionally been regarded as foreign to the ‘Westminster model’ of parliamentary democracy. This model is one based on indirect democracy. It is characterized by the election of representatives who campaign on the basis of party manifestos. Once returned to parliament it is the task of members of the majority party to form a government to implement the platform on which it was elected. While the role played by the electors is funda­mental to the democratic legitimacy of the system, it is nonetheless limited. Beyond their infrequent visits to the polling booth, the people delegate the task of government to others. They are not invited to pick and choose between different elements of any party’s platform. Rather, the governing party is regarded as enjoying a mandate to implement its programme as a whole.

It is obvious that referendums sit uneasily within such a model. Referendums confer on the electorate direct power of decision on a specific issue: one that is deliberately, some might say artificially, isolated from other concerns. When implemented within a Westminster-style parliamentary democracy, referendums also operate outside the broader context and culture of civic engage­ment and direct democracy that characterizes a system such as Switzerland, where referendums are a regular occurrence.³ It is little wonder, then, that in the UK and its fellow Westminster-type democracies, referendums have traditionally been viewed with suspicion, and they have remained rarely used. The 1975 vote on continuing membership of the ‘Common Market’ (as it was then dubbed) was the first ever UK-wide referendum. Only one further such referendum has since occurred: that on the adoption of the Alternative Vote (AV) electoral system for UK general elections, some thirty-six years later. Moreover, the latter came about through highly exceptional circumstances: the formation of the first UK peacetime coalition government since the days of Ramsay MacDonald, and concerning a policy that had not featured in the manifestos of either coalition partner in the preceding general election. Thus, one might reasonably regard the AV ballot as an exception proving the general rule: that referendums do not fit comfortably within, and are not normally a prominent part of, a Westminster democracy.

Wales has greater experience of referendums than the rest of the UK. Local referendums on Sunday drinking were an intermittent feature of Welsh life from the time of the 1961 Licensing Act (which repealed the uniform provisions of the 1881 Sunday Closing (Wales) Act) until the requirement for such votes was finally abandoned in 2003.⁴ More relevant to our concerns are the two nation-wide refer­endums on devolution held in 1979 and 1997. These votes, which we discuss in much greater detail below, were part of the evidence considered by the Lords’ Committee as they examined previous experience with referendums in the UK, and the prospect that referendums might become more frequent. Their Lordships’ overall conclusion was clear:

The balance of the evidence that we have heard leads us to the conclu­sion that there are significant drawbacks to the use of referendums. In particular, we regret the ad hoc manner in which referendums have been used, often as a tactical device, by the government of the day. Referendums may become a part of the UK’s political and constitu­tional practice. Where possible, cross-party agreement should be sought as to the circumstances in which it is appropriate for referen­dums to be used.

Notwithstanding our view that there are significant drawbacks to the use of referendums, we acknowledge arguments that, if referendums are to be used, they are most appropriately used in relation to funda­mental constitutional issues.

The committee did not believe it possible to give a precise defin­ition of a ‘fundamental constitutional issue’. But they did provide illustrative examples, including abolition of the monarchy, secession from the Union and the abolition of either House of Parliament.⁵

One of the central arguments of this book is that the referendum held in Wales on 3 March 2011 illustrates almost perfectly many of the problems that the House of Lords Select Committee identified. The referendums of 1979 and 1997 concerned the fundamental question of whether or not Wales should have devolution. By stark contrast, the subject of the 2011 vote – a choice between two systems of granting primary law-making powers to the National Assembly for Wales – cannot remotely be considered as constituting a funda­mental constitutional issue. Moreover, the referendum itself resulted from tactical manoeuvring, made to overcome internal divisions within one party, rather than following any all-party agree­ment or consensus. Perhaps predictably, far from being a shining example of participative democracy in action, the resulting campaign was generally uninspiring and at times dispiriting. And the referendum result left important questions unanswered; some even unasked.

In this opening chapter, we set the context for the 2011 Welsh referendum. We begin in very general terms, by reviewing briefly how referendums are understood as a form of democratic decision making. In particular, we discuss how the apparent simplicity of a straight Yes/No vote on a single issue is often complicated by appar­ently extraneous factors, thus leading many people to vote for reasons other than their views on what is ostensibly at stake.

Our focus then moves more specifically to Wales. We begin by examining the 1979 referendum. We consider the circumstances that led to the holding of a vote on devolution. We also draw atten­tion to the broader political and social context in which the vote was eventually held, the nature of the opposing campaigns and the overwhelming final result.

In the next section of the chapter, we move on to consider the 1997 referendum. We again spend some time considering the pecu­liar circumstances that led to the decision to hold a referendum. We also review the nature of the respective campaigns, and the context in which the vote was held, before explaining the very different result that transpired from eighteen years previously.

The final section of the chapter addresses the 2006 Government of Wales Act and its aftermath. In particular, we explore the gesta­tion of the 2006 Government of Wales Act and the decision to include within its provisions a referendum on a move to extended legislative powers as provided by Part IV of the Act. We conclude by discussing how the referendum provision of the 2006 Act became central to coalition negotiations between Labour and Plaid Cymru after the 2007 National Assembly election, and thus helped pave the way for a referendum to be held in 2011.

Referendums across the world

While the political systems of the contemporary world can be classi­fied according to many criteria, probably the most common distinction is that drawn between democracies andnon-democracies. Democratic political systems supposedly embody the principle of government by the demos, the people. In practice, the defining characteristic of these systems is usually understood to be that those who hold major political offices are chosen via free and fair elec­tions in which the great majority of the people are able to participate. The system is democratic not because the people directly make the major decisions themselves, but because they get to choose the decision-makers. Non-democratic systems differ because the rulers are not elected at all, or because the elections supposedly used to select them are significantly lacking in freedom and/or fairness. Of course, in practice, all democracies fall short of the ideal. The conduct of elections will usually be imperfect to some degree. And much power in a society is typically wielded by people who are not elected: those who lead powerful private-sector economic organiza­tions and those occupying significant public roles that are not filled via election (such as senior judicial roles, or the headship of central banks). But the regular conduct of reasonably free and fair elec­tions to fill at least some key political offices usually suffices for a political system to be regarded by analysts, and by its own people, as being broadly democratic.

Making free and fair elections the defining characteristic of a democratic political system implies an understanding of democracy as an indirect process. The people are not expected to decide for themselves; they choose those they wish to make the major decisions on their behalf. At subsequent elections, they may then pass judge­ment on the decisions and behaviour of those previously chosen as their representatives. The use of referendums does not obviously fit easily within this understanding. A referendum takes the power of decision over a specific question back from the elected representa­tives and returns it to the people, who decide a matter directly.

The use of direct democracy within political systems founded upon indirect democracy raises numerous questions. To give the people direct power of decision over one particular issue, out of the multitude that a political system may be grappling with at the time, suggests that the issue is in some way special or distinct. Analysts and citizens are likely to ask why this specific matter should be isolated from the normal procedures of indirect democracy and given to the people to decide. What is different about this issue from others?

A second set of questions may be raised about the specific nature of the choice that a referendum offers the people. While some issues may present two very obvious alternatives between which the people can be asked to choose, in many cases there will be consider­ably more room for discretion about the number of alternatives to be placed on the ballot, which particular options these will be, and how they will be phrased. As scholars of politics have long under­stood, considerable influence can be wielded by those with agenda-setting power: the ability to shape the nature of the options between which people must choose.

Many other questions can be asked about referendums as a form of democratic decision making. Should the referendum outcome constitute a legally binding decision for the political community, or merely be something for holders of political power to take into account? If a referendum result is binding, for how long should it remain so? Under what circumstances might the people legitimately be asked to reconsider their original verdict in a second vote? And if it is accepted that some issues should be given to the people to decide directly, then what remains the proper realm of representa­tive democracy? And what, if any, role should representatives have in direct democracy?

Although their use can be seen to raise these and other funda­mental questions, referendums have become increasingly prevalent within the democratic world.⁶ Very few of the established major democracies have never held a nation-wide referendum;⁷ among those that have not, the United States has made frequent use of referendums in many of its states. Nonetheless, the frequency with which referendums are held continues to vary substantially across political systems. In most democracies they remain much less common than representative elections.

There are a very small number of instances – of which Switzerland is by far the most well-known example – of plebiscitary democracies, where the frequent use of referendums to make many major polit­ical decisions, both at the national and sub-national level, has been built into the design of the political system. Elsewhere, referendums are used less frequently, and most commonly in particular types of situations. Referendums will often be used to approve or reject a new constitution for a political system, or for proposed constitu­tional amendments. Constitutions may also require them to be held to approve certain types of decision: for instance, provisions in the Irish constitution have necessitated a series of public votes over European Union treaty revisions in recent decades. Some political systems provide for citizen-led referendums to be held: usually requiring a certain number of validated signatures supporting a particular ‘initiative’ for it to be placed on the ballot paper. California is perhaps the most well-known example where such initiatives are frequently used, and have had major political consequences.⁸ In many other political systems there are few requirements to hold popular votes to decide major political issues. But referendums have often been used by political leaders as a means of resolving otherwise intractable political disputes: a governing political party that is deeply divided over an issue may find it highly convenient to ‘let the people decide’ rather than having to come to an agreement internally.⁹

How people vote in referendums remains rather less well under­stood than voting in elections. In some contexts, voting choices appear to be little more than a reflection of group identities. In the March 2011 South Sudanese referendum that overwhelmingly endorsed independence from the rest of Sudan, voting patterns appeared primarily driven by the fundamental rejection of Sudanese identity by nearly all the southern people. But such factors can be important not only in supposedly ‘tribal’ societies: national identi­ties and broader national sentiments were highly influential on voting in the 1992 Canadian and the 1980 and 1995 Quebec refer­endums.¹⁰ Similarly, those with a more Welsh identity were substantially more likely to have voted in favour of devolution for Wales in 1997 than those with a British identity.¹¹

Although parties are not ‘on the ballot’ in referendums, the obvi­ously political nature of many referendum questions means that parties often take stances on many referendum issues. And voters often appear to heed such stances. The cues offered by parties to voters are central to many explanations of referendum voting deci­sions.¹² The willingness of many voters to be guided by the positions taken by major parties and political leaders often appears to func­tion as a cognitive short cut, compensating for voters’ lack of knowledge on the issue ostensibly at stake.¹³ However, the type of cue followed may vary. Popular parties and leaders can attract support to causes that they endorse: such a process seemingly shaped the outcome of the UK’s 1975 referendum on European Community membership.¹⁴ But the electorate may also use a refer­endum to register discontent with an unpopular party or leader, as with President Mitterrand and the French Socialists in the 1992 French referendum on the Maastricht Treaty.¹⁵ Voters can even use referendums to indicate unhappiness with an entire political class, as occurred in the 1992/1993 New Zealand referendums on elec­toral reform.¹⁶

Finally, how people vote in a referendum may, of course, be driven by their views on the question on the ballot paper. The extent to which many people will have settled or deeply rooted views on the matter at hand will likely depend heavily on the nature and history of the issue. Does the referendum concern a matter that has been the subject of public debate for many years? If this is so, or to an even greater extent if the referendum relates to one of the defining cleavages in a society, then large proportions of people can be expected to have deeply rooted and well-developed opinions about the matter. If the referendum concerns a rather more esoteric issue, or one about which there is little tradition of widespread public debate, then few beyond some of the most highly educated and politically engaged can be expected to have thought much about the question. We certainly cannot assume that the votes cast in referendums will always reflect the settled will of the people on the question placed before them.

The 1979 devolution referendum

The Labour Party