Revolution or Evolution? The 2007 Scottish Elections 9780748642151

Citation preview

__________________________________________

REVOLUTION OR EVOLUTION? THE 2007 SCOTTISH ELECTIONS

__________________________________________

REVOLUTION OR EVOLUTION? THE 2007 SCOTTISH ELECTIONS

JOHN CURTICE, DAVID MCCRONE, NICOLA MCEWEN MICHAEL MARSH AND RACHEL ORMSTON

EDINBURGH UNIVERSITY PRESS

© John Curtice, David McCrone, Nicola McEwen, Michael Marsh and Rachel Ormston, 2009 Edinburgh University Press Ltd 22 George Square, Edinburgh www.euppublishing.com Typeset in Perpetua by Lesley Birse, and printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne A CIP record for this book is available from the British Library ISBN 978 0 7486 3899 4 (hardback) ISBN 978 0 7486 3898 7 (paperback) The right of John Curtice, David McCrone, Nicola McEwen, Michael Marsh and Rachel Ormston to be identified as authors of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

__________________________________________

CONTENTS

List of Tables Preface CHAPTER 1 A Watershed Election? CHAPTER 2 A Distinctive Scottish Voice? Identities, Values and Attitudes CHAPTER 3 What has Devolution Achieved? The Public’s View CHAPTER 4 Governing Scotland: The People’s Preferences? CHAPTER 5 Lost Labour Votes? Records, Personalities and Issues CHAPTER 6 How the SNP Won CHAPTER 7 Do Voters Care about Parties Any More? CHAPTER 8 A Personal Vote? How Voters Used the STV Ballot CHAPTER 9 Conclusion Technical Appendix Index

vii xi 1 17 42 56 72 104 126 158 182 193 209

  __________________________________________

TABLES

Chapter 1 1.1 Scottish Parliament Election Results, 1999–2007 1.2 Summary of Local Election Results, 2003 and 2007 Chapter 2 2.1 Trends in ‘Free Choice’ National Identity, 1997–2007 2.2 Trends in ‘Forced Choice’ National Identity, 1974–2007 2.3 Trends in Moreno National Identity, 1992–2007 2.4 Indicators of Left–Right Values in Scotland and England, 2000–7 2.5 Mean Left–Right Score by Moreno National Identity, 2000 and 2007 2.6 Indicators of Liberal–Authoritarian Values in England and Scotland, 2007 2.7 Trends in Attitudes towards University Tuition Fees, Scotland and England, 2000–7 2.8 Attitudes towards Paying for Personal Care, Scotland and England, 2005 and 2007 2.9 Attitudes in Scotland and England towards Private Companies Providing Public Services, 2007 2.10 Trends in Attitudes towards Economic Issues, Scotland and England, 1999–2006 Chapter 3 3.1 Expectations of Devolution, 1997 and 1999 3.2 Perceived Impact of Devolution on the Health Service and Education, 2001–7 3.3 Perceived Impact of Scottish Parliament on Ordinary People’s Say in how Scotland is Governed, 2001–7 3.4 How Good is Scottish Executive/UK Government at Listening to People before Taking Decisions? 3.5 Perceived Impact of Devolution on Scotland’s Voice in the UK, 2001–7  

vii

3 10 19 20 22 26 28 29 31 32 33 35 44 46 47 48 50

Revolution or Evolution? 3.6 Perceptions of Whose Economy Benefits More from Scotland Being Part of the UK, 2000–7 3.7 Perceptions of Scotland’s Share of Government Spending Compared with Other Parts of the UK, 2000–7 Chapter 4 4.1 Constitutional Preferences, 1997–2007 4.2 Perceived Impact of Scottish Parliament on Scotland’s Membership of the UK, 1997–2007 4.3 Who Should and Does have Most Influence in Scotland? 1999–2007 4.4 Attitudes towards Giving the Scottish Parliament More Powers, 1999–2007 4.5 Which Institution should Make Important Decisions about…? 4.6 Attitudes towards Funding Services in Scotland out of Taxes Collected in Scotland, 2001–7 4.7 Attitudes towards Scottish Ministers Rowing with UK Government by National Identity and Attitudes towards How Scotland should be Governed, 2007 Chapter 5 5.1 Overall Evaluations of Labour’s Performance in the UK Government and Scottish Executive, 2007 5.2 Evaluations of Policy Performance, 2003 and 2007 5.3 Attributions of Responsibility, 2007 5.4 Labour Loyalty by Evaluations of UK Government and Scottish Executive 5.5. Logistic Regression of Impact of Overall Evaluations on Voting Labour in 2007 5.6 Labour Loyalty by Evaluations of the NHS and Standard of Living 5.7 Logistic Regression of Impact of Policy Evaluations on Voting Labour, 2003 and 2007 5.8 Evaluations of Party Leaders, 2003 and 2007 5.9 Labour Loyalty by Evaluations of Labour Party Leaders 5.10 Logistic Analysis of Impact of Leadership Evaluations on Voting Labour, 2003 and 2007 5.11 The Balance of Opinion on the Issues of the Campaign 5.12 Vote by Policy Preference on Tuition Fees and Iraq 5.13 The Link between Vote and Policy Preference  

viii

51 52 58 59 61 62 63 64 67 76 77 79 80 82 83 84 87 88 89 93 94 96

Tables Chapter 6 6.1 Demographic Profile of SNP Support, 2003 and 2007 6.2 National Identity and Nationalist Support, 1999–2007 6.3 Constitutional Preference and Nationalist Support, 1999–2007 6.4 Attitudes towards More Powers and Nationalist Support, 1999–2007 6.5 What Is Uppermost in Voters’ Minds? 6.6 Constitutional Preference and Evaluations of SNP Leaders, 2003–7 6.7 Logistic Regression of SNP Support, 2007 6.8 Support for SNP and Evaluations of Tony Blair, 2003–7 Chapter 7 7.1 Strength of Party Identification, 1979–2007 7.2 Voters Voting for Different Parties across Different Kinds of Election/Ballot, 1999–2007 7.3 Perceptions of Difficulty of Completing Scottish Parliamentary Ballot Paper(s), 1999–2007 7.4 Voting for Other than First-Choice Party, 1999–2007 7.5 Ticket Splitting by Perceived Difficulty of Ballot Paper and Votes/Seats Relationship, 1999–2007 7.6 Not Voting for First Choice by Perceived Difficulty of Ballot Paper and Votes/Seats Relationship, 1999–2007 7.7 Not Voting for First Choice by Highest Educational Qualification, 1999–2007 7.8 Ticket Splitting by Strength of Party Identification, 1999–2007 7.9 Not Voting for First Choice by Party Identification, 1999–2007 7.10 Ticket Splitting by Preference for Single-Party or Coalition Government, 1999–2007 7.11 Reported Motivation for Constituency Vote, 1999–2007 7.12 Not Voting for First Choice by Motivation for Constituency Vote, 1999–2007 7.13 Ticket Splitting by Motivation for Constituency Vote, 1999–2007 Chapter 8 8.1 Number of Candidates and Parties Voted for 8.2 Voting for Multiple Candidates and Parties by Partisanship 8.3 Number of Candidates and Parties Voted for by Characteristics of First-Preference Party  

ix

107 108 110 111 113 114 117 119 130 132 134 135 137 138 139 140 140 143 146 147 148 162 163 165

Revolution or Evolution? 8.4 Loyal Party Voting by Partisanship 8.5 Perceptions of Difficulty of STV Ballot Paper and Understanding Allocation of Seats 8.6 Use of STV Ballot Paper by Highest Educational Qualification and Perceived Difficulty of Electoral System 8.7 Incidence of Alphabetic Voting in Scotland by Indicators of Capacity 8.8 Solidity of Party Support by Party of First-Preference Vote 8.9 Relative Liking of Parties and Candidates 8.10 Number of Parties Voted for by First-Preference Vote 8.11 Transfer Structure of Voters’ Preferences Technical Appendix A.1 2007 Scottish Social Attitudes Survey Response A.2 Confidence Intervals for Survey Findings

 

x

167 169 170 173 175 176 177 178 196 203

  __________________________________________

PREFACE

On 7 May 2007 the Scottish National Party came first in a national parliamentary election for the first time ever. At the same time in parallel local elections the Single Transferable Vote was used for the first time in Scotland since 1945. This book tells the story of that momentous day. It does so not from the perspective of the politicians who hopes to secure power. Rather it examines what happened from the perspective of the ordinary voter in whose hands the future of those politicians lay. Shortly after the election, interviewers from the Scottish Centre for Social Research interviewed a random sample of 1,500 people who had been eligible to vote in the elections. In those interviews, which comprised the latest in an annual series of Scottish Social Attitudes surveys, respondents were asked a wide range of questions about themselves, their attitudes and the electoral choices they had made on election day. This book aims to uncover the story revealed by their answers - and thereby assess the true significance of the results that were declared once the ballot boxes had been opened. Such an enterprise cannot be undertaken without the use of statistics. We could not possibly summarise the answers provided by 1,500 people without them, let alone go on to use them to make claims about what happened amongst voters in Scotland as a whole. Yet for some potential readers any text that makes heavy use of statistics can seem both dry and daunting. We have approached our task of painting a picture of the electoral mood of Scotland with the feelings of such readers very much in mind. Our principal focus throughout is on the picture revealed by our statistics, not on the technicalities behind them. We thus hope our text will prove accessible to all those with an interest in Scotland, the future of the United Kingdom, nationalism, elections or electoral systems, whether they feel comfortable with statistics or not. A book based on so large a research enterprise inevitably incurs many debts. Our first debt is to those who funded the survey work. The principal funders of the material presented here were the Leverhulme Trust, whose trustees have consistently proved highly generous in their support of our research on public attitudes since devolution, and the  

xi

Revolution or Evolution? Economic and Social Research Council, the principal source of public funding of academic social science research in the UK and without whose support the Scottish Social Attitudes series of surveys would never have begun. We also make considerable use of questions funded by the Scottish Government, whose interest in ascertaining the public’s perceptions of the operation of devolution has played a vital in enabling us to chart the public mood between elections as well as at election time. Many people have played a vital role in the task of turning interviews into a text. Nicky Cleghorn and Catriona Webster of the Scottish Social Attitudes survey research team made important contributions to questionnaire design and the administration of the survey. Joan Corbett from the Scottish Centre for Social Research and Ann Mair at the Social Statistics Laboratory at the University Strathclyde did sterling work in ensuring that the resulting data could easily be analysed. The Local Government Boundary Commission for Scotland generously agreed to provide us with details of the local government wards in which our interviews were conducted, a task with its secretary, Bob Smith, performed with admirable efficiency. Sarinder Hunjan of the Social Statistics Laboratory meticulously compiled a database of the local election results in the wards covered by the survey. Brian Taylor, the Political Editor of BBC Scotland, generously gave up his time to chair a conference in October 2007 when some of the first fruits of our research were unveiled. Lesley Birse at the Scottish Centre for Social Research performed with unquestioning good humour the unenviable task of formatting and cleaning our text. And Eddie Clark, James Dale and Nicola Ramsey of Edinburgh University Press have been both supportive and patient in guiding us through the editorial process. But without those 1,500 interviews this book would not have been possible at all. We must thus extend our thanks above all to our interviewers who travelled the length and breadth of Scotland, tirelessly seeking out those who had been selected for interview, and our respondents who generously gave an hour of their time to answer our questions. We hope they feel their efforts have proved worthwhile. John Curtice David McCrone Nicola McEwen Michael Marsh Rachel Ormston September 2009  

xii

CHAPTER 1 __________________________________________

A WATERSHED ELECTION?

Two battles were being fought when Scotland went to the polls on 1 May 2007. The first contest was for the 129 seats in the devolved Scottish Parliament, the third such election to be held since the parliament was established in 1999. The second tussle was for 1,222 seats on all of the country’s 32 local councils. Both events appeared to represent a watershed, or even a revolutionary change in the nature of Scottish politics. In the parliamentary election the Scottish National Party (SNP), whose principal objective is that Scotland should leave the United Kingdom and become an independent state, emerged with most votes and most seats. It was the first time that the party, formed as long ago as 1934, had succeeded in coming ahead of all of its rivals in a nationwide contest. Apparently the warnings of those who argued that devolution would put Scotland on a ‘slippery slope’ to independence were proving all too prescient (Dalyell, 1977). The local elections, meanwhile, were fought using a new more proportional electoral system. Instead of using the single-member plurality (or first-past-the-post) system that had been in place ever since the reorganisation of Scottish local government in the 1970s, councillors were elected using the single transferable vote in three- or four-member wards. This new system allowed voters both to express support for more than one party and to choose between candidates of the same party. As a result it supposedly shifted the emphasis of the elections away from being contests between parties and towards being races between individual candidates. At the same time the more proportional outcomes produced by the system would end the days of single-party – primarily Labour – rule on most councils. This book examines whether these two developments do indeed represent a revolution in Scottish politics. Did the success of the SNP indicate a growing wish amongst people in Scotland to terminate their country’s membership of the United Kingdom? And did the introduction of the new electoral system in local government ensure that  

1

Revolution or Evolution? considerations of personality rather than party were now at the forefront of voters’ minds when deciding how to vote? Or should we, perhaps, adopt a less apocalyptic view of the significance of the 2007 elections? Perhaps voters backed the SNP because they were unhappy with the performance of the previous Labour/Liberal Democrat coalition, and used the devolved election to replace one seemingly ineffective regime with another that might do a better job? And perhaps despite whatever incentives the new electoral system might have given voters to vote on the basis of personality rather than party, perhaps party still mattered to them more? If it is these alternative perspectives that seem in practice to be closer to the mark, then perhaps the 2007 elections should be regarded as no more than a further stage in the evolution of Scottish politics rather than marking some revolutionary break with the past. THE NATIONALIST BREAKTHROUGH For many of its advocates, devolution was designed to strengthen Scotland’s position within the Union (Aughey, 2001; Bogdanor, 1999; Paterson, 1998; Mackintosh, 1998). In creating the Scottish Parliament the United Kingdom would demonstrate it could give recognition to the distinctive sense of national identity felt by many people in Scotland and meet the demand that the already wide-ranging powers of the Scottish Office (the UK government department responsible for much of what happened in Scotland) should be brought under more effective democratic control. As a result, it was argued, people in Scotland would come to accept that their country did not need to become independent in order to secure a sufficient degree of autonomy. Indeed according to the former Labour Shadow Scottish Secretary, George Robertson, the creation of the Scottish Parliament would kill nationalism ‘stone dead’.

2

A Watershed Election? Table 1.1 Scottish Parliament Election Results, 1999–2007

SNP Labour Conservatives Liberal Democrats Greens Scottish Socialists Others SNP Labour Conservatives Liberal Democrats Greens Scottish Socialists Others

% constituency vote (no. of seats) 1999 2003 2007 28.7 ( 7) 23.8 ( 9) 32.9 (21) 38.8 (53) 34.6 (46) 32.2 (37) 15.6 ( 0) 16.5 ( 3) 16.6 ( 4) 14.2 (12) 15.4 (13) 16.2 (11) 0.1 ( 0) 1.0 ( 0) 6.2 ( 0) 0.0 ( 0) 1.7 ( 1) 3.5 ( 2) 2.0 ( 0) % regional list vote (no. of seats) 1999 2003 2007 20.9 (18) 31.0 (26) 27.3 (28) 33.6 ( 3) 29.3 ( 4) 29.2 ( 9) 15.5 (15) 13.9 (13) 15.4 (18) 12.4 ( 5) 11.8 ( 4) 11.3 ( 5) 3.6 ( 1) 6.9 ( 7) 4.0 ( 2) 2.0 ( 1) 6.7 ( 6) 0.6 ( 0) 5.7 ( 0) 8.9 ( 2) 10.0 ( 1)

Note: Greens did not fight any constituencies in 1999 or 2003. Sources: Hassan and Fraser, 2004; Electoral Commission, 2007.

Table 1.1 shows how the SNP, together with the other parties, have fared in elections to the new devolved parliament since it was created in 1999. Elections to the Scottish Parliament are held using a two-ballot additional-member system rather than the single-member plurality system that is used in elections to the UK House of Commons (for further details see Curtice, 2006). One vote is cast in a single-member plurality election for a local Member of the Scottish Parliament (MSP), of whom there are 73. A second vote is cast for a regional party list; this determines the allocation of 56 ‘additional’ seats (seven in each of eight separate regions) that are distributed so that the total number of seats (constituency and list) won by each party is as proportional as possible to its share of the regional list vote. As there is no obligation on voters to give both their votes to the same party (or indeed any requirement for parties to nominate both constituency and regional list candidates), the table shows separately the distribution of the vote (and associated number of seats) on the two ballots. 3

Revolution or Evolution? We can see that from the very outset of the devolution project there was good reason to doubt the validity of George Robertson’s prediction. The SNP might have trailed Labour by some considerable margin in the 1999 election, but, nevertheless, its share of the constituency vote was greater than it had been at any previous election to the House of Commons apart from October 1974 (when it won 30.4%). Moreover the greater proportionality afforded by the ‘top-up’ or ‘additional’ element of the electoral system meant that the party’s performance was better rewarded than under single-member plurality where the even geographical distribution of the nationalist vote puts the party at a disadvantage. The nationalists’ meagre reward in 1999 of seven constituency seats was ‘topped up’ with a further 26 list seats. The resulting tally of 33 seats meant that the SNP now not only had a substantial parliamentary presence for the first time ever (it had never had more than eleven seats in the over-600-seat UK House of Commons), but also actually enjoyed the status of being the principal opposition party in the chamber (to a coalition between Labour and the Liberal Democrats). Such an outcome looked more like a boost to the SNP’s fortunes than the first leg in its demise. Moreover, opinion polls conducted in the run up to the 1999 election campaign revealed an intriguing phenomenon; people were more willing to vote for the SNP in Scottish Parliament elections than they were in elections to the House of Commons. For example, the first detailed poll of Scottish Parliament vote intentions, conducted exactly a year before the May 1999 polling date by ICM for the Scotland on Sunday newspaper, found that whereas only 27% said that they would vote for the SNP in any immediate House of Commons election, no less than 36% would do so on the constituency vote for the Scottish Parliament – and as many as 39% on the regional list vote. Such a pattern continued to persist in every reading of public opinion thereafter. Whatever the ups and downs of SNP popularity, it became apparent that the party was always more likely to prosper in Scottish Parliament elections than in House of Commons contests. So as well as using an electoral system that treated SNP support more favourably, devolved elections apparently also provided a forum in which the SNP found it easier to win votes in the first place. Given the apparently propitious environment created by the advent of devolved elections, the nationalist performance in 2003 was undoubtedly 4

A Watershed Election? a disappointment. Alex Salmond, the party’s charismatic leader since 1990, had surprisingly stood down in 2000 and opted soon after to leave the Scottish Parliament, focusing his energies instead on being an MP (and the party’s parliamentary leader) at Westminster. His replacement, John Swinney, struggled to make an impact on the general public. Compared with 1999, the party’s share of the vote dropped by five points on the constituency vote and by six on the list. Not that Labour thrived either. Rather the election was graced by a dramatic rise in support for smaller parties, most notably the Greens and the far-left Scottish Socialist Party (SSP). The success of these two parties, neither of which has ever been represented in the House of Commons, in winning between them 13 seats meant that the resulting body was soon dubbed a ‘rainbow parliament’. The SNP took comfort from the fact that, because both the Greens and the SSP also support independence, the proportion of the regional list vote cast for parties that backed independence had in fact risen by just over two points. Perhaps more importantly, the SNP had still performed rather better than the 20.1% it had won in 2001 UK general election, while it retained its position as the principal opposition party, and thus as the main alternative to the Labour/Liberal Democrat coalition that was returned to power. By 2007 circumstances had changed. After another disappointing result in the 2004 European election, John Swinney stood down as party leader. Then to great surprise Alex Salmond successfully put himself forward as leader once more and committed himself to fighting the 2007 Scottish Parliament election as his party’s candidate to be First Minister. Meanwhile by now the gloss had finally come off the UK Labour government at Westminster, which trailed the Conservatives in Britainwide polls. In the summer of 2006 a poll conducted by Ipsos-MORI put the SNP two points ahead in voting intentions for the next Scottish Parliament election, and this was followed in the autumn by a flurry of polls that between them suggested the SNP was at least neck and neck with Labour in the race to come first. Both the press and politicians began to realise that the SNP might well pose a real threat to Labour’s dominance of Scottish politics. Labour’s reaction to the apparent threat was to repeat the campaign tactics it had used with seeming success in the first devolved election in 1999 – to suggest that independence, as championed by the SNP, would pose a threat to Scotland’s prosperity and stability. This time around it 5

Revolution or Evolution? claimed, for example, that independence would cost every family in Scotland an extra tax bill of £5,000 a year. The alleged threat posed by the SNP was regularly invoked by the UK Prime Minister, Tony Blair, and Chancellor, Gordon Brown, whose extensive involvement in the election campaign tended to leave the Scottish First Minister, Jack McConnell, in the shade. By such tactics it was hoped that sufficient voters would be persuaded to draw back from voting for the SNP. In contrast, the SNP said relatively little about independence and focused instead on emphasising its ability to provide Scotland with effective (devolved) government. Under the slogan, ‘It’s time’, it set out a detailed prospectus for what it would do in government over the next four years (Scottish National Party, 2007). True, that included holding a referendum on independence, but not until the SNP had been in power for three years and demonstrated its ability to govern effectively. Meanwhile the party set out a detailed new departmental structure for government in Scotland and a programme that would focus on five instrumental objectives ranging from making Scotland healthier to making it smarter. In the event the nationalists did just enough to win. Although they won their highest ever share of the Scotland-wide vote, they only led Labour by less than two points on the regional list vote and by less than one percentage point in the constituency contests. Because the ‘top-up’ element of the electoral system did not fully reverse Labour’s advantage in winning constituency seats – despite trailing in the nationwide vote Labour still won 37 such seats to the SNP’s 21 – these margins were only sufficient to give the party one seat more than Labour. Indeed with just 47 seats the party was a long way indeed from having an overall majority, and could conceivably be denied office should Labour and the Liberal Democrats strike another coalition deal that also secured the acquiescence of the Conservatives. However, an attempt by the Liberal Democrats’ UK party leader, Sir Menzies Campbell, to instigate such negotiations came to naught (Campbell, 2008), while the party’s Scottish leadership soon signalled that they were also unwilling to enter a coalition with the SNP. The only possible government seemed to be a minority SNP administration. So with the Conservatives and Liberal Democrats abstaining and the Greens backing him, on 16 May 2007 Alex Salmond was elected First Minister, and the SNP secured its first ever taste of power. 6

A Watershed Election? It is then little wonder that we might be uncertain as to the significance of the nationalist breakthrough in 2007. On the one hand not only did the SNP secure its highest share of the vote ever in the face of a strong attack on independence by its opponents, but the result was also another piece of evidence that the introduction of devolved elections has created an environment in which the SNP persistently find it easier to prosper. On this reading devolution does seem to have instigated greater popular support for the independence cause. On the other hand, even though the incumbent UK Labour government was unpopular the SNP still only secured less than a third of the vote and only did so after fighting an election campaign as an alternative government rather than as a movement for independence. Not only does this suggest a continuing lack of enthusiasm for the nationalist cause, but perhaps indicates that voters were doing no more than using the SNP to protest against the perceived inadequacies of the UK Labour government and/or the incumbent Labour/Liberal Democrat Scottish administration. Evidently there is a need to dig deeper if we are to assess adequately the significance of the 2007 parliamentary contest. A NEW WAY OF VOTING? How Scotland’s local councils should be elected was the subject of dispute and debate from the very early days of devolution. Two events started the ball rolling. First, in 1999 a commission established by the Scottish Office on the proper relationship between the new Scottish Parliament and local government argued that a new more proportional electoral system should be used to elect the country’s 32 local councils (McIntosh, 1999). Second, the Liberal Democrats, who have long been in favour of proportional representation, succeeded in their coalition talks with Labour after the 1999 election in securing a commitment to making ‘significant progress’ on local government electoral reform. As a result the new administration appointed a second commission that was charged with the task of recommending what particular form of proportional representation should be used in Scottish local elections. It reported in favour of the single transferable vote (STV) in multi-member wards (Kerley, 2001). This in fact was the system favoured by both the Liberal Democrats and 7

Revolution or Evolution? the Electoral Reform Society, the principal organisation campaigning in favour of electoral reform in the UK. Progress thereafter, however, was relatively slow, not least because of reluctance within much of the Labour party to embrace a reform from which it would undoubtedly suffer (Curtice and Herbert, 2005; Curtice and Ritchie, 1999). By the time of the 2003 election, the Scottish Executive (as the devolved government was then known) had undertaken a consultation on the proposal and published a draft bill, but was still not committed to changing the system. However, after making reform one of their key demands, the Liberal Democrats did secure such a commitment from Labour in the 2003 coalition negotiations. The single transferable vote was to be introduced in three- and four-member wards and be in place for the 2007 local elections. The relevant legislation, part of the Local Governance (Scotland) Act, received the Royal Assent on 28 July 2004. Subsequently the Local Government Boundary Commission for Scotland formulated the necessary new ward boundaries which were all in place before the end of 2006. This switch to STV was not merely of interest within Scotland. As we have already noted STV has long been the preferred system of the Liberal Democrats and the Electoral Reform Society. It had also enjoyed some previous use in elections in the UK. In particular, from 1918 until their abolition in 1950 it was used to elect the handful of MPs that represented university seats, including the three seats elected by graduates of the ancient Scottish universities. It was also used to elect the separate Scottish local education authorities that existed between 1919 and 1929. However, more recently the system has only been used in Northern Ireland, where since the 1970s it has been deployed in all elections other than to the UK House of Commons. It has been the Additional Member System (AMS) rather than STV whose merits have in recent years been most successfully argued in debates about electoral systems in Great Britain. Not only is AMS used in elections to the Scottish Parliament but also in those to the National Assembly of Wales and the Greater London Assembly. However, if the introduction of STV in Scotland were to be judged a success the tide of the electoral reform debate might be turned back in favour of STV. STV is very different from either the two-ballot AMS system used in Scottish Parliament elections or the single-member plurality system previously used in Scottish local elections (Lakeman, 1970). Instead of 8

A Watershed Election? being asked just to place an ‘X’ against the name of an individual candidate or party, voters are invited to place the individual candidates standing in their ward in rank order. Those candidates may be standing in the name of particular parties, but this need not necessarily constrain the choice that voters make. They can express a preference for candidates from different parties, perhaps, for example, marking a ‘1’ against a Conservative candidate, ‘2’ against a Labour one, etc. Note too that as there is more than one seat at stake in each ward, more than one candidate may be standing for the same party. Nevertheless, a voter is free to express support for one such candidate but not another. And even if a voter does want to support more than one candidate from the same party, they are obliged to place them in a preferred order. In short, STV encourages voters to consider the merits of individual candidates rather than just the parties for which they are standing. Moreover, any switch from party- to personality-based voting could have important implications for the way in which councillors fulfil their role. No longer could a councillor simply rely on the popularity of their party locally in order to secure election. Rather, they would have to ensure they were personally popular too. Of course, such popularity might be secured in a variety of ways. But if, for example, voters valued someone who spoke up for the interests of their area and was assiduous in dealing with individual voters’ complaints about the local council, then councillors would find themselves under pressure to perform such tasks. Indeed they might even compete with each other in terms of the quality of the local service they provide. This possibility in fact addresses one of the key aspects in the debate about electoral reform in the UK. Advocates of single-member plurality argue that it provides a strong incentive for elected representatives to look after the interests of their constituency and constituents and that this feature would be lost under any multi-member system of proportional representation (Curtice and Shively, 2009; Farrell and McAllister, 2006). Proponents of STV, on the other hand, argue that STV provides an even stronger incentive for elected representatives to provide their constituents with a good local service (Carey and Shugart, 1995). So it matters whether the argument that STV encourages personality politics is correct or not. In practice, international experience suggests that while STV can encourage voters to focus on candidates rather than parties, there is no 9

Revolution or Evolution? guarantee that this will happen. On the one hand, in the Republic of Ireland, where STV has been used ever since the founding of the State in 1922, voters do often appear to vote on the basis of what they think of individual candidates rather than (or as well as) their parties and in so doing do take into account the quality of the constituency service they provide (Marsh, 2000; Sinnott, 1995). On the other hand, the position in both Malta and Australia is very different. In Malta very few voters are willing to give any kind of preference to a candidate from another party (Hirczy de Miño and Lane, 2000). In Australia (outside Tasmania at least) voters have long followed advice issued by parties on how to order their ballots (Reilly and Maley, 2000; Wright, 1986); indeed since 1983 the order in which parties would like voters to place the candidates has been indicated on the ballot paper together with an easy means of endorsing that order. So it seems that we cannot presume that the introduction of STV in local elections in Scotland will necessarily have persuaded voters to vote on the basis of personality rather than party – with possible implications for the role of councillors. If voters care only about party, then STV allows them to behave accordingly. They can just vote for the candidates of their preferred party and no other, and they are able to place those candidates in the order recommended by that party. Changing the electoral system can change the opportunities available to voters, but voters will only exploit those opportunities if they are motivated to do so. Table 1.2 Summary of Local Election Results, 2003 and 2007

SNP Labour Conservatives Liberal Democrats Independents Greens Scottish Socialists Others

2003 % vote 24.3 32.9 15.2 14.6 9.5 0.0 3.4 0.2

Source: Bochel and Denver, 2007.

10

seats 181 509 123 175 231 0 2 1

2007 % vote 27.9 28.1 15.6 12.7 10.7 2.2 0.9 1.8

seats 363 348 143 166 187 8 1 6

A Watershed Election? As expected, Labour did indeed lose out from the introduction of STV, though its losses were exacerbated by the fact that its share of the vote fell as compared with 2003 (see Table 1.2). Not only did the party win 161 fewer seats, but it was left in overall majority control of just two councils rather than 13 as in 2003. But radical though the partisan consequences of the new system might have been, they tell us nothing about the more subtle but potentially more profound possibility that its introduction might have encouraged voters to vote on the basis of personality rather than party.1 To assess that possibility we need to examine how voters used the STV ballot paper, and how their use relates to their views about the parties and candidates standing in their area. Here too we evidently need to dig more deeply. OUR EVIDENCE Our evidence on what underlay the nationalist success and how voters used the local election ballot paper comes from survey research. In the weeks and months immediately after the 2007 election, professional interviewers from the Scottish Centre for Social Research (ScotCen) interviewed, face to face, a representative sample of adults aged 18 and over as part of the Scottish Social Attitudes survey (for further details see the technical appendix). The survey covered a wide range of topics of interest to us here, including how people voted in the two elections, their views about policies, the parties and their leaders, and their attitudes towards devolution and independence. This means we can link how people behaved in the elections with a wide range of views and attitudes that might account for why they voted as they did. But, of course, if we wish, for example, to investigate whether support for independence has risen or fallen, we require to know more than people’s attitudes towards Scotland’s constitutional future in 2007. Rather, we need to know how their attitudes in 2007 compare with what they were in earlier years. Fortunately, the Scottish Social Attitudes survey was not only conducted after the 2007 election, but every year from 1999 onwards. This means we can chart how attitudes in Scotland have developed since the advent of devolution. In addition, we can compare our findings about the reasons as to why people voted as they did in 2007 with similar evidence for the devolved elections in 1999 and 2003 – and indeed on occasion the UK 11

Revolution or Evolution? House of Commons elections held in 2001 and 2005 too. Moreover, some of the questions included on Scottish Social Attitudes were previously asked in surveys conducted in Scotland after the 1992 and 1997 UK general elections (Bennie et al., 1997; Brown et al., 1999) and the referendum on whether the devolved Scottish Parliament should be created that was held in September 1997 (Taylor and Thomson, 1999), thereby enabling us to compare the position now with that immediately prior to the advent of devolution. Our analysis of these data falls into three parts. First of all, in Chapters 2 to 4 we look at how people in Scotland have reacted to the experience of eight years of devolved government. In so doing our primary goal is to establish whether devolution has strengthened or weakened the basis of public support in Scotland for the Union with the rest of the United Kingdom. In Chapter 2 we consider some of the underlying forces that could have helped undermine support for the Union. Perhaps, for example, the creation of the Scottish Parliament has influenced people’s sense of national identity? Perhaps they are now less likely to feel British, and thus an emotional affinity with the rest of the United Kingdom, and more likely to feel distinctively Scottish? Perhaps, too, devolution has resulted in debates about policy in Scotland being conducted entirely separately from the equivalent debates in England, with the result that people in Scotland increasingly have different policy preferences from their counterparts south of the border? Then in Chapter 3 we consider how people in Scotland have reacted to the experience of devolution. Do they feel that devolution has improved the way in which their country is governed? And has their experience changed their views about the way in which the Union operates? As we have seen, the advocates of devolution argued that people would come to feel that devolution has improved the way Scotland is governed, and that as a result they would be persuaded to hold a more benign view towards the Union. Our analysis puts that argument to the test. Thereafter in Chapter 4 we look directly at trends in attitudes towards how Scotland should be governed. Did the SNP’s success in the 2007 election really signify a growing demand for independence? And even if people in Scotland still have doubts about the merits of independence, does this necessarily mean that they are content with the degree of autonomy afforded their country by the current arrangements? But even if it were the case that people in Scotland want more 12

A Watershed Election? autonomy and that support for independence has increased, it does not necessarily follow that this is why the SNP achieved its breakthrough in 2007. After all, we should not assume that people who favour independence necessarily support the SNP – or vice-versa (Bond and Rosie, 2002; Bond and Rosie, 2003). As we have already noted, there were other reasons why voters might have opted for the SNP, such as dissatisfaction with the performance of the UK Labour government, unhappiness with the Labour/Liberal Democrat coalition in Scotland, or perhaps support for policies put forward by the SNP other than independence. So in the second part of the book we look more specifically at why people voted as they did in the parliamentary election. In Chapter 5 we examine the role that evaluations of the performance of government and voters’ attitudes towards some of the key policy debates that surfaced during the campaign played in determining how people voted. Then in Chapter 6 we focus specifically on how we can best account for the success of the SNP in 2007. In the third part of the book we shift our focus towards the local elections. One important implication appearing to flow from the international experience of STV is that how voters behave depends on how important parties are to them in the first place. If they feel a strong attachment to a party, and are inclined to support it through thick and thin, then perhaps there is little reason why the introduction of STV should persuade people to pay attention to the merits of individual candidates. But if they do not feel such an attachment then perhaps the change of electoral system might serve to weaken further the already loose ties between parties and voters yet further. So in Chapter 7 we consider recent trends in the strength of people’s attachments to parties in Scotland. In part this is achieved by looking at how strongly attached to a party people say they feel. But we can also garner valuable evidence by looking at how people behave under the two-vote AMS system used in Scottish Parliament elections. For voters who feel strongly attached to a party would seem unlikely to indulge in the practice of voting for different parties on the two ballots. Thereafter in Chapter 8 we examine how voters behaved in the local elections themselves. Did many people avail themselves of the opportunity to express support for more than one party? Or did most voters support all of the candidates of their preferred party and no others? In addressing these questions we are able to compare our findings directly with the evidence for 13

Revolution or Evolution? Ireland, whose 2002 National Election Study asked many of the relevant questions included in our 2007 survey. As a result we can directly assess how closely Scotland conforms to the Irish model of STV personality politics. In the concluding chapter we return to the original question posed at the beginning of this chapter. In the light of our analysis, does the 2007 election indeed mark a revolutionary break in Scotland’s links with the rest of the United Kingdom and in the importance of personality in local elections? Or does it simply mark an evolution in the country’s politics? Our quest for answers now begins with a look at the nation’s sense of identity. NOTE 1. For analyses of the partisan impact brought about by the change of system see Baston, 2008; Bennie and Clark, 2008; Curtice, 2007; Denver and Bochel, 2007; Electoral Reform Society, 2008. REFERENCES Aughey, A. (2001), Nationalism, Devolution and the Challenge to the United Kingdom, London: Pluto Press. Baston, L. (2008), ‘The Wider Impact of the 2007 Scottish Local Government Elections’, Representation, 44: 341–54. Bennie, L., Brand, J. and Mitchell, J. (1997), How Scotland Votes, Manchester: Manchester University Press. Bennie, L. and Clark, A. (2008), ‘The Transformation of Local Politics? STV and the 2007 Scottish Local Government Elections’, Representation, 44: 225–38. Bochel, H. and Denver, D. (2007), Scottish Council Election Results 2007: Results and statistics, Lincoln: University of Lincoln. Bogdanor, V. (1999), Devolution in the United Kingdom, Oxford: Oxford University Press. Bond, R. and Rosie, M. (2002), ‘National Identities in Post-Devolution Scotland’, Scottish Affairs, 40: 34–53. Bond, R. and Rosie, M. (2003), ‘Identity Matters’, in Bromley, C., Curtice, J., Hinds, K. and Park, A. (eds), Devolution – Scottish Answers to Scottish Questions?, Edinburgh: Edinburgh University Press. 14

A Watershed Election? Brown, A., McCrone, D., Paterson, L. and Surridge, P. (1999), The Scottish Electorate: The 1997 general election and beyond, London: Macmillan. Campbell, M. (2008), My Autobiography, London: Hodder and Stoughton. Carey, J. and Shugart, M. (1995), ‘Incentives to Cultivate a Personal Vote: a Rank Ordering of Electoral Formulas', Electoral Studies, 14: 417–39. Curtice, J. (2006), ‘Forecasting and Evaluating the Consequences of Electoral Change: Scotland and Wales’, Acta Politica, 41: 300–14. Curtice, J. (2007), ‘STV Goes Tartan: a Preliminary Analysis of its Use in the 2007 Scottish Local Elections’, Representation, 43: 209–16. Curtice, J. and Herbert, S. (2005), STV in Local Government Elections: Modelling the 2003 results, SPICE Briefing 05/31, Edinburgh: Scottish Parliament Information Centre. Curtice, J. and Ritchie, K. (1998), Electing Scotland’s Councillors, London: Electoral Reform Society. Curtice, J. and Shively, W. P. (2009), ‘Who Represents us Best? One Member or Many?’, in Klingemann, H.-D., The Comparative Study of Electoral Systems, Oxford: Oxford University Press. Dalyell, T. (1977), Devolution – The End of Britain?, London: Jonathan Cape. Denver, D. and Bochel, H. (2007), ‘A Quiet Revolution: STV and the Scottish Council Elections of 2007’, Scottish Affairs, 61: 1–17. Electoral Commission (2007), Scottish Elections 2007: Electoral administration issues arising from the Scottish parliamentary and local government elections 3 May 2007, Edinburgh and London: Electoral Commission. Electoral Reform Society (2008), Local Authority Elections in Scotland, 3 May 2007: Report and analysis, London: Electoral Reform Society. Farrell, D. and McAllister, I. (2006), ‘Voter Satisfaction and Electoral Systems: Does Preferential Voting in Candidate-Centred Systems Make a Difference?’, European Journal of Political Research, 45: 723–49. Hassan, G. and Fraser, D. (2004), The Political Guide to Modern Scotland, London: Politico’s. Hirczy de Miño, W. and Lane, J. (2000), ‘Malta: STV in a Two-Party System’, in Bowler, S. and Grofman, B. (eds), Elections in Australia, Ireland and Malta under the Single Transferable Vote, Michigan: University of Michigan Press. 15

Revolution or Evolution? Kerley, R. (chmn) (2001), Report of the Renewing Local Democracy Working Group, Edinburgh: Scottish Executive. Lakeman, E. (1970), How Democracies Vote: A study of majority and proportional systems, London: Faber and Faber. McIntosh, N. (chmn) (1999), The Report of the Commission on Local Government and the Scottish Parliament, Edinburgh: Scottish Office. Mackintosh, J. (1998), ‘A Parliament for Scotland’, reprinted in Paterson, L. (ed.), A Diverse Assembly: The Debate on the Scottish Parliament, Edinburgh: Edinburgh University Press. Marsh, M. (2000), ‘Candidate Centred but Party Wrapped: Campaigning in Ireland under STV’, in Bowler, S. and Grofman, B. (eds), Elections in Australia, Ireland and Malta under the Single Transferable Vote, Ann Arbor: University of Michigan Press. Paterson, L. (ed.) (1998), A Diverse Assembly: The debate on the Scottish Parliament, Edinburgh: Edinburgh University Press. Reilly, B. and Maley, M. (2000), ‘The Single Transferable Vote and the Alternative Vote Compared’, in Bowler, S. and Grofman, B. (eds), Elections in Australia, Ireland and Malta under the Single Transferable Vote, Ann Arbor: Michigan University of Michigan Press. Scottish National Party (2007), Manifesto 2007, Edinburgh: Scottish National Party. Sinnott, R. (1995), Irish Voters Decide: Voting Behaviour in Elections and Referendums since 1918, Manchester: Manchester University Press. Taylor, B. and Thomson, K. (1999), Scotland and Wales – Nations Again? Cardiff: University of Wales Press. Wright, J. (1986), ‘Australian Experience with Majority-Preferential and Quota-Preferential Systems’, in Grofman, B. and Lijphart, A. (eds), Electoral Laws and their Political Consequences, New York: Agathon.

16

CHAPTER 2 __________________________________________

A DISTINCTIVE SCOTTISH VOICE? IDENTITIES, VALUES AND ATTITUDES Establishing the Scottish Parliament was always potentially a risky enterprise for those who wanted to maintain the Union between Scotland and England. There were at least two key ways in which its possible impact on public opinion north of the border might undermine the link between the two countries. First, the creation of distinctive political institutions that recognised and symbolised Scotland’s nationhood might stimulate and encourage people in Scotland to feel Scottish rather than British. And if people’s sense of British national identity were to be eroded, then so also might be their wish to have that identity recognised through Scotland’s continued membership of the United Kingdom (Thatcher, 1998). Second, the establishment of a separate forum for political debate and decision-making in Scotland would seem to make it more likely that public opinion in the country would diverge from that elsewhere in the United Kingdom. After all, people in Scotland would debate policy issues amongst themselves without being influenced any more by those living in England, while the government in Scotland could implement and promote policies that were very different from those being pursued in England. Once people in Scotland realised they had a different set of priorities from those living in England, they would increasingly chafe at the remaining restrictions that membership of the Union imposed on their ability to decide policy for themselves (Curtice, 2005). Advocates of devolution had, of course, a riposte to these arguments. They suggested that, in the absence of devolution, support for the Union was being undermined because of its failure to recognise the distinctive public opinion that already existed in Scotland. The ‘Thatcherite’ policies pursued by Conservative governments from 1979 onwards, governments that never secured a mandate north of the border, offended the country’s alleged social democratic sensibilities (Keating, 2005a; McCrone, 1998). Devolution would enable Scotland to avoid having right-wing policies thrust upon it by a government that only wielded power as a result of winning English votes. Instead, public policy could reflect the distinctive  

17

Revolution or Evolution? needs and wishes of people in Scotland, and thereby reduce the dangers of frustration with the Union (Bednar, 2005; Curtice and Seyd, 2009; De Vries, 2000; Paterson, 1998). Meanwhile by demonstrating that the British state could make provision for specifically Scottish institutions, people in Scotland would appreciate that their distinctive Scottish identity could co-exist with the wider British identity that they shared with the United Kingdom as a whole (Aughey, 2001: chapter 7; Bogdanor, 1999; Mackintosh, 1998). Here the domestic argument touched upon a longstanding debate about whether dividing power between state and sub-state governments makes it easier or more difficult to maintain the unity of multi-national states (Friedrich, 1968; Stepan, 1999). On the face of it, the success of the SNP in the 2007 Scottish election would seem to cast doubt on the realism of the claims made by the advocates of devolution. Surely it must signify that frustration with the Union had finally boiled over? Perhaps it is testament to the fact that British national identity has indeed withered on the vine since the creation of the Scottish Parliament, choked by an ever-growing sense of Scottish identity that expresses itself in support for an avowedly nationalist party. Or perhaps public opinion in Scotland has indeed increasingly diverged from that in England, and as a result voters wanted to eject from office the party that still held the reins of power at Westminster because it was regarded as too unwilling to implement policies different from those being pursued by the UK government? So in this chapter we will consider whether Scotland has indeed grown apart from the rest of the United Kingdom since 1999, either gradually during that time or maybe suddenly in 2007. Do people in Scotland no longer feel British? Do they no longer share the same values as their neighbours south of the border? And do they now favour policies that are different from those that find favour in England? A DISTINCTIVE SENSE OF IDENTITY? Our first task is to examine how people in Scotland think of themselves in terms of national identity. Has their sense of feeling Scottish come to displace their willingness to embrace a British identity? Table 2.1 provides an initial answer by examining what people say when asked which of those identities associated with one or more parts of Great Britain or Ireland best describe how they think of themselves (see the Appendix to this book 18

A Distinctive Scottish Voice? for further details). Few opt for an identity other than British or Scottish, so we only show the proportion choosing these. However, as people could choose both of these identities, we also show how many did that. Table 2.1 Trends in ‘Free Choice’ National Identity, 1997–2007

Scottish British Both Sample size

1997 % 82 52 38 882

1999 % 84 47 35 1,482

2000 % 87 52 43 1,663

2001 % 86 50 41 1,605

2002 % 83 55 43 1,665

Scottish British Both Sample size

2003 % 84 58 47 1,508

2004 % 83 52 38 1,637

2005 % 85 52 43 1,549

2006 % 84 43 33 1,594

2007 % 80 47 33 1,508

Source: 1997: Scottish Election Study; 1999–2007: Scottish Social Attitudes.

Two points immediately stand out. First, the vast majority of people in Scotland – over four in five – already made claim to being Scottish in 1997, before devolution was in place. In contrast only around half said that they were British. To that extent the creation of the Scottish Parliament appears to have been a recognition of a distinctive sense of national identity that already existed north of the border. Second, remarkably little seems to have changed since the Scottish Parliament has been in place. Over the last decade the proportion saying they are Scottish has consistently been 80% or more, while the proportion feeling British has hovered around the 50% mark. Perhaps the proportion saying they were Scottish was already so high it could hardly go any higher, but even so it is notable there is little apparent sign here of any marked decline in the still widespread sense of British identity in Scotland. We should note too that many people appear to be happy to acknowledge both identities. True, the proportion of all people in Scotland volunteering both identities has fluctuated somewhat – between a third and only a little short of a half. Nevertheless, the perception that it is possible to be both is apparently so widespread that regularly many 19

Revolution or Evolution? readily acknowledge both identities without being prompted to do so. That might be thought to suggest that any recognition of people’s distinctive sense of Scottish identity need not necessarily be at the expense of their willingness to acknowledge a British identity. Still, to volunteer that you feel both Scottish and British is not necessarily to imply you feel both with equal intensity. Perhaps people’s sense of being British is very much secondary in intensity to their attachment to being Scottish. Indeed, in saying they are British many people could simply be acknowledging their legal citizenship rather than revealing an affective attachment. Consequently in Table 2.2 we show what the pattern of national identity looks like after those who lay claim to more than one identity are asked to choose just one. One incidental advantage of this approach is that information on the pattern of responses to such a ‘forced choice’ question on identity is available over a much longer period of time than just the last decade. Table 2.2 Trends in ‘Forced Choice’ National Identity, 1974–2007

Scottish British Sample size

1974 % 65 31 588

1979 % 56 38 658

1992 % 72 25 957

1997 % 72 20 882

1999 % 77 17 1,482

2000 % 80 13 1,663

Scottish British Sample size

2002 % 75 18 1,665

2003 % 72 20 1,508

2004 % 75 19 1,637

2005 % 79 14 1,549

2006 % 78 14 1,594

2007 % 72 19 1,508

2001 % 77 16 1,605

Source: 1974–97: Scottish Election Study; 1999–2007: Scottish Social Attitudes.

Our suspicion that when forced to choose people are more likely to prioritise their sense of feeling Scottish rather than any British identity they may also have is confirmed. Whereas during the course of the last decade the proportion saying that British is one of the identities they acknowledge has hovered at around a half, the proportion saying that British is the one single identity that best describes themselves had never exceeded one fifth. In contrast the proportion choosing Scottish as the one identity that best describes themselves has consistently been around three20

A Distinctive Scottish Voice? quarters, only a little below the 80% or so who say they are Scottish at all. Nevertheless, this alternative approach to the measurement of national identity still does not suggest that the advent of devolution has had much impact on the pattern of national identity. The proportion saying in 2007 that they are Scottish is exactly the same as it was in 1997, while there is virtually no difference in the proportion claiming to be British. Relatively few people in Scotland may feel British above all else, but this state of affairs seems to predate devolution rather than have been occasioned by it. Not that the position has always been thus during the last decade. As the table reveals, in 1979, shortly after Scotland failed to vote for devolution in sufficient numbers for it to be implemented, well over a third said that they were British, while almost a third had done so five years before that. This suggests the claim made by devolution’s advocates that failure to recognise Scotland’s distinctiveness from 1979 onwards served to undermine its attachment to the rest of the United Kingdom has some validity. Certainly insofar as the demand for devolution was linked to a distinctive sense of identity the pressure for change was evidently considerably greater by the time of the second devolution referendum in 1997. Both the primacy of Scottish identity over British identity, but also the fact that many people in Scotland acknowledge both, is confirmed by another approach to asking people about their identity. Simply asking people whether they are Scottish or British, both or neither, is a very crude approach to ascertaining national identity when people may acknowledge more than one identity, but to varying degrees. Better, perhaps, to ask people to choose which amongst a graduated set of descriptions best describes for themselves the relative balance of the two identities. The results of this so-called ‘Moreno question’ approach to asking about national identity are shown in Table 2.3 (Moreno, 1988; for further details see the Appendix to this book).

21

Revolution or Evolution? Table 2.3 Trends in Moreno National Identity, 1992–2007 1992 1997 % %

Scottish not British More Scottish than British Equally Scottish and British More British than Scottish British not Scottish Sample size

1999 %

2000 %

2001 %

2003 %

2005 %

2006 %

2007 %

19

23

32

37

36

31

32

33

27

40

38

35

31

30

34

32

32

30

33

27

22

21

24

22

22

21

28

3

4

3

3

3

4

4

4

5

3

4

4

4

3

4

5

5

6

957

882 1,482 1,663 1,605 1,508 1,549 1,594 1,508

Source: 1992–7: Scottish Election Study; 1999–2007: Scottish Social Attitudes.

When asked about their identity in this way, a majority of people in Scotland consistently choose some combination of Scottish and British, that is one of the three middle categories in Table 2.3. Since 1999 at least, the proportion doing so has hovered at on or just below 60%, though before then it seems to have been somewhat higher. There is no sign of any decline in adherence to some form of dual national identity since devolution has been in place.1 But at the same time those who say they are ‘more Scottish than British’ have always vastly outnumbered – sometimes by as much as ten to one – those who say they are ‘more British than Scottish’. Indeed, they are even consistently more numerous than those stating they are equally Scottish and British. People in Scotland may feel both identities, but typically it is their Scottish identity to which they are more strongly attached. 22

A Distinctive Scottish Voice? At the same time we can see that around a third or so of people in Scotland do explicitly reject the idea that they are British, whereas only around one in twenty explicitly reject being Scottish. But once again, it is not clear that much of the responsibility for this position can be laid at the door of devolution. True, the proportion saying they are Scottish and not British has been consistently higher since 1999 than it was before. On the other hand, there is not any evidence of a secular trend thereafter. Indeed the proportion saying they were Scottish and not British was lower in 2007 than at any time since 1997. The most that can be argued is that the initial creation of the devolved institutions encouraged some people explicitly to reject the idea that they were British, but that it has not had any discernible impact thereafter. Certainly, there is no reason to believe that the outcome of the 2007 election was the product of any marked new decline in adherence to Britishness. Still, we should perhaps consider a couple of possible caveats to the picture we have drawn so far. The first is that we might want to bear in mind that not everyone who lives in Scotland was born in Scotland. Perhaps the overall figures mask the fact that a sense of British identity has declined during the course of the last decade amongst those who were born in Scotland, but that this has been counteracted by either an opposite trend amongst the remainder of the population, or perhaps an influx of people newly resident in Scotland. Of this, however, there is little sign.2 True, those who were born in Scotland are less likely to say they are British; for example, in 2007 only 15% did so on our ‘forced choice’ measure. However, this is nothing new; in 1999 too, only 13% of those born in Scotland said they were British. Moreover, the twopoint difference between the two dates in the proportion saying they were British matches exactly the difference in that proportion between 1999 and 2007 amongst the population as a whole. It could be, of course, that while the broad distributions remain roughly the same over time, those who say they are Scottish not British have increasingly adopted a more ‘exclusive’ sense of national identity, whereby they are unwilling to accept as fellow Scots those who cannot claim either Scottish parentage or birth (Bryant, 1997; Kiely et al., 2005). There is certainly some evidence that those who regard themselves as Scottish and not British may be more likely to adopt an exclusive attitude towards those who might be considered ‘Other’. For example, amongst those who in 2006 said they were Scottish not British, 27% 23

Revolution or Evolution? indicated that they would be unhappy if a close relative of theirs were to marry someone who was Muslim. This compares with 23% of those who said they were equally Scottish and British and just 13% of the small number who said they were British not Scottish. More starkly, 61% of those who said they were Scottish not British agreed that Scotland would begin to lose its identity if more Muslims were to come to live in the country. In contrast, just 43% of those who stated they were equally Scottish and British and 39% of those who claimed they were only British adopted that view (see also Bromley and Curtice, 2003: 34; Bromley et al., 2007). However, it is not clear that the tendency of those who say they are Scottish and not British to be somewhat less likely on occasion to adopt an inclusive approach has been fostered by devolution.3 Certainly when the same questions about attitudes towards Muslims were previously asked in 2003, much the same pattern existed then too. For example, 24% of those who said they were Scottish not British said they would be unhappy if a close relative married a Muslim, compared with 20% of those who said they were equally Scottish and British and 15% of those who claimed only to be British. Equally, 44% of those who said they were only Scottish agreed that Scotland’s identity could be threatened by the presence of more Muslims, compared with 36% of those who adhered equally to both identities and 29% of those who said they were only British. However, we should bear in mind that the time span between 2003 and 2006, which is the only period for which we have the necessary data, is a relatively short one and does not encompass the whole of the period since the Scottish Parliament was created. Still, that possible uncertainty aside, there is little evidence that devolution has boosted Scottish identity, whether of an exclusive character or otherwise, at the expense of British identity. True, fewer people feel British than feel Scottish, while even amongst those who do acknowledge a British identity, it often sits alongside and has a lower priority than a concomitant Scottish identity. But for the most part this was just as true immediately before the advent of devolution at it was when the SNP came to power in 2007. On the other hand, there is reason to believe that during the period between the two devolution referendums, when attempts to create devolved institutions were opposed by the UK government, there was a marked decline in British identity north of the border. 24

A Distinctive Scottish Voice? DIFFERENT VALUES? The second key question we address is whether people in Scotland have increasingly come to uphold a different set of values from their counterparts in England. By values we mean normative objectives that people might support or oppose, such as equality or freedom (Rokeach, 1973). We do not mean the attitudes that people might adopt towards more specific policy proposals, which, while they may be influenced by their values, may also reflect their perceptions as to the impact a policy would have on the achievement of a particular normative objective. Of particular interest to us here is whether or not people in Scotland have come increasingly to value more equality as compared with their counterparts in England, as arguably it is equality that is the value that underlies support for the more social democratic outlook on the world to which it has been argued people in Scotland are more sympathetic (see also Curtice, 2005; Rosie and Bond, 2005). To assess support for greater equality we can look at the pattern of answers to five questions that between them are designed to tap people’s views about the merits of greater equality and of further government action to secure that objective (Evans et al., 1996; Heath et al., 1994). Each question has been asked regularly over the last decade in England as well as in Scotland, and therefore between them these questions enable us to ascertain whether or not the two countries have grown apart in their support for equality since the advent of devolution. The detailed wording of each question is shown in Table 2.4. Note that each one is worded in a pro-equality or more ‘left-wing’ direction.

25

Revolution or Evolution? Table 2.4 Indicators of Left–Right Values in Scotland and England, 2000–7 % agree 2000 2002 2004 2005 2006 2007 There is one law for the rich and one for the poor Scotland 67 61 53 49 52 57 England 64 61 51 53 55 52 Scotland– -3 0 +2 -4 -3 +5 England Ordinary working people do not get their fair share of the nation’s wealth Scotland 71 64 63 56 56 62 England 61 61 53 55 55 58 Scotland– +10 +3 +10 +1 +1 +4 England Management will always try to get the better of employees if it gets the chance Scotland 62 59 56 53 52 59 England 61 60 51 54 52 52 Scotland– +1 -1 +5 -1 0 +7 England Big business benefits owners at the expense of workers Scotland 61 57 56 50 50 53 England 55 56 47 52 51 49 Scotland– +6 +1 +9 -2 -1 +4 England Government should redistribute income from the better off to those who are less well off Scotland 50 45 40 41 39 37 England 38 37 31 33 34 32 Scotland– +12 +8 +9 +8 +5 +5 England Sample size Scotland 1,506 1,507 1,514 1,409 1,312 1,312 England 2,515 2,419 2,185 3,005 3,195 3,057 Source: Scottish Social Attitudes, 2000–7; British Social Attitudes, 2000–7.

26

A Distinctive Scottish Voice? We can see why it is sometimes claimed that people in Scotland are more social democratic or ‘left-wing’ than their counterparts in England. In the first year for which we have data, 2000, people in Scotland were consistently more likely to agree with our propositions than were the inhabitants of England. This was particularly true of the statements that ‘ordinary working people do not get their fair share of the nation’s wealth’ and that ‘government should redistribute income from the better off to those who are less well off’. Equally, however, the differences between Scotland and England should not be exaggerated. If Scotland could legitimately be described as being predominantly a social democratic country at the end of the twentieth century, then England qualified for that status too (see also Brown et al., 1996; Brown et al., 1999; Curtice, 1996; Miller et al., 1996 for similar analyses and conclusions for the period immediately before 1999). On each of the four propositions with which over half of people in Scotland were in agreement, so also were well over half of those in England. Moreover, there is little sign that attitudes in Scotland have subsequently diverged further from those in England. On average, in 2000, people in Scotland were six points more likely than were those in England to agree with one of our propositions. In 2007 the equivalent figure was five points. This, however, is not to say that support for equality has been unwavering since the advent of devolution. Rather, in every case the proportion of people in Scotland who agree with the proposition is lower now than it was in 2000. This shift somewhat away from left-wing values, however, has occurred in tandem with a similar move amongst people in England. That the two countries should experience much the same change in the climate of public opinion over the same time period suggests that, despite devolution, the two publics can still be moved by much the same forces. Nevertheless, although the differences in public opinion between Scotland and England may be small, a further analysis demonstrates why support for social democratic values might be regarded as distinctively Scottish. In Table 2.5 below we show for each Moreno national identity category respondents’ average score across all five of our left–right items in both 2000 and 2007. This figure is calculated by, first, giving a respondent a score of 1 where they strongly agreed with an item, 2 where they simply agreed, through to 5 where they strongly disagreed, and then 27

Revolution or Evolution? adding up this score across all five items and dividing the resultant total by 5. This means that the lower the score the more left-wing and proequality someone is. Table 2.5 Mean Left–Right Score by Moreno National Identity, 2000 and 2007 2000 Scottish not British More Scottish than British Equally Scottish and British More British than Scottish British not Scottish All

2.3 2.6 2.6 2.9 2.7 2.5

Sample size 552 453 317 50 59 1,506

2007 2.5 2.6 2.8 3.3 2.9 2.7

Sample size 327 409 357 71 81 1,312

Source: Scottish Social Attitudes, 2000, 2007.

This exercise reveals that those who say they are Scottish and not British consistently have the lowest score, and thus are the most left-wing. Limited though the differences between public opinion north and south of the border may be, it seems that strong adherence to a Scottish national identity is particularly intertwined with a more social democratic outlook on the world (Rosie and Bond, 2005). Not that those who say they are Scottish and not British have been immune to the drift away from leftwing values in recent years; however, as this shift occurred to more or less the same extent across all groups, the relatively left-wing character of those who say they are Scottish and not British has not been disturbed. But equally, of course, neither is it anything new. A second way in which it is sometimes thought that the values upheld by people in Scotland are distinctive is that they are more likely to adopt a socially conservative outlook than their counterparts in England. For example, writing at the time when there was significant opposition to a proposal to repeal a law that made it illegal for local authorities to ‘promote’ homosexuality, Tim Luckhurst, a former editor of The Scotsman, argued that, ‘Scotland remains a backward, repressed and socially conservative country’ (Luckhurst, 2001). Previous analyses have cast doubt on the validity of this proposition (Curtice, 2005; Park, 2002), but perhaps we should assess it further with more recent data. We can do so by looking at the pattern of responses to six questions that between 28

A Distinctive Scottish Voice? them are designed to tap the degree to which people value personal freedom in social and moral matters as opposed to valuing collective adherence to a set of common moral standards. The detailed wording of the items that form this ‘libertarian–authoritarian’ scale is shown in Table 2.6. In each case the question is worded in the authoritarian or conservative direction. We can see that in practice there are only small differences between the two countries in the proportion that agree with each proposition – and such differences as do exist do not support the contention that Scotland is the more conservative society.4 Indeed, if anything, people in Scotland are a little more likely to be liberal in their values than people in England. But essentially, as previous research has shown, the values of the two countries as measured by this scale are still largely similar to each other. Table 2.6 Indicators of Liberal–Authoritarian Values in England and Scotland, 2007 % agree

Scotland England

Young people today don’t have enough respect for traditional British values People who break the law should be given stiffer sentences For some crimes, the death penalty is the most appropriate sentence Schools should teach children to obey authority The law should always be obeyed, even if a particular law is wrong Censorship of films and magazines is necessary to uphold moral standards Sample size

Scotland– England

71

77

-6

83

79

+4

53

55

-2

79

84

-5

40

42

-2

59

62

-3

1,312

3,057

Source: Scottish Social Attitudes, 2007; British Social Attitudes, 2007.

So it seems that there is no consistent evidence that since the advent of devolution the values upheld by people in Scotland have diverged from those supported by people in England. Scotland has always been a little more social democratic or left-wing than England, but that is no more the case now than it was a decade ago. Meanwhile, the two countries share much the same 29

Revolution or Evolution? perspective on social and moral matters. But can we still say the same thing when we look at attitudes towards more specific aspects of public policy? ATTITUDINAL AUTONOMY? After all, devolution has certainly resulted in rather different public policies being implemented on the two sides of the border. This divergence has been particularly evident in the delivery of public services, many of which such as the health service, schools and universities, are now the responsibility of the devolved institutions. Two sets of differences have been apparent in particular. First, Scottish governments have been less keen than their UK counterparts on charging people for the use of public services. Thus upfront university tuition fees, first introduced by the UK Labour government in 1998, were initially converted into a smaller charge that was only paid after people had graduated and then subsequently scrapped entirely by the new SNP administration in 2007. In 2002 the then Scottish Executive introduced so-called free personal care for older people, in line with the recommendations of the Sutherland Commission (Sutherland, 1999), recommendations that were rejected by the UK government. Indeed these two policies, the abolition of up-front tuition fees and the introduction of free personal care, are often regarded as iconic examples of how devolution has resulted in differences of policy between Scotland and England (Keating, 2005b; Morris, 2007; Schmuecker and Adams, 2005: 34). Meanwhile in 2007 the SNP government announced the phased withdrawal of prescription charges, a policy that had already been implemented in Wales, but not in England. Second, Scottish governments have been less willing to use private contractors to help provide public services. This is most apparent in the management of the health service. As of 2005 in England the Department of Health expected Primary Care Trusts, the bodies principally responsible for commissioning health services, to spend 10% of their budget on buying services from private providers, rising to 15% by 2008 (Greer, 2005: 508; Talbot-Smith and Pollock, 2006: 69). In Scotland as of 2005 only £10m of an £8bn budget was expected to be spent in this way (Scottish Executive, 2005). Meanwhile, in 2007 the new SNP Health Secretary, Nicola Sturgeon, declared within just a few weeks of attaining 30

A Distinctive Scottish Voice? office that, ‘We reject the very idea that markets in health care are the route to improvement’, thereby signalling opposition in principle to the contracting-out of health service provision to private providers (Sturgeon, 2007; see also Scottish Government, 2007). But are these different policies an indication of divergent public attitudes towards the provision of public services? Table 2.7 looks at attitudes on both sides of the border in recent years towards the payment of university tuition fees. We should note, though, that the question that was asked (in both England and Scotland) in 2007 was somewhat different from what was asked between 2000 and 2003; in 2007 people were asked about their attitude towards the payment of tuition fees either while someone was studying or afterwards, whereas previously they were asked specifically about the payment of fees while someone was studying.5 We thus should exercise a little caution in talking about any trends between 2003 and 2007. Table 2.7 Trends in Attitudes towards University Tuition Fees, Scotland and England, 2000–7 % by column All should pay Some should pay, depending on income None should pay Sample size

All should pay Some should pay, depending on income None should pay Sample size

2000 Scotland England 5 8

2001 Scotland England 4 7

56

61

63

57

38 1,663

30 959

31 1,605

33 935

2003 Scotland England 5 8

2007 Scotland England 6 8

63

62

63

66

29 1,508

28 3,709

30 1,508

25 2,617

Note: Between 2000 and 2003 the question referred specifically to the payment of tuition costs while someone was studying. In 2007 it referred to either while they are studying or afterwards. Source: Scottish Social Attitudes, 2000–7; British Social Attitudes, 2000–7.

31

Revolution or Evolution? An interesting picture emerges (see also Paterson, 2002). Despite the iconic status of Scotland’s policy towards tuition fees and the complete abolition of such fees by the SNP government, it seems that only a minority of people in Scotland believe that no students at all should have to contribute towards the costs of their tuition. Moreover, it is not even the case that the size of that minority has consistently been larger in Scotland than it has been in England. Rather, in each and every year attitudes in Scotland have been largely much the same as they are in England. On this subject at least, devolution has opened up a policy gap where no significant attitudinal difference exists. The position is not, however, the same in respect of free personal care for older people (see also Ormston et al., 2007). As Table 2.8 shows, in Scotland over half support such a policy, whereas in England less than half do so. But even so, the difference of opinion is modest rather than dramatic. People in Scotland are some fifteen points or so more likely than their counterparts in England to feel that the government should pay for the cost of someone’s personal care irrespective of that person’s income or wealth. Still at least the difference in public policy does reflect a difference in majority public opinion.6 Table 2.8 Attitudes towards Paying for Personal Care, Scotland and England , 2005 and 2007 % by column Government should pay irrespective of person’s money Person should pay irrespective of money Who pays should depend on person’s money Sample size

2005

2007

Scotland

England

Scotland

England

57

42

55

41

2

3

1

4

41

54

43

54

1,549

2,721

1,508

1,735

Source: Scottish Social Attitudes, 2005, 2007; British Social Attitudes, 2005, 2007.

Much the same can be said about attitudes towards prescription charges. In Scotland the balance of opinion appears to be in favour of their abolition whereas in England the opposite seems to be the case. Thus in 2007, 46% of people in Scotland agreed with the proposition that ‘nobody 32

A Distinctive Scottish Voice? should have to pay prescription charges for medicine they need, even if they can afford to do so’, while 41% disagreed. The equivalent figures for England are 36% and 49% respectively. But again the difference of opinion is modest rather than dramatic; people in Scotland are around ten points more likely to favour the abolition of prescription charges, while the balance of opinion is only just in favour.7 Still we might note another reason why free personal care and the abolition of prescription charges might be regarded as expressions of a distinctively Scottish outlook; both policies are particularly popular amongst those who say they are Scottish not British. As many as 62% of that group say that the government should pay for personal care irrespective of someone’s financial resources, compared with just 52% of those who feel equally Scottish and British and 49% of those who only feel British. Equally, 54% of those who feel exclusively Scottish believe that nobody should have to pay prescription charges, compared with 40% of those who feel equally British and Scottish and 39% of those who state they are either primarily or only British. In contrast, but in tandem with the absence of much difference of outlook with England, there is no consistent link between national identity and attitudes towards tuition fees. Table 2.9 Attitudes in Scotland and England towards Private Companies Providing Public Services, 2007 Private companies running state schools

Scotland England

Private companies running NHS hospitals

Private companies providing personal care

Support

Oppose

Support

Oppose

Support

Oppose

% 12 20

% 65 53

% 17 23

% 67 55

% 28 31

% 51 43

Sample size

1,508 1,735

Source: Scottish Social Attitudes, 2007: British Social Attitudes, 2007.

But what of the second major difference of public policy that has arisen between England and Scotland in the wake of devolution, that is, in the degree to which public services are delivered by private sector organisations? Table 2.9 gives an indication of public attitudes towards 33

Revolution or Evolution? private companies running or providing state schools, NHS hospitals and personal care. It suggests that the Scottish public are indeed wary about all three, albeit somewhat less so in the case of personal care where the use of private providers is not uncommon. But this wariness is also to be found amongst people in England. The most that can be said is that people in Scotland are even more inclined to be wary than their counterparts in England. So the differences that have emerged between Scotland and England since the advent of devolution in how public services are delivered are in part a reflection of differences in public attitudes – but only in part. There is majority support for free personal care and the abolition of prescription charges in Scotland, while this is not the case in England. On the other hand, it is not clear that there is much difference between the two countries in their attitude towards the payment of university tuition fees, while Scotland is simply somewhat more doubtful than England about having private companies deliver public services. Moreover, even where there are differences of view, it is not clear that they had grown any bigger by the time of the 2007 election. We should not, however, presume that any differences of attitude between Scotland and England are confined to those matters where the devolution settlement does make it possible for public policy to be different in Scotland from that in England. After all, where devolution has made it possible for Scotland to pursue a different policy that is more in tune with a distinctive public opinion north of the border, this should have increased satisfaction with the existing devolution settlement rather than engendered discontent. Perhaps one of the frustrations felt by people in Scotland – and thus perhaps one reason for the appeal of the SNP in 2007 – is that the devolved parliament is unable to reflect what is a distinctive public opinion north of the border on some of those matters that are still reserved to the UK government and parliament, such as many financial matters. Table 2.10 considers this possibility by showing recent trends on both sides of the border in attitudes towards the balance of taxation and spending, the distribution of income, and unemployment benefit. We might note that each of these is often regarded as a ‘left–right’ issue and thus as ones where we might anticipate that people in Scotland would have distinctive views from their neighbours in England should they indeed be inclined towards a social democratic outlook. 34

A Distinctive Scottish Voice? Table 2.10 Trends in Attitudes towards Economic Issues, Scotland and England, 1999–2006 % agree 1999 2000 2001 2002 2003 2004 2006 Benefits for unemployed people are too low and cause hardship Scotland 36 43 41 41 45 na 33 England 32 40 34 28 36 na 22 Scotland– +4 +3 +7 +13 +9 +11 England Increase taxes and spend more on health, education and social benefits Scotland 55 54 63 60 58 56 41 England 58 51 60 61 51 49 46 Scotland– -3 +3 +3 -1 +7 +7 -5 England Gap between those with high incomes and those with low incomes is too large Scotland 84 85 85 na na 82 78 England 81 82 79 na na 72 75 Scotland– +3 +3 +6 +10 +3 England Sample size Scotland 1,482 1,663 1,605 1,665 1,508 1,637 1,549 973– 942– 1,849– 1,806– England 1,798– 1,932– 2,761 2,718 2,887 2,897 2,734 2,721 2,775 na: not asked In most years sample sizes in England vary between questions; figures shown are the minimum and maximum sizes. Source: Scottish Social Attitudes, 1999–2006; British Social Attitudes, 1999–2006.

To some extent people in Scotland do have different views from people in England. People in Scotland have certainly been consistently more likely in recent years to be sympathetic to the view that unemployment benefit is too low. Of course unemployment has historically been higher north of the border. However, by 2006 this was no longer the case (Office for National Statistics, 2008: Table 5.14), yet even so the difference of attitude had, if anything, widened. Nevertheless, on our other items the differences of outlook are relatively small and 35

Revolution or Evolution? certainly have not widened over time. People in Scotland are a little more likely than those in England to believe that the gap in incomes is too large. Differences in the level of support for more public spending, even at the expense of higher taxation, are not only small but are not even always in the same direction. For the most part it seems unlikely that frustration at the Scottish Parliament’s inability to tread a different policy path in these areas gave rise to an electoral revolution in 2007. CONCLUSION The creation of distinctively Scottish political institutions does not seem to have undermined adherence north of the border to a British national identity. True, the strength of attachment to British identity generally looks weaker than that to Scottish identity, but this was no more the case in 2007 than it was in 1999. Equally, the creation of a separate process for determining much public policy in Scotland does not appear to have caused the values upheld by people in Scotland to diverge further from those supported by people in England. Indeed people in Scotland – including those who feel only Scottish – seem to have shared in the marginal drift away from the left that England has also experienced over the last decade. In short, there seems to be no reason to believe that in these respects at least the Union was under significantly greater strain in 2007 than it had been previously. Meanwhile, devolution appears to have had some success in ensuring that public policy reflects public attitudes north of the border towards specific policy issues. This can be said of the provision of free personal care (where opinion is rather different from that in England), prescription charges (where the difference of opinion is less marked) and the use of private providers in the health service (where the policy pursued by the Scottish government seems to have been more in tune with opinion on both sides of the border). On the other hand it is not clear that the abolition of university tuition fees is in tune with Scottish public opinion, which appears to support such fees in much the same way as opinion does in England. Still, and just as importantly, there is not much evidence that attitudes in Scotland towards those matters that are still decided by the UK government have become markedly at variance with public opinion in England. There is thus no particular reason why these issues should have 36

A Distinctive Scottish Voice? become an increasing source of tension within the Union. So it is not clear so far as the identities, values and attitudes of people in Scotland are concerned that the electoral success of the SNP in 2007 was a sign of renewed pressure on the Union or a wish for a revolutionary break with the past. Devolution even seems to have had some success in reflecting those distinctive public policy preferences that do exist north of the border. But of course this does not necessarily mean that devolution was thought by the public at the time of the 2007 election to have delivered on its objectives. We thus turn in the next chapter to public perceptions of devolution’s successes and failures. NOTES 1. Indeed, intriguingly, the proportion choosing one of the three middle categories is, at 63%, slightly higher in 2007 than at any time since 1997, though we should bear in mind that this could simply be the result of random sampling variation. 2. As many as 18% of the people in our 2007 survey said that they were born outside Scotland, including 10% who were born in England. However, the overall figure is little different from the equivalent of 17% in 2003 though it is a little higher than the 14% recorded in 1999. The 2001 Census estimated the proportion born outside Scotland was 13% (see http://www.scrol.gov.uk/scrol/browser/profile.jsp). 3. Note also that if we exclude those who are immigrants to Scotland from England, much of this difference by national identity disappears (Hussain and Miller, 2006). 4. There is also no evidence that those who say they are Scottish and not British are particularly socially conservative. If we calculate the mean score across all six items on this scale in an analogous manner to that deployed in Table 2.5, we find that, at 3.9, the mean score for this group is little different from that for all respondents (3.8). 5. Between 2000 and 2003 respondents were also asked separately their views about paying tuition fees after someone has finished studying. People in both countries were somewhat more inclined to think that in these circumstances everyone should pay rather than just some, but, as in Table 2.7, there are only small and inconsistent differences of view between the two countries. 37

Revolution or Evolution? 6. Though of course we should bear in mind the possibility that public opinion has been influenced by public policy rather than vice-versa. 7. The difference of view on this subject is certainly far less than the equivalent difference between England and Wales, where prescription charges had already been abolished. No less than 62% of people in Wales agreed that nobody should have to pay prescription charges, while only 29% disagreed. REFERENCES Aughey, A. (2001), Nationalism, Devolution and the Challenge to the United Kingdom State, London: Pluto Press. Bednar, J. (2005), ‘Federalism as a Public Good’, Constitutional Political Economy, 16: 189–205. Bogdanor, V. (1999), Devolution in the United Kingdom, Oxford: Oxford University Press. Bromley, C. and Curtice, J. (2003), Attitudes to Discrimination in Scotland, Edinburgh: Scottish Executive Social Research. Bromley, C., Curtice, J. and Given, L. (2007), Attitudes to Discrimination in Scotland 2006, Edinburgh: Scottish Government Social Research. Brown, A., McCrone, D. and Paterson, L. (1996), Politics and Society in Scotland, Basingstoke: Macmillan. Brown, A., McCrone, D., Paterson, L. and Surridge, P. (1999), The Scottish Electorate: The 1997 election and beyond, Basingstoke: Macmillan. Bryant, C. (1997), ‘Citizenship, National Identity and the Accommodation of Difference: Reflections on the German, French, Dutch cases’, Journal of Ethnic and Migration Studies (formerly New Community), 23: 157–72. Curtice, J. (1996), ‘One Nation Again?’, in Jowell, R., Curtice, J., Park, A., Brook, L. and Thomson, K. (eds), British Social Attitudes: The 13th report, Aldershot: Dartmouth. Curtice, J. (2005), ‘Brought together or driven apart?’, in Miller, W. (ed.), Anglo-Scottish Relations from 1900 to Devolution and Beyond, Oxford : Oxford University Press for the British Academy. Curtice, J. and Seyd, B. (2009), ‘Introduction’, in Curtice, J. and Seyd, B. (eds), Has Devolution Worked?, Manchester: Manchester University Press. 38

A Distinctive Scottish Voice? De Vries, M. (2000), ‘The Rise and Fall of Decentralization: a comparative analysis of arguments and practices in European countries’, European Journal of Political Research, 38: 193–224. Evans, G., Heath, A. and Lalljee, M. (1996), ‘Measuring Left–Right and Libertarian–Authoritarian Values in the British Electorate, British Journal of Sociology, 47: 93–112. Friedrich, C. (1968), Trends of Federalism in Theory and Practice, London: Pall Mall Press. Greer, S (2005), ‘The Territorial Bases of Health Policymaking in the UK after Devolution’, Regional and Federal Studies, 15: 501–18. Heath, A., Evans, G. and Martin, J. (1994), ‘The Measurement of Core Beliefs and Values: the development of balanced socialist/laissez faire and libertarian/authoritarian scales’, British Journal of Political Science, 24: 115–31. Hussain, A. and Miller, W. (2006), ‘Islamophobia and Anglophobia in PostDevolution Scotland’, in Bromley, C., Curtice, J., McCrone, D. and Park, A. (eds), Has Devolution Delivered?, Edinburgh: Edinburgh University Press. Keating, M. (ed.) (2005a), Scottish Social Democracy: Progressive ideas for public policy, Oxford: Peter Lang. Keating, M. (2005b), The Government of Scotland: Public policy making after devolution, Edinburgh: Edinburgh University Press. Kiely, R., Bechhofer, F. and McCrone, D. (2005), ‘Birth, Blood and Belonging: Identity claims in post-devolution Scotland’, Sociological Review, 53: 150–71. Luckhurst, T. (2001), ‘Scotland Returns to the Dark Ages’, New Statesman, 21 May. McCrone, D. (1998), ‘Thatcherism in a Cold Climate’, reprinted in Paterson, L. (ed.), A Diverse Assembly: The debate on the Scottish Parliament, Edinburgh: Edinburgh University Press. Mackintosh, J. (1998), ‘A Parliament for Scotland’, reprinted in Paterson, L.(ed.), A Diverse Assembly: The debate on the Scottish Parliament, Edinburgh: Edinburgh University Press. Miller, W., Timpson, A. and Lessnoff, M. (1996), Political Culture in Contemporary Britain: People and politicians, principles and practice, Oxford: Oxford University Press. Moreno, L. (1988), ‘Scotland and Catalonia: The path to Home Rule’, in McCrone, D. and Brown, A. (eds), The Scottish Government Yearbook 1988, Edinburgh: Unit for the Study of Government in Scotland. 39

Revolution or Evolution? Morris, N. (2007), ‘The Big Question: Is Scotland becoming Separated from the Rest of the United Kingdom?’, The Independent, 15 June. Office for National Statistics (2008), Regional Trends 40, Basingstoke: Palgrave Macmillan. Ormston, R., Curtice, J. and Fawcett, H. (2007), ‘Who Should Pay for My Care – When I’m 64?’, in Park, A., Curtice, J., Thomson, K., Phillips, M. and Johnson, M. (eds), British Social Attitudes: The 23rd report, London: Sage, 2007. Park, A. (2002), ‘Scotland’s Morals’, in Curtice, J. McCrone, D. Park, A. and Paterson, L. (eds), New Scotland, New Society, Edinburgh: Edinburgh University Press. Paterson, L. (1998), ‘Scottish Home Rule: Radical Break or Pragmatic Adjustment?’, in Elcock, H. and Keating, M. (eds), Remaking the Union: Devolution and British politics in the 1990s, London: Frank Cass. Paterson, L. (2002), ‘Governing from the Centre: Ideology and Public policy’, in Curtice, J., McCrone, S., Park, A. and Paterson, L. (eds), New Scotland, New Society, Edinburgh: Edinburgh University Press. Rokeach, M. (1973), The Nature of Human Values, New York: Free Press. Rosie, M. and Bond, R. (2005), ‘Social Democratic Scotland?’, in Keating, M. (ed.), Scottish Social Democracy: Progressive ideas for public policy, Oxford: Peter Lang. Schmuecker, K. and Adams, J., (2005), ‘Divergence in Priorities, Perceived Policy Failure and Pressure for Convergence’, in Adams, J. and Schmuecker, K. (eds), Devolution in Practice 2006: Public policy differences within the UK, London: Institute for Public Policy Research. Scottish Executive (2005), Delivering for Health, Edinburgh: Scottish Executive. Scottish Government (2007), Better Health, Better Care: Action plan, Edinburgh: Scottish Government. Stepan, A. (1999), ‘Federalism and Democracy: beyond the US model’, Journal of Democracy, 10: 19–34. Sturgeon, N. (2007), ‘A Different Path to Reform’. Speech at NHS Confederation Annual Conference, reported at: http://www.scotland.gov.uk/News/Releases/2007/06/2108 3635. Sutherland, S. (chmn) (1999), With Respect to Old Age: Long-term care – rights and responsibilities, Cm 4192, London: Stationery Office. 40

A Distinctive Scottish Voice? Talbot-Smith, A. and Pollock, A. (2005), The New NHS: A guide, London: Routledge. Thatcher, M. (1998), ‘Don’t Wreck the Heritage we All Share’, reprinted in Paterson, L. (ed.), A Diverse Assembly: The debate on the Scottish Parliament, Edinburgh: Edinburgh University Press.

41

CHAPTER 3 __________________________________________

WHAT HAS DEVOLUTION ACHIEVED? THE PUBLIC’S VIEW According to its advocates, devolution was meant to improve the way that Scotland was governed. First, as we noted in the previous chapter, it was hoped that the creation of the Scottish Parliament would ensure that public policy would be better attuned to the needs and wishes of the people of Scotland. In part this was an argument about ensuring that the wishes of the people of Scotland could not be overridden by the votes cast by people in England. But it was also an argument about the functional benefits of devolving decisions away from London. Scotland’s economic growth had long tended to lag behind that of southern England, and the country had seemed to have been ill-served by centralised rule (Paterson, 1998). Now following globalisation, the ability of even large states like the UK to steer their economies was limited. In this new environment the main economic role of governments was to make their part of the world an attractive place for companies to locate, by providing a skilled workforce and a pleasant working environment – ‘supply side’ measures that were pursued more effectively at a regional rather than a statewide level (Keating and McGarry, 2001; Ohmae, 1995). As the then Secretary of State for Scotland, Donald Dewar, put it in his foreword to the UK government’s white paper on Scottish devolution, ‘the Scottish Parliament will be in a position to encourage vigorous sustainable growth in the Scottish economy. Policies on health, housing and education will respond more directly to Scotland’s needs’ (Scottish Office, 1997; see also Scottish Constitutional Convention, 1995). Second, devolution would improve the process of government in Scotland. As the then Minister of State in the Scottish Office, Henry McLeish, argued, ‘the establishment of the Scottish Parliament offers the opportunity to put in place a new sort of democracy in Scotland, closer to the Scottish people’ (Scottish Office, 1998). Several key elements of the way in which the Scottish Parliament and Executive operated were explicitly designed to encourage citizens to become involved and influential in decision making, with the aim of achieving a radical break 42

What has Devolution Achieved? from the procedures employed at Westminster. Anyone could submit a petition to a Petitions Committee of the Scottish Parliament that was obliged to consider the issues raised and empowered to pursue them with the relevant government department (Carman, 2006). The Scottish Executive was required to engage in considerable public consultation before proposing measures to Parliament. Meanwhile, Parliament itself was to begin the process of legislative scrutiny by taking evidence from outside organisations and individuals. These more open, participative processes would, it was hoped, help restore some of the trust and confidence in politics and politicians that had seemingly been lost in recent years (Curtice, 2002; Dalton et al., 2003; Dewar, 1998). For those advocates of devolution who were keen to preserve the Union, there was a further important benefit that they hoped would accrue. If devolution improved the perceived effectiveness of government in Scotland and engaged people more closely in the political process, then people’s satisfaction with the way in which the Union was operating should increase as well. That increased satisfaction should then in turn help to maintain and even strengthen public support for the Union. But equally, of course, if the hopes people invested in devolution were to be dashed then perhaps support for the Union would be eroded. Maybe the electoral success of the SNP in 2007 was the result of growing disenchantment with the way in which devolution, and thus the Union, was operating in practice, leaving the electorate thirsting for a more revolutionary break with the past? In this chapter, therefore, our main task is to consider how well or badly devolution is thought to have performed since its introduction in 1999, and whether we can detect evidence of heightened disappointment at the time of the 2007 election. We focus on two questions in particular. First, is devolution thought to have succeeded or failed in producing more effective policies? And, second, have people come to feel more closely involved in their government or not? Thereafter we examine whether there is indeed any evidence of growing discontent with the operation of the Union in practice. But we begin by examining whether at the outset the people of Scotland shared the hopes and aspirations for devolution that its advocates contended it would deliver.

43

Revolution or Evolution?

EXPECTATIONS There is considerable evidence that they did. Table 3.1 shows what impact it was thought the creation of the Scottish Parliament would have on, first, two of the policy areas for which it would acquire responsibility, the health service and education, and, second, how much say people would have in how they are governed (for data on a more extended set of expectations see Paterson et al., 2001; Seyd, 2009). In each case the first reading shown in the table was taken immediately after the referendum in September 1997 in which a majority voted for the creation of the Scottish Parliament, while the second observation was taken soon after the first elections to the new parliament in 1999. Table 3.1 Expectations of Devolution, 1997 and 1999 Expected impact of a Scottish Parliament on …

Better/ increase/ more %

… the standard of the health service in Scotland … the standard of education … giving ordinary people more say over how Scotland is run Expected impact of a Scottish Parliament on …

%

Worse/ reduce/ less %

65

28

6

71

19

3

79

16

2

Better/ increase/ more %

… the standard of the health service in Scotland … the standard of education … giving ordinary people more say over how Scotland is run

1997 No difference

1999 No difference %

Worse/ reduce/ less %

49

41

4

56

35

3

64

32

2

Sample size: 1997: 676; 1999: 1,482. Source: Scottish Referendum Study, 1997; Scottish Social Attitudes Survey, 1999.

44

What has Devolution Achieved? Hopes were high indeed when the public voted in favour of devolution in 1997. Nearly two-thirds (65%) thought having a Scottish Parliament would result in a better health service, while even more thought it would bring about higher standards of education (71%) and more say for ordinary people in how Scotland was run (79%). Even so, by the time the first election had been held in 1999 and the parliament had initially been established, there were signs that a somewhat more sanguine mood was setting in. Now only a little under half thought that having the devolved parliament would result in a better health service (49%), while less than two-thirds thought its establishment would improve the standard of education or give ordinary people more say in the running of their country. But even so, very few indeed expected devolution to have a negative impact on the governance of Scotland. So the parliament was created on a wave of widespread optimism about what it might achieve. Trouble is, that meant that it would be difficult indeed for devolution to live up to expectations, not least perhaps because it lacked the necessary powers to do so (Mitchell, 2004). There seemed to be plenty of scope for disappointment with the devolution project. IMPROVED EFFECTIVENESS? In Table 3.2 we show what impact the parliament was actually thought to be having on the health service and education once devolution was in place. Note, however, that until 2006 the survey question on the health service asked people what impact they thought having a parliament would have rather than what actual impact it was thought to have had. This formulation means that in this case respondents’ answers still mixed their hopes with their perceptions of what had actually been achieved. It seems that for many people devolution has indeed not lived up to the hopes they once had for it. Consistently only around a quarter have felt that having the Scottish Parliament was increasing the standard of education, far fewer than the 71% who at the time of the referendum thought it would have a positive impact. Meanwhile in the case of the health service, hopes that devolution would bring about an improvement diminished further in the early years of its implementation. Thereafter once survey respondents were asked what impact devolution had actually 45

Revolution or Evolution? had, only a quarter or so thought it had served to bring about improvement, much the same as in the case of education. Table 3.2 Perceived Impact of Devolution on the Health Service and Education, 2001–7 2001 % 45

2002 % na

42

na

9 2001 %

na 2002 %

2003 2006** % % 37 22

2007 % 26

Impact of Scottish Parliament on NHS

Better No difference Worse

Impact of Scottish Parliament on standard of education Sample size

Increasing

27

25

23

30

28

No difference

59

58

59

52

54

Reducing

5

6

7

6

3

1,605

1,665

1,508

1,594

1,508

46

62

59

10 9 2003 2006** % %

6 2007 %

na: not asked ** Up to and including 2003 people were asked whether the parliament ‘will’ make the NHS better, worse or make no difference. From 2006 onwards they were asked whether they thought the parliament ‘is’ making a difference. The question on education asked about whether the parliament ‘is’ making a difference throughout. Source: Scottish Social Attitudes, 2001–7.

Even so, we should also note two further points. First, very few people have ever said that devolution has made the health service or education any worse. Second, there was no sudden new rush of disappointment in 2007 that might have caused people to lose faith in devolution and consider the merits of independence instead. So it seems unlikely that the perceived ineffectiveness of devolution in delivering better policy outcomes could account for the SNP’s success in that year’s election.1

46

What has Devolution Achieved?

MORE ENGAGEMENT? Meanwhile, Table 3.3 shows what impact devolution is thought to have had on the degree to which ordinary people have a say in how they are governed. Once again, fewer people have felt that having the parliament has increased people’s involvement in the process than initially expected that it would. Typically somewhere between three and four in ten have felt that devolution has increased ordinary people’s say, far fewer than the nearly eight in ten who believed in 1997 that it would. Nevertheless that still means the parliament was more likely to have been considered successful at enhancing people’s involvement in the political process than it was at improving the effectiveness of public policy. Table 3.3 Perceived Impact of Scottish Parliament on Ordinary People’s Say in how Scotland is Governed, 2001–7

More No difference Less Sample size

2001 % 38

2002 % 31

2003 % 39

2004 % 31

2005 % 37

2006 % 37

2007 % 47

56

62

54

60

55

55

45

4 1,605

4 1,665

4 1,508

6 1,637

5 1,549

5 1,594

3 1,508

Source: Scottish Social Attitudes, 2001–7.

Moreover, there is no sign that disenchantment grew over time; the distribution of opinion in 2006 was almost exactly the same as it was in 2001. However, the balance of opinion was rather different in 2007. The proportion thinking that the parliament had given ordinary people more say in how they are governed increased by ten percentage points to just under half (47%), much higher than in any previous year. But, of course, far from suggesting growing disappointment with devolution at the time of the 2007 election, this result seems to suggest the very opposite was happening. Here we should remember, however, that our 2007 survey was conducted after the election, by which time the SNP had already come to power and had begun to make its initial pronouncements and decisions as 47

Revolution or Evolution? a government. Although the aim of the survey was to capture public opinion at the time of the election, it is possible that people’s answers to our survey were sometimes influenced by subsequent events. Maybe the advent of the new government persuaded some people that they did now have more say in how they were governed? If this is so, however, it suggests an interesting paradox. Perhaps the initial impact at least of having a pro-independence SNP government come to power was to help restore faith in the operation of the devolution settlement. Whatever may have been the position immediately before the 2007 election, maybe afterwards disappointment with what devolution was achieving may actually have fallen away somewhat (Curtice, 2008). This is an important point to which we shall return. Table 3.4 How Good is Scottish Executive/UK Government at Listening to People before Taking Decisions? 2004 2005 2006 2007 % % % % Very/quite Scottish 32 34 36 43 good Executive Not very / 58 54 53 44 not at all good Very/quite UK 15 16 18 21 good Government Not very / 80 77 76 71 not at all good Sample size 1,637 1,549 1,594 1,508 Source: Scottish Social Attitudes, 2004–7.

In the meantime we have a second measure available to us of how far people feel they have some involvement or influence in the decisionmaking process following the introduction of devolution. Since 2004 the Scottish Social Attitudes survey has been asking, ‘How good do you think the Scottish Executive is at listening to people’s views before taking decisions?’ The same question has then also been asked of the UK government. As Table 3.4 shows, typically only around a third or so have said the Executive were ‘very’ or ‘quite good’ at listening to decisions. However, this is around twice the proportion saying the same of the UK 48

What has Devolution Achieved? government. That might be taken to indicate that the devolved institutions are thought by some at least to have increased people’s influence on decision making. Moreover, there is evidence here too that evaluations of the Executive (and perhaps to some degree the UK government also) were particularly favourable in 2007. So once again there are signs that, by the time of the 2007 election was over at least, people were rather more likely to think they had some influence over those who governed them. PERCEPTIONS OF THE UNION For many people, then, devolution does not seem to have lived up to the high hopes that they once had for it. But few think it has done much harm – they simply wonder whether it has made much difference – while there is no evidence that people had become particularly disillusioned in the run up to the 2007 election. Indeed, so far as giving people more influence and involvement is concerned, the very opposite was true in the immediate wake of the election at least. But what impact, if any, has the experience of the first eight years of devolution had on perceptions of the functioning of the Union? Is there any reason to believe, as the more unionist advocates of devolution hoped would be the case, that it has persuaded people that the Union is now working more effectively so far as Scotland is concerned? Or has the very opposite happened, thereby providing some of the impetus that helped take the SNP to power? People in Scotland certainly anticipated that the creation of the Scottish Parliament would strengthen the way in which the Union operated from Scotland’s point of view. At the time of the 1997 referendum as many as 70% said that devolution would strengthen Scotland’s voice in the United Kingdom, and in this case expectations remained just as high immediately after the first devolved election in 1999. But once again this meant that there was plenty of room for disappointment.

49

Revolution or Evolution? Table 3.5 Perceived Impact of Devolution on Scotland’s Voice in the UK, 2001–7

Stronger No difference Weaker Sample size

2001 % 52

2002 % 39

2003 % 49

2004 % 35

2005 % 41

2006 % 43

2007 % 61

40

52

41

55

50

49

32

6

7

7

7

6

6

4

1,605

1,665

1,508

1,637

1,549

1,594

1,508

Source: Scottish Social Attitudes, 2001–7.

Indeed, here too typically fewer people have said that devolution has brought about a stronger voice for Scotland than initially hoped it would. Nevertheless, it has consistently been more likely to be considered successful at delivering a stronger voice than it has been at either improving policy outcomes (compare Table 3.2) or giving people a role in decision making (Table 3.3). In most years the proportion saying that devolution has given Scotland a stronger voice has been above the 40% mark. In other words a significant proportion of the Scottish public have felt that on this criterion at least devolution was living up to its promise. Meanwhile, once again there is no evidence that the 2007 election was graced by a particularly high level of disenchantment with devolution. Rather the opposite was the case; as many as 61% now said that devolution had strengthened Scotland’s voice in the UK. One possibility is that holding a Scottish election itself persuades people that devolution is having that effect, for the proportion thinking that devolution was strengthening Scotland’s voice also rose by ten points between 2002 and 2003. After all, when such an election is held political developments in Scotland receive a relatively high level of attention in the UK-wide media and senior British party leaders frequently come north of the border to campaign. Even so, the increase in the proportion saying that devolution strengthens Scotland’s voice was, at 18 points, even more marked between 2006 and 2007 than it was between 2002 and 2003. Once again it seems that the coming to power of the SNP was accompanied by increased satisfaction with Scotland’s position within the Union. 50

What has Devolution Achieved? Indeed, much the same seems to be true of perceptions of the economic benefits of the Union. As Table 3.6 shows, typically people in Scotland have been inclined to say that England’s economy benefits more than Scotland’s does from Scotland being part of the UK. True, the degree to which the Union is thought to be to Scotland’s disadvantage should not be exaggerated; typically as many people have said that the two economies benefit equally from the Union as have said that England’s benefits more. Nevertheless, a substantial minority of people in Scotland have consistently felt some sense of grievance about the economic consequences of the Union. Crucially, however, that minority has not grown in size. And in 2007 it was at its lowest level yet. Indeed, almost as many people now felt that Scotland’s economy benefited more from the Union than believed that England’s did. So once again it seems that in 2007 people were more likely to be satisfied with the way in which the Union was operating. Table 3.6 Perceptions of Whose Economy Benefits More from Scotland Being Part of the UK, 2000–7

England Scotland Both equal Sample size

1999 % 36 22 36 1,482

2000 % 42 16 36 1,663

2001 % 38 18 39 1,605

2003 % 30 24 40 1,508

2005 % 36 21 34 1,549

2007 % 27 25 39 1,508

Source: Scottish Social Attitudes, 1999–2007.

One of the biggest sources of dispute about the way in which the Union operates concerns Scotland’s share of government spending. There is little disagreement that, in so far as it is possible to assign government spending to one part of the UK rather than another, spending per head is higher in Scotland than in England (Scottish Government, 2008). What is contested is whether this reflects a higher level of need (McLean et al., 2008) or whether, once revenues from North Sea Oil are taken into account, the higher level of spending is matched by higher revenues (Allardyce, 2008). In any event, as Table 3.7 shows, many people in Scotland have inclined to the view that their country receives less than its fair share of public spending. Typically around a half endorse that view, whereas only 51

Revolution or Evolution? around one in ten feel that Scotland actually gets more than its fair share. But once again perceptions were different in 2007. By then only just over a third felt that Scotland received less than its fair share, no more than believed that Scotland secured more or less its fair share. So, once again, it seems that the installation of a nationalist regime in Edinburgh coincided with decreasing dissatisfaction with the Union. Table 3.7 Perceptions of Scotland’s Share of Government Spending Compared with Other Parts of the UK, 2000–7

More than fair share Pretty much fair share Less than fair share Don’t know/not answered Sample size

2000 %

2001 %

2003 %

2005 %

2007 %

10

10

11

10

16

27

36

34

32

37

59

48

47

49

36

4

6

7

8

11

1,663

1,605

1,508

1,549

1,508

Source: Scottish Social Attitudes, 2000–7.

How might we account for this unexpected trend? As we have already remarked, it may be a consequence of the SNP’s success rather than a mood that had already taken hold of the nation at the time that it voted. Perhaps the installation of a government that vowed to put Scotland’s interests first, and was not constrained (as the previous Labour-led administration had appeared to be) from airing its disagreements with the UK government in public (Laffin et al., 2007; Trench, 2005; 2007),2 began to convince people that devolution could be used to improve the way Scotland was governed within the framework of the Union after all. This is not what the SNP itself anticipated a spell of successful nationalist government would bring; rather. it hoped that by providing effective government it would persuade people that Scotland could and should govern itself.3 It seems, however, that there is no guarantee that this is a link the public necessarily makes.

52

What has Devolution Achieved?

CONCLUSION In some respects devolution has come to be regarded as a disappointment. Considerably fewer people feel that it has actually improved the way that Scotland is governed than initially hoped it would bring a benefit. However, that does not mean that people feel that devolution has had an adverse impact; very few do. They are simply inclined to say it has not made much difference. Nor is there any evidence that disappointment with devolution had grown in the period prior to the 2007 election. Rather, when it was over at least, the very opposite seems to have been the case. As a result, perhaps, there is also little evidence of growing disenchantment with the operation of the Union. Rather, by the time the 2007 election was over, the Union was regarded in a more favourable light than it had been at any time since the advent of devolution. In short there is little evidence here that the SNP success in the 2007 election signified a wish to instigate a revolutionary break with the past. But still we have as yet to look directly at the level of support for independence, and whether or not people were happy with the current devolution settlement. So we now turn to looking at how people believed Scotland should be governed. NOTES 1. Note that much the same conclusion might be drawn from the data presented in Table 5.2 below. Note also that none of the figures for 2007 in that table are markedly worse than the equivalent data for 2006 (Ormston, 2007: 12). 2. See also the further discussion of this difference in Chapter 4 below. 3. For example, when in May 2008 he was celebrating a year in office, Alex Salmond argued, ‘I firmly believe that as the SNP deliver good government in the devolved areas and secure the trust of the people, so we will build the case for Scotland to be governed equally well in all areas’ (Salmond, 2008).

53

Revolution or Evolution?

REFERENCES Allardyce, J. (2008), ‘Oil would make independent Scotland rich’, Sunday Times, 18 May. Carman, C. (2006), Assessment of the Scottish Parliament's Public Petitions System, 1999–2006, Edinburgh: Scottish Parliament Corporate Body. Curtice, J. (2002), ‘Devolution and Democracy: New trust or old cynicism?’, in Curtice, J., McCrone, D., Park, A. and Paterson, L. (eds), New Scotland, New Society?, Edinburgh: Edinburgh University Press. Curtice, J. (2008), ‘‘How Firm are the Foundations? Public attitudes towards the Union in 2007’, in Devine, T. (ed.), Scotland and the Union 1707–2007, Edinburgh: Edinburgh University Press. Dalton R., Scarrow, S. and Cain, B. (2003), ‘New Forms of Democracy? Reform and Transformation of Democratic Institutions’, in Cain, B., Dalton, R. and Scarrow, S. (eds), Democracy Transformed? Expanding political opportunities in advanced industrial democracies, New York: Oxford University Press. Dewar, D. (1998), ‘The Scottish Parliament’, Scottish Affairs: Special issue on understanding constitutional change, 4–12. Keating, M. and McGarry, J. (2003), ‘Introduction’, in Keating, M. and McGarry, J. (eds), Minority Nationalism and the Changing International Order, Oxford: Oxford University Press. Laffin, M, Shaw, E. and Taylor, G. (2007), ‘The Parties and Intergovernmental Relations’, in Trench, A. (ed.), Devolution and Power in the United Kingdom, Manchester University Press. McLean, I., Lodge, G. and Schmuecker, G. (2008), Fair Shares? Barnett and the politics of public expenditure, London: Institute for Public Policy Research. Mitchell, J. (2004), ‘Scotland: Expectations, Policy Type and Devolution’, in Trench, A. (ed.), The State of the Nations 2004, Exeter: Imprint Academic. Ohmae, K. (1995), The Decline of the Nation State: The rise of regional economies, New York: Free Press. Ormston, R. (2008), Scottish Social Attitudes Survey 2007 – Core Module Report 1: Attitudes to government in Scotland, Edinburgh: Scottish Government Social Research. Available at http://www.scotland.gov.uk/Publications /2008/05/16095134/13. 54

What has Devolution Achieved? Paterson, L. (1998), ‘Scottish Home Rule: Radical Break or Pragmatic Adjustment?’, in Elcock, H. and Keating, M. (eds), Remaking the Union: Devolution and British Politics in the 1990s, London: Frank Cass. Paterson, L., Brown, A., Curtice, J., Hinds, K., McCrone, D., Park, A., Sproston, K. and Surridge, P. (2001), New Scotland, New Politics?, Edinburgh: Edinburgh University Press. Salmond, A. (2008), ‘SNP marks “a year of achievement”’. Available at http://www.snp.org/node/13750. Scottish Constitutional Convention (1995), Scotland’s Parliament, Scotland’s Right, Edinburgh: Scottish Constitutional Convention. Scottish Government (2008), Government Revenue and Expenditure Scotland 2006–2007, Edinburgh: Scottish Government. Scottish Office (1997), Scotland’s Parliament, Cm 3658, Edinburgh: Stationery Office. Scottish Office (1998), Report of the Consultative Steering Group on the Scottish Parliament, Edinburgh: Scottish Office. Seyd, B., (2009), ‘The Citizens’ Response: The performance of the devolved bodies’, in Curtice, J. and Seyd, B. (eds), Has Devolution Worked?, Manchester: Manchester University Press. Trench, A., (2005), ‘Intergovernmental Relations within the UK: The pressures yet to come’, in Trench, A. (ed.), The Dynamics of Devolution: The state of the nations 2005, Exeter: Imprint Academic. Trench, A, (2007), ‘Washing Dirty Linen in Private: the processes of intergovernmental relations and the resolution of disputes’ in Trench, A. (ed.), Devolution and Power in the United Kingdom, Manchester: Manchester University Press.

55

CHAPTER 4 __________________________________________

GOVERNING SCOTLAND: THE PEOPLE’S PREFERENCES? So far we have failed to detect much sign that the outcome of the 2007 election reflected a revolutionary break with the past. There is no evidence of a marked divergence between Scottish and English values, or of an intensification of Scottish national identity that might have contributed to the SNP’s electoral success. Nor did the SNP’s success and Labour’s failure reflect growing dismay with the Scottish Parliament’s ability to achieve the policy outcomes people want, or to involve Scots more in the process of government. The Scottish people had at least as much (or as little) faith in devolution and in the operation the Union at the time of the 2007 election as they had done in previous years. But what of attitudes towards the SNP’s central policy objective, indeed its raison d’être, that is Scottish independence? Surely the party’s electoral success must have been accompanied by increased support for separation from the rest of the United Kingdom? After all, as we noted in chapter 1, the SNP had promised to hold an independence referendum if it secured office and its stance was discussed extensively during the election campaign. Or even if people did not want to go so far as to leave the United Kingdom the SNP’s success must surely at least have denoted an increased wish for a radical shift of power away from London and towards Edinburgh? Experience suggested that heightened support for the SNP often produced an increased impetus towards self-government (Finlay, 2004; Mitchell, 1996). So this chapter examines the development of attitudes towards how Scotland should be governed since the advent of devolution. We focus first on the level of support for independence before looking to see what support there might be for changing the terms of the existing devolution settlement.

56

The People’s Preferences? GROWING SUPPORT FOR INDEPENDENCE? The level of support recorded for Scottish independence varies considerably from survey to survey depending on the way in which the question is asked (Curtice, 2008). If respondents are simply asked whether they are in favour of Scotland being independent or not, without any reference to what is meant by independence, then relatively high and sometimes even majority levels of support are registered. If in contrast independence is characterised as complete separation from the rest of the UK, then support is relatively low. The Scottish Social Attitudes survey has tried to steer an acceptable path between these two positions, on the one hand spelling out not only what independence but also what devolution mean, but on the other hand trying to avoid pejorative language in doing so. Crucially its question gives respondents the opportunity to choose between a range of ways in which Scotland might possibly be governed, rather than simply state whether they are for or against one particular option. The question reads as follows: Which of these statements comes closest to your view? Scotland should become independent, separate from the UK and the European Union Scotland should become independent, separate from the UK but part of the European Union Scotland should remain part of the UK, with its own elected parliament which has some taxation powers Scotland should remain part of the UK, with its own elected parliament which has no taxation powers Scotland should remain part of the UK without an elected parliament This question has been asked in exactly the same way since 1997 and thus, irrespective of the merits or otherwise of its particular wording, gives us the best possible guide to whether support for independence has risen or fallen over time.1 To simplify matters, in Table 4.1 we combine those who chose one of the first two options into a single group labelled ‘independence’, and those who chose either the third or the fourth option into a group of supporters of some form of devolution.2 57

Revolution or Evolution? Table 4.1 Constitutional Preferences, 1997–2007 Sept 1997 % 37 41 17 676

1999

2000

2001

2002

Independence Devolution No parliament Sample size

May 1997 % 28 54 18 882

% 27 60 9 1,482

% 30 54 12 1,663

% 28 59 9 1,605

% 29 52 13 1,665

Independence Devolution No parliament Sample size

2003 26 55 13 1,508

2004 32 45 17 1,637

2005 34 44 14 1,549

2006 30 53 10 1,594

2007 24 62 9 1,508

Source: May 1997: Scottish Election Study; September 1997: Scottish Referendum Study; 1999–2007: Scottish Social Attitudes.

As can be seen typically the question has found that just under 30% of people in Scotland are in favour of independence, while over half are in favour of devolution. One striking exception, ironically, occurred immediately after the 1997 referendum that endorsed devolution; then no less than 37% said they favoured independence, nearly outstripping those who backed devolution. Perhaps the fact that Scotland was finally to get her own parliament meant that independence now seemed more credible too, but whatever the reason this mood had disappeared by the time of the first devolved election in 1999. More importantly for our purposes, however, were signs in 2004 and 2005 that a trend towards increasing support for independence might be setting in. By 2005 support had risen to 34%, not far below its post-referendum peak. Perhaps this provided a platform for the SNP’s success? It seems not. For by 2007 the trend had been reversed. Indeed, far from coinciding with an unusually high level of support for independence, the 2007 election coincided with the lowest level ever recorded by the social attitudes series.3 So once again, an obvious explanation for the success of the SNP eludes us. Its success did not signify increased support for its central policy plank of Scottish independence; instead it coincided with an unusually low level of support for independence. 58

The People’s Preferences? Moreover, we can tie this low level of support with the evidence we uncovered in the previous chapter of increased satisfaction with the Union. Support for independence is much higher (47% in 2007) amongst those who think that England’s economy benefits more from Scotland’s membership of the Union than it is either amongst those who think the two economies of the two countries benefit equally (18%) or those who think Scotland’s benefits most (11%). Equally, those who think that Scotland gets less than its fair share of spending are twice as likely to support independence than those who think it does secure at least its fair share. It seems as though at least part of the reason why support for independence was so low in 2007 was because the Union was now thought to be operating more effectively – thanks, perhaps to the style and behaviour of the newly installed SNP government!4 Table 4.2 Perceived Impact of Scottish Parliament on Scotland’s Membership of the UK, 1997–2007 % say more likely that Scotland will leave UK Scotland will stay in UK No difference Sample size

1997 Ref 42 32 19 676

1999

2000

2001

2003

2007

36 27 28 24 31 25 27 31 27 43 41 40 1,482 1,663 1,605 1,508

32 24 39 1,508

Note that in 1997 and 1999 respondents were asked whether ‘having a Scottish Parliament will make it more likely that Scotland eventually leaves/stays in the United Kingdom’, whereas from 2000 onwards they were asked whether ‘having a Scottish Parliament is going to make it more likely’. Source: 1997: Scottish Referendum Study; 1999–2007: Scottish Social Attitudes.

Nevertheless, the accession of the SNP to government did make it seem more likely that independence would eventually happen. As Table 4.2 shows, just under a third (32%) felt that way in 2007 compared with just under a quarter four years earlier (24%). The increase was more marked amongst supporters of independence (from 44% in 2003 to 57% in 2007) and those who opposed any kind of parliament (from 19% to 39%) than it was amongst supporters of devolution (19% to 25%). In other words the SNP’s success raised the hopes of those who want independence and the fears of those who opposed any kind of Scottish 59

Revolution or Evolution? Parliament, whereas those who preferred devolution were more likely to take the development in their stride. In any event, overall fewer people still thought that devolution was moving Scotland closer to independence than did at the time of the 1997 referendum. STRENGTHENING DEVOLUTION? That the majority of people in Scotland have consistently expressed a clear preference for devolution rather than independence should not, however, be taken as evidence that all is thought to be well with the current constitutional settlement. One clue that this is not the case is to be seen in Table 4.3, which compares what people said when they were asked who does have most influence over the way Scotland is run and who ought to do so. Ever since the parliament’s establishment, consistently between twothirds and three-quarters have declared that the devolved institutions ought to have most influence over the way Scotland is governed. However, the proportion believing the devolved institutions do have most influence is far short of that. Even though it seems that gradually more people have come to regard the devolved institutions as relatively influential, even by 2007 still only just over a quarter (28%) reckoned that it was the Scottish Executive that had most influence over what happened north of the border. Almost half (47%) still believed that the UK government had the greatest influence. So while the gap between expectations and perceived delivery may have become somewhat less than it was, the perceived influence of the devolved institutions still fell far short of what people believed should be the case.

60

The People’s Preferences? Table 4.3 Who Should and Does have Most Influence in Scotland? 1999–2007 1999 2000 2001 2003 % % % % Which ought to have most influence? Scottish 74 72 74 66 Executive UK 13 13 14 20 government

2004 %

2005 %

2006 %

2007 %

67

67

64

71

12

13

11

14

Which does have most influence? Scottish na 13 15 17 19 23 24 28 Executive UK na 66 66 64 48 47 38 47 government Sample size 1,482 1,663 1,605 1,508 1,637 1,549 1,594 1,508 Note: Until 2004 the question referred to the Scottish Parliament rather than the Scottish Executive. In 2004 one half of the sample were asked about the Scottish Parliament, one half the Scottish Executive; there was no significant difference between the answers of the two groups. From 2005 onwards all respondents were asked about the Scottish Executive. Respondents were also offered ‘local councils in Scotland’ and the ‘European Union’ as other possible answers. Relatively few chose these options and to simplify matters the relevant figures are not shown. na: not asked Source: Scottish Social Attitudes, 1999–2007.

It thus, perhaps, should not come as much surprise that a majority have consistently been inclined to agree that the Scottish Parliament should be given more powers. For example, in the wake of the 2007 election, two-thirds agreed that the Scottish Parliament should be given more powers, while only 16% disagreed. It would seem that there is a clear appetite amongst the Scottish public for a devolution settlement that gives Edinburgh significantly greater influence.

61

Revolution or Evolution? Table 4.4 Attitudes towards Giving the Scottish Parliament More Powers, 1999–2007 1999 The Scottish Parliament should be given more powers Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree Sample size

2001

2003

2005

2007

% % % % % 14 20 13 17 22 42 48 47 46 44 20 14 16 17 15 18 12 16 13 13 3 4 6 5 3 1,482 1,605 1,508 1,549 1,508

Source: Scottish Social Attitudes, 1999–2007.

However, we must take care before reaching that conclusion. Firstly, although higher than at any time since 1999, at 22% the proportion that ‘strongly’ agreed in 2007 that the parliament should have more powers was still relatively low; for every person who strongly agreed there were two who only just agreed. This suggests that giving the Scottish Parliament more powers may not be an intensely held view or considered a high priority by many. Secondly, the statement put to respondents is vaguely worded, with no indication of which powers should be transferred to the Scottish Parliament. The apparent consensus in favour of more powers might break down if people were presented with proposals for specific extra powers. Certainly the parties themselves have put forward very different sets of proposals. Thus shortly after coming to power in 2007, the SNP government suggested that a new devolution settlement short of independence might involve giving the Scottish Parliament powers over energy, oil and gas reserves, employment and labour market policies, equal opportunities, competition policy, postal services, economic and fiscal policy, and a broad range of taxes, tax credits and the benefit system, including pensions (Scottish Government, 2007). In contrast, writing as the then Labour leader in the Scottish Parliament, Wendy Alexander suggested a much more modest set of additional powers, including aspects of labour market policy and some revenueraising power (Alexander, 2008). Meanwhile a commission established by the Liberal Democrats and chaired by the former Presiding Officer, Lord 62

The People’s Preferences? Steel, made the case for a greater extension of the parliament’s powers than envisaged by Labour but less than that suggested by the SNP, spanning energy policy, transport, the civil service, the operation of the parliament (including elections), and taxation (Steel Commission, 2006). However, in practice, there does seem to be a reasonable degree of consensus amongst the Scottish public. As we might anticipate and as shown in Table 4.4, over three in five believe that the Scottish Parliament should make the important decisions concerning the NHS and schools – areas already under the parliament’s jurisdiction. But support is just as strong for the idea that the Scottish Parliament should control welfare benefits, which under the existing settlement are wholly reserved to Westminster and which neither the Steel Commission nor Wendy Alexander suggested should be devolved. On the other hand, most people oppose devolving to Edinburgh decision-making authority over defence and foreign policy – powers typically associated with an independent sovereign nation-state. It seems that a majority of people in Scotland want the powers of the Scottish Parliament to extend over most areas of domestic affairs, including ones that many unionist politicians do not wish to devolve, and simply draw the line at defence and foreign affairs. So while they may not want formal independence, they still seemingly favour a high degree of autonomy. Table 4.5 Which Institution should Make Important Decisions about…?

% 63 62 63

UK government at Westminster % 25 13 18

Local councils in Scotland % 10 23 16

33

58

*

Scottish Parliament Health Service Schools Welfare benefits Defence and foreign affairs

Sample size 1,508. * Less than 0.5%. Source: Scottish Social Attitudes, 2007.

63

EU % * * 1 4

Revolution or Evolution? Further evidence of this comes when we look at attitudes towards taxation. As we have already noted in Chapter 3, the funding arrangements remain one of the most contentious features of the existing devolution settlement. In fact that debate is not just about what the level of public spending in Scotland should be, but also about who should decide what that level should be and how spending in Scotland should be financed. Under the devolution settlement the Scottish Parliament has considerable autonomy over how its budget is spent, but very little responsibility for raising revenue. In addition it is not allowed to borrow money on its own account at all. Around 90% of its revenue comes in the form of a block grant directly transferred from the UK government, which thus effectively controls the overall level of spending incurred by the devolved institutions. In practice that level is largely determined by a formula, known as the Barnett formula, under which changes to the level of the grant reflect changes in the level of spending on comparable programmes in England (Bell and Christie, 2007: 73–4). Critics of this arrangement argue that it gives too little freedom to the devolved institutions to decide how much to spend and too little responsibility for ensuring the Scottish economy is strong enough to provide a buoyant flow of tax revenues (Hallwood and Macdonald, 2005; Steel Commission, 2006). Table 4.6 Attitudes towards Funding Services in Scotland out of Taxes Collected in Scotland, 2001–7 Now that Scotland has its own parliament, it should pay for services out of taxes collected in Scotland Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree Sample size Source: Scottish Social Attitudes, 2001–7.

64

2001

2006

2007

% 7 45 18 24 3 1,605

% 5 46 16 25 4 1,508

% 8 49 16 20 2 1,508

The People’s Preferences? It seems that there is considerable support for giving the Scottish Parliament taxation powers. In 2001 and 2003, just over half agreed that ‘Now that Scotland has its own parliament it should pay for services out of taxes collected in Scotland’, while only just over a quarter disagreed with the statement. By 2007 the proposition seemed to be slightly more popular still, with 57% in favour and just 22% opposed. However, few people are either strongly in favour or strongly opposed. Perhaps opinion on the issue is therefore rather ‘soft’ and support might well disappear should people become persuaded that making the Scottish Parliament raise its own revenues would result in either higher taxes or lower public spending. This, however, does not appear to be the case, at least as judged by the answers respondents gave when they were also asked: At the moment, most taxes in Scotland are decided by the UK government. Say that instead all taxes in Scotland were to be decided by the Scottish Parliament. Do you think as a result taxes in Scotland would be higher, lower or would it make no difference? No less than 53% said that they thought that taxes would be higher under this arrangement; just 9% believed they would be lower. And while those in favour of services in Scotland being paid for out of taxes collected in Scotland were somewhat less likely than those who were opposed to think that taxes would increase if all taxes were decided by the Scottish Parliament, still as many as 49% believed that it would. Evidently support for the idea of greater ‘fiscal autonomy’ for the Scottish Parliament does not depend on considerations of its perceived financial consequences. This chimes with previous research that suggested that people voted in favour of the creation of the Scottish Parliament in 1997, including for it to have the ability to vary the basic rate of income tax by 3p in the pound, even though they believed that taxes would go up as a result (Surridge and McCrone, 1999: 45–6). So although independence remained very much a minority preference in 2007, this does not mean that people in Scotland were content with the existing devolution settlement. The majority appear to want a devolved parliament and government that run most aspects of domestic affairs in Scotland, including taxation and welfare benefits which presently remain the responsibility of the UK government, and that as a result the devolved 65

Revolution or Evolution? institutions are clearly the principal players north of the border. They remain unconvinced that the existing devolution settlement meets their aspirations. Here perhaps was a mood for greater autonomy, if not necessarily independence, to which the SNP might have been able to appeal in 2007. However, it is far from clear that this mood was either new or more intense in 2007 than previously. So it is still far from clear why it might have helped SNP secure victory. TERRITORIAL ASSERTIVENESS Constitutional scholars are inclined to agree with the Scottish public that the formal, constitutionally delineated powers of the devolved institutions in Scotland leave those institutions in a less powerful position than the UK government and parliament (Trench, 2007). But the potential power and influence of the Scottish Parliament and government are not wholly determined by their constitutional status. Even on matters that formally are the responsibility of the UK government, the devolved institutions may be able to promote Scotland’s case with UK-level institutions with considerable effect by claiming to be the legitimate voice of the Scottish nation. Indeed, we have already seen in Chapter 3 that most people hoped that devolution would strengthen Scotland’s voice in the United Kingdom, but that prior to our 2007 survey at least many had doubts about how far this had been achieved. Perhaps the SNP were able to tap into these doubts? Certainly during the 2007 campaign the SNP argued that the government in Scotland should adopt a more assertive approach in its dealings with the UK government. It wanted to ‘maximise the role and influence of the Scottish government across the full range of reserved and devolved policy areas’ and ensure that the relationship between the two institutions comprised a ‘partnership of equals’ (SNP, 2007; see also Swenden and McEwen, 2008). Labour in contrast repeatedly emphasised its wish to work in ‘partnership’ with the UK government without disturbing the current balance of powers or relations (Scottish Labour Party, 2007).

66

The People’s Preferences? Table 4.7 Attitudes towards Scottish Ministers Rowing with UK Government by National Identity and Attitudes towards How Scotland should be Governed, 2007 Government ministers in the Scottish Executive should always stand up for what they think is best for Scotland, even if it means having serious rows with the UK government at Westminster Strongly Agree Neither/ Sample agree disagree size National identity: Scottish not British % 55 40 4 393 More Scottish than % 42 48 9 452 British Equally Scottish and % 27 58 13 410 British More British than % 35 34 30 76 Scottish British not Scottish % 35 51 13 94 Constitutional preference: Independence % 66 31 3 330 Devolution % 34 57 10 953 No parliament % 25 48 27 142 Scottish Parliament should have more 330 17 6 powers: % 77 6 670 35 59 Strongly agree % 60 15 231 23 Agree % Neither agree nor 194 61 20 % 17 disagree 43 46 30 Disagree % 25 Strongly disagree All % 40 48 12 1,508 Source: Scottish Social Attitudes, 2007.

67

Revolution or Evolution? The more assertive tone adopted by the SNP seemed to be the one that was closer to the mood of the Scottish public. No less than 88% agreed that ‘Scottish Ministers should always stand up for what they think is best for Scotland even if it means having serious rows with the UK government at Westminster’, with as many as 40% agreeing strongly. True, as Table 4.7 shows those who favoured independence, or who strongly agreed the parliament should have more powers exhibited the greatest enthusiasm for the adoption of a tough stance towards Westminster, together with those who said they were Scottish and not British. In other words, the idea appealed most strongly to those of a nationalist bent. But even so, the proposition also secured majority support across all other groups, including those who said they opposed any kind of Scottish Parliament and who indicated they were British not Scottish. It seems that even those most out of sympathy with devolution, let alone independence, accepted that the devolved institutions had a responsibility to pursue Scotland’s case at Westminster with vigour. Unfortunately, previous surveys did not address this issue and so it is impossible to state whether there was a greater demand in 2007 than previously for the adoption of a more assertive stance towards the UK government, a demand that might have helped the SNP enjoy electoral success for the first time. All we can note is that in 2007 the SNP’s stance matched the public mood, and that indeed, as we have already noted in Chapter 3, no sooner had it acquired the reins of power than people were much more likely to think that Scotland had a stronger voice within the UK. There thus seems little doubt that on this issue at least the SNP were singing a popular tune. CONCLUSION So once again it seems that an obvious, common sense explanation of the SNP’s success in 2007 is in error. It was not the result of a growing demand for independence, and thus a revolutionary break from the rest of the United Kingdom. This does not mean, however, that there was not a mood from which the SNP could profit. The powers assigned to the Scottish Parliament under the existing devolution settlement seem to fall short of those that a majority of people in Scotland believe it should be able to wield. Equally, there was a wish for a more assertive tone in the 68

The People’s Preferences? stance that was adopted by the devolved government towards the UK government. However, there is no consistent evidence that this mood was any stronger in 2007 than it had been previously. So even in this respect there is little evidence that a revolution was afoot. We have, then, uncovered a picture of remarkable stability in the identities and attitudes of the Scottish public since the advent of devolution. There was no gradual or sudden change in the public’s outlook that might plausibly have accounted for the nationalists’ electoral success. We thus need to look more directly at how and why voters behaved as they did when they cast their ballots. One explanation could be that they thought little of the record of the incumbent Labour-led administration and wanted to see if someone else could do a better job. Perhaps in 2007 devolution evolved into a system that, like many a liberal democracy, was capable of replacing an unpopular government with a more popular opposition. It is to that possibility that we turn next. NOTES 1. Of course we cannot be sure that the meaning of each option has not changed over time. For example, the reference in the third option to ‘some taxation powers’ might in the early years of devolution have been taken to refer to the power the parliament already has to vary the basic rate of income tax by up to 3 pence in the pound. By 2007, by which time this limited power was widely regarded as impractical, it might be regarded as a reference to giving the parliament wider taxation powers than it currently has. This argument does, however, assume that the public have played close attention to the elite level debate about Scotland’s constitutional future. 2. Typically those favouring independence within the EU have outnumbered those preferring independence outside it by around two to one, while the proportion favouring devolution without taxation powers has never exceeded one in ten. 3. Should this finding appear incredible, it might be noted that, irrespective of how they worded their question, all opinion polls that asked the same question about constitutional preferences identified a fall in support during the course of the 2007 election campaign (Curtice, 2008). 69

Revolution or Evolution? 4. The decline in support for independence between 2005 and 2007 cannot, however, simply be accounted for by the more positive perceptions of the Union that existed in 2007. Support for independence even fell amongst those who had negative views. Thus for example, in 2005, 51% of those who thought England’s economy benefited more from the Union favoured independence, four points higher than in 2007. However, even bigger declines, of nine and thirteen points respectively, were registered amongst those who thought the two economies benefited equally and those who believed Scotland’s benefited most. In other words, those with more positive views about the functioning of the Union were particularly unlikely to support independence in 2007. REFERENCES Alexander, W., (2008), Scottish Labour: New Directions: Change is what we do, Glasgow: Scottish Labour Party. Bell, D. and Christie, A. (2007), ‘Funding devolution: the power of money’, in Trench, A. (ed.), Devolution and Power in the United Kingdom, Manchester: Manchester University Press. Curtice, J. (2008), Where Stands the Union Now? Lessons from the 2007 Scottish parliament elections, London: Institute for Public Policy Research. Finlay, R. (2004), Independent and Free: Scottish politics and the origins of the Scottish National Party, 1918–1945, Edinburgh: John Donald. Hallwood, P. and MacDonald, R. (2005), ‘The Economic Case for Fiscal Federalism in Scotland’, in Coyle, D., Alexander, W. and Ashcroft, B. (eds), New Wealth for Old Nations: Scotland’s economic prospects, Princeton: Princeton University Press. Mitchell, J. (1996), Strategies for Self-Government, Edinburgh: Polygon. Scottish Labour Party (2007), Building Scotland, Glasgow: Scottish Labour Party. Scottish National Party (2007), Manifesto 2007, Edinburgh: Scottish National Party. Steel Commission (2006), Moving to Federalism: A new settlement for Scotland, Edinburgh: Scottish Liberal Democrats. Surridge, P. and McCrone, D. (1999), ‘The 1997 Scottish Referendum Vote’, in Taylor, B. and Thomson, K. (eds), Scotland and Wales, Nations Again?, Cardiff: University of Wales Press. 70

The People’s Preferences? Swenden, W. and McEwen, N. (2008), ‘Scotland after Devolution: Regional autonomy in the shadow of hierarchy?’, paper prepared for presentation at the 4TAD conference, Milan, 12–14 June. Available at http://www.4tad.org/4TAD_workshop1.htm. Trench, A. (2007), ‘The Framework of Devolution: The formal structure of devolved power’, in Trench, A. (ed.), Devolution and Power in the United Kingdom, Manchester: Manchester University Press.

71

  CHAPTER 5 __________________________________________

LOST LABOUR VOTES? RECORDS, PERSONALITIES AND ISSUES So far we have looked at the climate of opinion in which the 2007 Scottish parliamentary election was fought. We have uncovered little evidence to support some of the more obvious and immediate explanations of the SNP’s success. People in Scotland were no more likely than before to regard themselves as Scottish rather than British. Scotland had not grown apart from England in its values and attitudes. Devolution may have brought its disappointments, but it is not clear that dissatisfaction with its performance had increased over recent years. And, above all, it is not apparent that support for independence had grown. But of course in focusing on people’s sense of national identity, their values and their constitutional preferences, we are making an important assumption – that these are the things that actually make a difference to the way that people vote. Politicians might talk about the respective merits of independence and devolution, but it does not follow that people decide how to vote on the basis of their view of that debate. Parties may invoke the symbols of Scottish or British national identity, but perhaps they are mistaken in believing that such symbols actually resonate in voters’ hearts and minds. So in this second part of the book we look directly at what considerations appear to have influenced the way that people voted in the 2007 parliamentary election. We do so in two parts. In this chapter, we look at two possible influences on the way in which people voted that we have not considered so far. First, we consider whether voters took into account their views about how well or badly they believed the Labour/Liberal Democrat coalition had performed since it was first formed in 1999? After all, if they held an unfavourable view this might well have given them sufficient reason to want to see whether the SNP could do a better job instead. Second, we examine whether the way in which people voted reflected their views on some of the key differences in the policy proposals put forward by the parties. In so doing our main focus in both cases is how far these possible influences might have cost Labour support. In the next  

72

Lost Labour Votes? chapter we look directly at what factors were associated with people voting for and switching to the SNP. It should hardly come as much surprise that it might be thought the outcome of the 2007 parliamentary election was influenced by people’s evaluations of the performance of the incumbent Labour/Liberal Democrat coalition government. Indeed such a process might well be regarded as an indication that devolution was evolving as originally intended. After all, it is frequently argued that holding governments to account for their performance in office is one of the key functions that democratic elections are designed to fulfil (Katz, 1997; Powell, 2000; Schumpeter, 1976). If governments are aware they might be ejected from office at the next election should they be thought to have performed badly, they have a strong incentive to meet public expectations. Equally we would consider it unremarkable that the way in which people vote reflects their views about the respective merits of the policy platforms of the parties. Such behaviour helps ensure the policies that are implemented by governments reflect the wishes of a majority of the electorate. Moreover, a presumption that this is precisely what would happen in devolved elections underlay many of the arguments that were deployed in favour of creating the Scottish Parliament in the first place. Advocates of devolution wanted to ensure, for example, that allegedly unpopular policies, such as the ‘poll tax’ that was introduced in the 1980s, could not be visited on Scotland in future. Equally, they wanted to ensure that if the people of Scotland were unhappy with the performance of the incumbent government, they would be able to turn it out of office and not have their views overridden by votes cast in England. But in making these arguments they made some crucial assumptions about how voters would behave. They assumed voters would vote against parties that promote unpopular policies. And they anticipated that voters would actually act on their evaluations of the performance of the incumbent government. Yet in truth there is no guarantee that in any election voters will vote on the basis of their views of the incumbent government or their policy preferences (Anderson, 2007; Bartels, 1996; Butler and Stokes, 1974; Delli Carpini and Keeter, 1996; Evans and Andersen, 2006; Kiewet, 1983; Milner, 2002; Page and Shapiro, 1992; Popkin, 1991; Whitten and Palmer, 1999). They might be heavily influenced by a strong sense of partisanship that inclines them to the view that anything done or proposed by their favourite party is inherently good. As a result their 73

Revolution or Evolution? evaluations of the government’s performance and their policy preferences become a reflection of their party preference, rather than vice-versa. Even if this is not the case, if they are to vote effectively on the basis of the government’s record or the issues of the day, voters require some knowledge and understanding of the incumbent government’s performance in office and the various parties’ policy proposals; perhaps they do not have the inclination or ability to acquire such knowledge. But if there are doubts about the ability of any kind of election to promote accountable and representative government, there are particular reasons to wonder whether devolved elections are capable of doing so (Curtice, 2006; 2009). One reason is that once there is more than one level of government, it becomes more difficult for voters to assign credit or blame for what has happened (Anderson, 2006; Gélineau and Bélanger, 2005). If, for example, voters in Scotland are unhappy at the level of unemployment in their country, should they blame the UK government, which has responsibility for macroeconomic policy, or the Scottish government, which has particular responsibility for attracting firms and jobs to Scotland? Moreover, even if they are clear that blame should be attached to the Scottish government, which of the two parties in the coalition should be considered responsible? Holding governments to account for their actions is arguably particularly difficult where, as in Scotland, there is coalition government and proportional representation (Nadeau et al., 2002; Powell, 2000; Powell and Whitten, 1993). The existence of a more complex political structure is not, however, the only reason why voters might be less likely to vote on the basis of their evaluations and policy preferences in devolved elections. Another possibility is that they are less likely to have the motivation to do so than in UK general elections (Reif, 1984; Reif and Schmitt, 1980). As we have already seen in Chapter 4, most people in Scotland feel that the devolved institutions are less powerful than their UK-level counterparts at Westminster. Perhaps, as a result, they take the view that the outcome of Scottish Parliament elections matters less too (Bromley, 2006). In that event, voters might think it makes more sense to regard a Scottish Parliament election as an opportunity to express their views about the performance of the current UK government (and especially so if they are unhappy with its performance) than to vote on the basis of their evaluations of the work of the Scottish government. Certainly, according 74

Lost Labour Votes? to the theory of ‘second order’ elections, this is often what happens in elections to bodies that are considered less important. So in considering whether voters voted on the basis of their evaluation of the performance of the incumbent government or the policy proposals put forward by the parties, we take into account these doubts and scepticism about the effectiveness of devolved elections. Thus, we do not just ask whether voters abandoned the Labour party because they were unhappy with its performance in office in Edinburgh. We also consider the possibility that responsibility for the party’s defeat lay at the door of the party at Westminster, because voters used the election as an opportunity to express their dislike of the party’s performance as the UK government. Equally, as well as examining whether people did not vote Labour because they did not like the party’s policy platform for the devolved institutions over the next four years, we also assess whether, instead, it was the parties’ stances on those issues that are primarily the preserve of Westminster that held the greater sway. ACCOUNTABILITY We begin by looking at how far voters appear to have held the incumbent Scottish Executive administration to account for its record in office, or whether they seem instead to have been influenced by their evaluations of the performance of the UK government. To do so we look first of all at how well or badly the two administrations are thought to have performed before considering what impact those evaluations appear to have had on the way in which people voted in the 2007 parliamentary election. In Table 5.1 we show how people responded when asked simply and directly, first, how good or bad a job of running Britain they thought the UK Labour government had done in recent years, and, second, how good or bad a job of running Scotland Labour ministers in the Scottish Executive had done. Neither group seems to have been particularly unpopular. Just under half (49%) reckoned the UK Labour government had done a ‘fairly good’ or ‘very good’ job, whereas only slightly below a quarter (23%) reckoned they had done a ‘fairly bad’ or ‘very bad’ one. Labour’s record on the Scottish Executive received slightly fewer plaudits, but fewer brickbats too. Just under two in five (39%) felt the party had done a good job in Edinburgh, while only one in five (20%) thought they 75

Revolution or Evolution? had done a bad job. On this evidence it is not immediately obvious that voters’ judgements of either government could have cost Labour many votes. Table 5.1 Overall Evaluations of Labour’s Performance in the UK Government and Scottish Executive, 2007

Very good Fairly good Neither good nor bad Fairly bad Very bad Sample size

Evaluation of Labour’s Performance in UK Government Scottish Executive % % 6 2 43 37 25 33 15 15 8 5 1,312 1,312

Source: Scottish Social Attitudes, 2007.

Of course, as we have already noted, Labour was not the only party in office in Edinburgh; the Liberal Democrats were its junior coalition partners. However, few voters appear to have formed a clear view one way or the other about how well the Liberal Democrat ministers in the Scottish Executive had performed. As many as 18% indicated they were unable to say (more than twice as many as the 8% who were unable to evaluate the Labour ministers) while another 41% said they were neither good nor bad. It thus seems that relatively few voters were strongly aware of the Liberal Democrats’ role and performance in the Scottish Executive, and thus we do not pay any further attention to the impact that evaluations of that party’s ministers had on how people voted. As well as asking respondents how well or badly they thought Labour had performed in London and Edinburgh, we also asked them whether they thought various services and aspects of life in Scotland had got better or worse over the course of the last year. So, for example, as can be seen in Table 5.2, one in five (20%) reckoned that the quality of education had increased, rather more than the one in eight (12%) who thought it had got worse. Those with a positive evaluation of recent trends on the economy, the standard of living and the standard of public transport also outnumbered those with a negative one, though in the case of the standard 76

Lost Labour Votes? of the National Health Service (NHS) the opposite was true. Even so, on balance people appear to have had a better opinion of the NHS than they had done at the time of the 2003 election, as indeed was also true of the quality of education, while evaluations of the general standard of living were no worse. Again, it is not immediately obvious that perceptions of Labour’s record in office should have cost it dear. Table 5.2 Evaluations of Policy Performance, 2003 and 2007 In last 12 months*

Has

2003 %

2007 %

Increased Stayed the same Fallen

20 25 46

19 45 26

Increased Stayed the same Fallen

25 27 29

20 45 12

Increased Stayed the same Fallen

na na na

27 38 18

Increased Stayed the same Fallen

34 36 24

29 41 20

na na na 1,508

31 36 11 1,508

Standard of NHS

Quality of education

Standard of public transport

General standard of living

Scotland’s economy Got stronger Stayed same Got weaker Sample size na: not asked * in 2003 the question referred to the last four years. Source: Scottish Social Attitudes, 2003, 2007.

77

Revolution or Evolution? But whom did voters regard as primarily responsible for these trends – the Scottish Executive, the UK government or neither? After all, even if they were voting on the basis of their perceptions of recent trends in the NHS, it can hardly be argued that in doing so voters were holding the Scottish Executive to account if in fact they felt that responsibility for those trends lay primarily with the UK government. The devolution settlement might assign formal responsibility for the health service to the Scottish Executive – and indeed both education and much of public transport too – but there is no guarantee that the public will see things that way. After all, we have already seen in Chapter 3 that in general people in Scotland regard the UK government as more influential than the Scottish Executive. Indeed, Table 5.3 confirms the need for caution. While at the time of the 2007 election more people reckoned primary responsibility for recent trends in both education and public transport lay with the Scottish Executive than felt it lay with the UK government, the same was not true of trends in the standard of the NHS. Just over a third (34%) said the UK government was primarily responsible for trends in the NHS, while a little under a quarter (23%) believed the Scottish Executive’s policies were mainly responsible. However, as we might anticipate given its continued responsibility for macroeconomic policy, the area where people were most likely to assign responsibility to the UK government was the general standard of living in Scotland. Even so, when respondents were asked who was responsible for the strength of Scotland’s economy, they were then slightly more likely to consider responsibility to lie with the Scottish Executive.

78

Lost Labour Votes? Table 5.3 Attributions of Responsibility, 2007 Outcome Standard of NHS

Quality of education

Result primarily of

%

Scottish Executive policies UK government policies Some other reason Don’t know/Both

23 34 17 26

Scottish Executive policies UK government policies Some other reason Don’t know/Both

28 22 15 35

Scottish Executive policies UK government policies Some other reason Don’t know/Both

31 17 22 30

Scottish Executive policies UK government policies Some other reason Don’t know/Both

19 37 22 23

Scottish Executive policies UK government policies Some other reason Don’t know/Both

28 24 15 33 1,508

Standard of public transport

General standard of living

Scotland’s economy

Sample size Source: Scottish Social Attitudes, 2007.

Still, whatever the precise balance of opinion on each subject, in all cases there is both a significant proportion of people who regard the Scottish Executive as responsible and a significant proportion, the UK 79

Revolution or Evolution? government. If we find that those voters with negative evaluations were less likely to vote for the Labour party, we will still have to establish further whom they regarded as responsible before we can be clear which government they were holding to account. But we should note at this stage that typically only between a half and two-thirds believed that either government was primarily responsible for what had happened. Some believed neither government was responsible, while around between a quarter and a third either said they did not know who was responsible1 or believed that both governments were. Many voters evidently did not find it easy to be clear which layer of government influences what. Table 5.4 Labour Loyalty by Evaluations of UK Government and Scottish Executive Evaluation of Labour’s record in % 2003 Labour voters voting Labour in 2007 Evaluation Good Sample size Not good Sample size

UK government

Scottish Executive

64 241 38 195

62 191 49 132

Source: Scottish Social Attitudes, 2007.

But is there any reason to believe that voters actually acted on their perceptions of the performance of the incumbent Scottish Executive or the UK Labour government? Is there any evidence that those with a negative evaluation of either government were less likely to vote Labour?2 Table 5.4 provides one simple approach to answering that question. It confines its attention to those respondents who claim to have voted Labour on the constituency ballot at the last Scottish Parliament election in 2007. In the first column this group of previous Labour voters is divided, first of all, between those who thought that the UK government had done a good job of running Britain in recent years and those who did not (including those who said it had done neither a good nor a bad job). In the second column, meanwhile, the group is divided between those who said Labour ministers in the Scottish Executive had done a good job of running Scotland and those who thought it had not. In each case the table 80

Lost Labour Votes? shows the proportion of former Labour voters who turned out and backed Labour again in 2007.3 Those who reckoned that Labour had done a good job in government were indeed more likely to turn out and vote for Labour once again in 2007. Moreover, this was true of evaluations of the party’s performance at Westminster and in Edinburgh. Thus, for example, nearly two-thirds (64%) of 2003 Labour voters who thought the UK government had done a good job turned out and voted for Labour again in 2007, whereas less than two-fifths (38%) of those who had a less favourable impression did so. However, we should note that the gap between these two groups in their likelihood of remaining loyal to Labour – a difference of 26 percentage points – is twice as large as the equivalent gap – 13 points – when the group is divided according to its opinion of the party’s record on the Scottish Executive.4 This would seem to suggest that a less-thanfavourable opinion of Labour’s record in office at Westminster was more likely to deter voters from backing Labour in the 2007 parliamentary election than was an unfavourable perception of its record in Scotland. However, there are some important limitations to this analysis. We are only looking at the behaviour of former Labour voters; it takes no account of the possibility that some people who did not vote Labour switched towards the party in 2007 because they were impressed by the job it had done in either London or Edinburgh. It does not allow for the possibility that those who felt Labour had done a ‘very’ bad job were more likely to have been reluctant to back the party than those who just felt it had performed indifferently. Meanwhile, we have not established that the differences in Table 5.4 are statistically significant once we take into account the number of people on whom our statistics are based. All three limitations can be addressed by undertaking some multivariate analysis. In this analysis, the results of which are shown in Table 5.5, we seek to establish what made a difference to whether someone voted Labour or not. In doing so, we first consider how respondents voted on the constituency vote in 2003. As indicated by the large positive coefficient for the relevant entry in the table, this reveals, as we would expect, that those who voted Labour in 2003 were significantly more likely to support Labour in 2007 than were those who had abstained four years previously. As it happens the same is also true of those who voted for one of the smaller ‘other’ parties in 2003, from whose decline in support in 2007 Labour was evidently one of the beneficiaries.5 81

Revolution or Evolution? Table 5.5. Logistic Regression of Impact of Overall Evaluations on Voting Labour in 2007 Voted Labour 2007 v. Not voted Labour Coefficient (and standard error) Recall constituency vote 2003 Conservative Labour Liberal Democrat SNP Other (Abstained) Evaluation of Labour in devolved government Labour in UK government Nagelkerke R-squared Sample size

-0.52 2.73 0.43 -0.18 1.68

(.81) (.24)* (.43) (.50) (.55)*

0.24 (.15) 0.76 (.15)* 48% 1,061

In the case of recall vote the coefficients are simple contrast coefficients. The baseline category is indicated in brackets. Evaluations of Labour’s performance are included as interval level variables. Those who were unable to say how well or badly Labour had done were coded as though they said ‘neither good nor bad’. * Significantly different from zero at the 5% level. Source: Scottish Social Attitudes, 2007.

Because we have now taken into account the link between how people voted in 2003 and whether or not they supported Labour in 2007, any other variable can only emerge as important in our multivariate analysis because there is a link between that variable and switching to or away from Labour between 2003 and 2007. So when we now also include in our analysis people’s evaluation of Labour’s performance in the UK government and on the Scottish Executive, as detailed earlier at Table 5.1, we are establishing whether such evaluations apparently led some people to switch their votes. The result is clear. Evaluations of Labour’s performance at Westminster are significantly associated with switching to or away from Labour in the 2007 Scottish parliamentary election; the same, however, is not true of evaluations of the party’s performance in the Scottish Executive. In other words the impression we formed from the 82

Lost Labour Votes? analysis in Table 5.4, that voters’ views about the UK government made more difference to their chances of voting Labour than did perceptions of the Executive, is confirmed. Table 5.6 Labour Loyalty by Evaluations of the NHS and Standard of Living % 2003 Labour voters voting Labour in 2007 Evaluation Increased Sample size Stayed same Sample size Fallen Sample size

Evaluation of Standard of NHS 62 81 56 164 49 107

Standard of living 54 126 57 151 52 71

Source: Scottish Social Attitudes, 2007.

What, however, do we discover if we look at evaluations of particular public services and the economy? Are these more detailed views linked to Labour’s fortunes? Table 5.6 gives us an initial impression. It repeats the kind of analysis of the loyalty of Labour voters that we undertook in Table 5.4, but on this occasion we divide those voters according to their perceptions of recent trends in the health service and the standard of living. There is a sharp contrast between the two columns in the table. In the first column we can see that those 2003 Labour voters who felt that the standard of the NHS had fallen were as much as 13 points less likely than those who thought the standard had increased to remain loyal to Labour in 2007, with those who thought the standard was unchanged falling in between. In contrast, the second column shows that the loyalty of former Labour voters was much the same amongst those who felt that the standard of living had fallen as it was amongst those who thought it had increased. Again, multivariate analysis (see Table 5.7) confirms the impression of our simpler analysis of rates of Labour loyalty. In this case, we allowed our analysis to include any of the five evaluations we introduced in Table 5.2, but only if they were statistically significant. And as we can see, only 83

Revolution or Evolution? evaluations of the NHS passed this test. After taking into account how they had voted in 2003, those who felt that the standard of the NHS had fallen were less likely to support Labour in 2007 than were those who thought standards had increased. Indeed, those who said they did not know whether standards had increased or fallen were reluctant to support Labour too. Table 5.7 Logistic Regression of Impact of Policy Evaluations on Voting Labour, 2003 and 2007 Voted Labour 2007 v. Not voted Labour Coefficient (and standard error) Recall constituency vote 2003 Conservative Labour Liberal Democrat SNP Other (Abstained) Evaluation of standard of NHS (Increased) Stayed same Fallen Don’t know Nagelkerke R-squared Sample size

- 0.94 3.04 0.45 -0.43 1.29

(.80) (.24)* (.42) (.46) (.52)*

-0.36 (.22) -0.77 (.25)* -0.90 (.40)* 43% 1,209

Coefficients are simple contrast coefficients that show the impact on the probability of voting Labour of being in the relevant category rather than in the baseline category. The baseline category is indicated in brackets. * Significantly different from zero at the 5% level. Source: Scottish Social Attitudes, 2007.

So it seems that evaluations of what is principally a responsibility of the Scottish Executive rather than the UK government, the state of the health service, were linked to Labour’s fortunes in 2007. Moreover, while not as negative as four years earlier, those evaluations were on balance negative. Thus evaluations of the health service may well have cost Labour some 84

Lost Labour Votes? votes in 2007. However, this does not necessarily mean that voters were holding the Scottish Executive to account. For we have already seen in Table 5.3 that in fact more people reckoned responsibility for the state of the NHS lay with the UK government than felt it could be accounted for by the policies of the Scottish Executive. Indeed, if we look a little more deeply at whom voters held responsible for the state of the health service, the degree to which blame for the state of the health service in Scotland was often laid at the door of the UK government becomes even more apparent. For those who felt that the standard of the health service had fallen were particularly likely to regard the UK government as responsible. Thus, whereas only 34% of those who believed that the standard of the health service had increased believed the UK government was responsible (while 50% reckoned the Scottish Executive was), no less than 45% of those who felt the health service had got worse did so (with just 14% blaming the Scottish Executive). So people were particularly likely to blame the government in London when they felt that things had gone wrong (see also Given and Ormston, 2006; Ormston and Sharp, 2007; Park and McCrone, 2006). However, to examine further whom voters were holding accountable for the state of the health service, we added an additional feature to our multivariate analysis. We created a variable that enabled us to establish whether people were particularly likely to act on their evaluations of the NHS if they felt that the Scottish Executive was primarily responsible for recent trends in the standard of the service.6 This did not, however, prove to be statistically significant. In other words people who thought the standard of the NHS had fallen were disinclined to vote Labour in 2007 irrespective of who they thought was responsible for that fall. It simply so happens that most of them blamed the UK government. So what lessons emerge from our analysis so far? First of all, it casts doubt on the claim that the creation of Scottish Parliament elections has served to ensure that the performance of the government in Scotland is held to account by the electorate. Voters might have been moved by their views about the state of the NHS in Scotland, but although formal responsibility for the NHS lies with the Scottish Executive, in practice many voters felt that it was the UK government that should get the blame. Moreover, those who blamed the UK government for what they regarded as a decline in the NHS seem just as likely to have been discouraged from voting Labour as were those who blamed the administration in Edinburgh. 85

Revolution or Evolution? Meanwhile, when we compare directly the apparent influence of evaluations of the UK government and of the Scottish Executive on how people voted, it seems to have been the former that moved them the more. But what were the consequences of this for Labour? It suggests the party’s defeat cannot be blamed on Edinburgh rather than Westminster. Its performance at Westminster might – despite the party’s low standing in British opinion polls at the time – have been evaluated a little more highly than that in Edinburgh, but disenchantment with the party at Westminster was the more powerful disincentive to voting Labour. Meanwhile, Labour may have lost some ground because of what on balance were negative evaluations of the performance of the health service in particular. But as those evaluations were no more negative than four years previously – indeed they were actually less so – this on its own cannot account for the party’s defeat in 2007. LEADERSHIP The approach we have taken so far is not, however, the only one that might be used to assess whether voters held their devolved rulers to account in 2007. It presumes that what voters were concerned about was the success or otherwise of the performance of a government as a whole. But perhaps what voters focused on instead were the personalities at the head of those governments. It is, after all, often argued that democratic elections have become ‘presidentialised’, that is that the demands of the modern media mean the spotlight of attention is concentrated on the party leaders and that as a result voters vote for the leader that they consider best rather than which party as a whole they most prefer (Bean and Mughan, 1989; Clarke et al., 1979; Clarke et al., 2004; Glaser and Salmon, 1991; Graetz and McAllister, 1987; Mughan, 1995; Mughan, 2000; Stewart and Clarke, 1992). So perhaps Labour lost ground because voters were unhappy with either the Scottish First Minister, Jack McConnell, or the UK Prime Minister, Tony Blair?

86

Lost Labour Votes? Table 5.8 Evaluations of Party Leaders, 2003 and 2007

Tony Blair Good (7–10) Middle (4–6) Bad (0–3) Don’t Know Jack McConnell Good (7–10) Middle (4–6) Bad (0–3) Don’t Know John Swinney/ Alex Salmond Good (7–10) Middle (4–6) Bad (0–3) Don’t Know Sample size

2003 %

2007 %

41 39 18 2

39 35 24 2

20 48 18 14

23 45 22 10

9 37 32 22 1,508

39 35 14 13 1,508

Source: Scottish Social Attitudes, 2007.

Table 5.8 shows how popular these two leaders were in 2007, and how this compares with the position in 2003. In both cases respondents to our survey were asked to give the leader a mark out of 10 to show how good or bad a job they had done as First and Prime Minister respectively. In the table we have summarised these scores by showing the proportion who gave each leader a ‘good’ score, that is a mark of seven or more, a ‘middle’ score of between four and six, and a ‘bad’ one of three or less. Tony Blair clearly had more admirers than Jack McConnell. Just under two-fifths (39%) gave the Prime Minister a good score whereas slightly less than a quarter (23%) did the same for Jack McConnell. However there were also slightly more people who gave the Prime Minister a bad score than gave one to Jack McConnell. It seems, as indeed had also been the case in 2003, that many a voter in Scotland was largely indifferent to Jack McConnell’s performance, with over half either giving him a 87

Revolution or Evolution? ‘middle’ score or else proving unable to say how good or bad a job he was doing. Moreover, Mr McConnell certainly seems to have been less attractive to voters than the SNP leader, Alex Salmond. When asked how good a job they thought he would do as First Minister, nearly three times as many people gave Mr Salmond a good score as marked him down. In this respect Mr Salmond was very different from his predecessor as SNP leader, John Swinney. In 2003 over one in five (22%) were unable to mark Mr Swinney at all, while over three times as many gave him a bad score as reckoned he merited a good one.7 Labour was evidently facing stiffer competition from the SNP than had been the case four years previously. Table 5.9 Labour Loyalty by Evaluations of Labour Party Leaders % 2003 Labour voters voting Labour in 2007 Evaluation Good Sample size Middle Sample size Bad Sample size

Evaluation of Blair 62 237 47 99 35 38

McConnell 68 137 49 172 42 53

Source: Scottish Social Attitudes, 2007.

But did what voters think about Tony Blair and Jack McConnell make any difference to the way they voted? Table 5.10 repeats the kind of analysis of loyalty to the party of those who voted Labour in 2003 we undertook previously in Tables 5.4 and 5.6. It shows that those 2003 Labour voters who thought highly of Mr Blair in 2007 were far more likely to support the party again than were those who had a low opinion of him; the former group were no less than 27 percentage points more likely to support Labour again. But the picture is much the same in respect of Mr McConnell too. Those who gave the incumbent First Minister a low score were 26 points less likely than were those who gave him a high score to remain loyal to Labour. It seems that voters did indeed use the election 88

Lost Labour Votes? to hold their First Minister accountable, though only in tandem with regarding it as an opportunity to express a judgement about the Prime Minister too. Table 5.10 Logistic Analysis of Impact of Leadership Evaluations on Voting Labour, 2003 and 2007 Voted Labour 2007 v. Not voted Labour (1) (2) Recall constituency vote 2003 Conservative Labour Liberal Democrat SNP Other (Abstained) Evaluation of McConnell Blair Salmond Nagelkerke R-squared Sample size

-0.66 2.89 0.46 -0.18 1.80

(.80) (.24)* (.42) (.47) (.54)*

0.16 (.05)* 0.20 (.04)* 47% 1,205

-0.78 2.89 0.52 0.22 1.91

(.81) (.24)* (.43) (.48) (.54)*

0.19 (.05)* 0.20 (.04)* -0.19 (.04)* 49% 1,205

Main cell entry is the parameter coefficient; entry in brackets represents the associated standard error. In the case of recall vote the coefficients are simple contrast coefficients. The baseline category is indicated in brackets. Evaluations of the party leaders are included as interval level variables scored from 0 to 10. Those who said they could not score a leader were regarded as having given him a score of 5. * Significantly different from zero at the 5% level. Source: Scottish Social Attitudes, 2007.

This impression is confirmed by multivariate analysis of all voters that takes into account how people voted on the constituency vote in 2003. As can be seen in the first column of Table 5.8, both evaluations of Mr Blair and evaluations of Mr McConnell were significantly associated with a decision to vote Labour; moreover, the strength of that association is more or less the same in both cases. So here there is some evidence of voters using the Scottish Parliament election to hold their incumbent First 89

Revolution or Evolution? Minister to account – but only in tandem with using the occasion to express a judgement about the Prime Minister too. Even so, as we have already noted, neither Mr Blair nor Mr McConnell was particularly unpopular. So even if the voting decision of some voters was influenced by their views about the two Labour leaders, it is not immediately apparent that this should have been to the party’s significant disadvantage, But we also noted earlier the popularity of the SNP leader and candidate for First Minister, Alex Salmond. Perhaps Labour’s difficulty was not that their leaders were particularly unpopular, but that their allure paled somewhat when compared with the SNP leader. To assess whether the popularity of the SNP leader cost Labour votes, in the second column of Table 5.8 we add people’s evaluations of Mr Salmond to our analysis. It confirms that even after taking into account their opinions of the two Labour leaders, those who thought highly of Mr Salmond were less likely to vote Labour. It looks as though Labour’s difficulty in 2007 was not only its own record, but also the attractiveness of the main alternative for many voters. The particular appeal of the SNP will be discussed in more detail in the next chapter. ISSUES What, however, of the second role that devolved elections are intended to fulfil, to help ensure that the views of those who govern Scotland are representative of the opinions of the population of Scotland – and especially so in respect of those issues that actually lie within the power of the Scottish Parliament to determine? Were voters’ views on the key areas of controversy between the parties reflected in how they voted? And might it therefore have been the unpopularity of its policy proposals that accounted for Labour’s defeat? Table 5.11 below summarises where the public stood on some of the key political issues of the day.8 In the first half of this table are those issues that lie within the competence of the Scottish Parliament to determine, while below are a few where responsibility primarily lies with the UK government. Not that this distinction is always straightforward. For example, while the UK government is responsible for making decisions about whether any new nuclear power stations should be built anywhere in the UK, the Scottish government has responsibility for making planning 90

Lost Labour Votes? decisions about any proposals to build nuclear power stations in Scotland. Thus the second of the items we have labelled as ‘primarily reserved’ (that is a responsibility of the UK government) is one where both levels of government potentially have a veto power, and the SNP and the Liberal Democrats at least wanted to use the leverage this gave the devolved institutions to stop any nuclear power stations being built in Scotland. Equally, although the Scottish Parliament is responsible for local government taxation in Scotland – and so could replace the current council tax (a tax on the value of private houses) with a local income tax (as proposed by both the SNP and the Liberal Democrats) – such a decision would mean an end to the payments made by the UK government to those on low incomes who are eligible for a welfare benefit to help them pay their council tax. If the UK government failed to compensate its Scottish counterpart for that lost revenue (as the incumbent UK Labour government said it would), then any change could only be made by increasing the overall level of local taxation raised in Scotland or increasing the level of financial support from the Scottish government to local councils. Either way, the switch would have a financial cost. Nevertheless, in most cases the difficulty of determining whether an issue is devolved or reserved does not arise. And there were a number of issues that clearly lay within the jurisdiction of the Scottish Parliament to determine for itself and which were a source of division between the parties during the 2007 election campaign. For example, as has already been noted in Chapter 2, up-front tuition fees for university students (that is fees paid while a student is studying) had already been abolished by the incumbent Labour/Liberal Democrat coalition. Now both the SNP and the Liberal Democrats wanted to abolish the so-called ‘graduate endowment’ to which students in Scotland were required to contribute after they had graduated, a policy that was opposed by both Labour and the Conservatives. Equally, both the SNP and the Liberal Democrats wanted to get rid of all prescription charges in Scotland, a policy that had already been implemented by the devolved government in Wales, while neither Labour nor the Conservatives endorsed the idea. Although neither the Conservatives nor Labour wanted to abolish the council tax, they did make proposals for reducing the amount of tax that had to be paid by older people, while it was the Conservatives together with the SNP who particularly wanted to cut the amount of local tax (known as ‘rates’) paid 91

Revolution or Evolution? by businesses. Meanwhile, as we might anticipate, the Greens were keen on introducing congestion charges for driving into city centres at peak times (a policy with which the Liberal Democrats had also toyed but eventually shied away from) while it was Labour who emphasised the need to impose a night-time curfew on young people who had caused trouble. In short, there was no shortage of devolved issues that divided the parties and which voters might reasonably have been expected to take into consideration when deciding how to vote. However, there were reserved issues that gave rise to controversy too. Shortly before the 2007 election the UK Labour government announced that it intended to replace the nuclear submarines that carry Britain’s nuclear weapons: a set of Trident missiles supplied by the United States based on the River Clyde on the west coast of Scotland. While this decision was backed by the Conservatives, it was opposed by both the SNP and the Liberal Democrats. Meanwhile, the legitimacy and effectiveness of Britain’s involvement in the 2003 invasion of Iraq continued to be the subject of argument, opposed as the war had always been by both the SNP and the Liberal Democrats. In short, there was no shortage of issues that were the responsibility of the UK Labour Government and which might have given voters reason to consider casting a protest vote against Labour in the Scottish parliamentary election. Some of the positions taken by Labour on devolved issues were more popular than others. As Table 5.11 shows, few disagreed that older people should pay less council tax. But at the same time asked to choose between a local property tax (such as the council tax that Labour along with the Conservatives wanted to retain) and a local income tax, only just over one in ten (11%) favoured a property tax. Meanwhile, as has already been seen in Chapter 2, Labour’s reluctance to abolish the graduate endowment seems to have been in tune with public opinion; over twothirds (69%) reckoned that at least some students should make a contribution towards the cost of their tuition. On the other hand, cutting business taxes, as espoused by the Conservatives and the SNP, was relatively popular too.

  92

Lost Labour Votes? Table 5.11 The Balance of Opinion on the Issues of the Campaign Agree %

Neither %

Disagree %

88 83

4 -

8 11

60

15

24

56

23

15

46

13

40

30

-

69

24

12

62

64

16

19

51

18

29

34

19

42

Primarily Devolved Issues Cut council tax for over 65s Introduce local income tax+ Curfew troublesome young people at night Cut business taxes to strengthen Scotland’s economy Abolish all prescription charges No student should have to pay tuition fees* £2 congestion charge to drive though city centres Primarily Reserved Issues Britain was wrong to go to war with Iraq No more nuclear power stations built in Scotland British government right to renew Trident Sample size =1,508

+ Question asked respondents whether it was better for local taxation to be based on people’s income or the value of their properties. The agree column shows the percentage who said income, the disagree column those who said property. * Based on answers to a question that asked people to choose between all students paying, some or none at all. Those giving one of the first two answers are counted as ‘Disagree’; this comprised 63% who said some students should pay and 6% that all should pay. Source: Scottish Social Attitudes, 2007.

In contrast, on the key reserved issues, Labour’s positions were consistently unpopular. Britain’s involvement in the Iraq war, on which public opinion in Scotland had been evenly divided at the time of the 2003 93

Revolution or Evolution? election (Curtice, 2006), was now widely unpopular. Equally, just over half (51%) opposed the idea of building new nuclear power stations in Scotland, a possibility that Labour had refused to rule out. Meanwhile, rather more people (42%) disagreed with the recently announced decision to renew Britain’s independent nuclear capability than endorsed it (34%). If voters were indeed inclined to use the Scottish election to cast a protest vote against policies being pursued by the Labour government in London, then it could well have cost Labour dear. Table 5.12 Vote by Policy Preference on Tuition Fees and Iraq Policy Position: No student should have to pay tuition fees % voted Conservative Labour Total for parties that oppose policy Liberal Democrat SNP Green/SSP/Solidarity Total for parties that support policy Sample size

Amongst those who support this policy position 13 34 47 12 36 4 52 244

Policy Position: Britain wrong to go to war in Iraq % voted Conservative Labour Total for parties that oppose policy Liberal Democrat SNP Green/SSP/Solidarity Total for parties that support policy Sample size

oppose this policy position 13 36 49 13 33 3 49 592

Amongst those who support this policy position 12 30 42 14 38 4 56 573

Source: Scottish Social Attitudes, 2007.

94

oppose this policy position 15 49 65 10 23 2 35 165

Lost Labour Votes? But did the way that people voted reflect their policy positions, either on reserved or devolved issues? Table 5.12 shows in detail how we can address that question, using the example of one devolved issue, university tuition fees, and one reserved one, the Iraq war. In the first column of the first part of the table we show what proportion of those who believed all tuition fees should be scrapped voted for each of the parties, with those parties that opposed that particular policy (which included Labour) distinguished from those that supported it. In the second column we show the same information for those that opposed getting rid of all tuition fees. If indeed voters’ views on this issue were reflected in how they voted, we should find that those who supported getting rid of tuition fees should have been more likely than those who were opposed to vote for those parties that advocated scrapping fees. This, however, is not what we find. Amongst those who supported the abolition of tuition fees, as many as 47% actually voted for parties that opposed that policy. This is hardly any different from the 49% level of support for those parties amongst those voters who did believe in retaining tuition fees. Equally, just 52% of those who supported scrapping tuition fees voted for parties that promoted that policy, little different from the level of support, 49%, amongst those voters who opposed abolishing fees. In short, on this devolved issue, at least, it seems that the way in which people voted did not reflect their views at all. The same cannot be said, however, for the reserved issue examined in the second half of the table, Britain’s involvement in the Iraq war. Nearly two-thirds (65%) of those who disagreed that Britain was wrong to have gone to war in Iraq (that is, they supported the war) voted for parties that backed the war. In contrast amongst those who agreed the war was wrong, only 42% did so. Note in particular that, at 49%, support for Labour was notably higher amongst supporters of the war than it was among the conflict’s opponents. Meanwhile, only 35% of those who supported the war voted for parties that opposed it, well short of the 56% level of support recorded for those parties amongst opponents of the war. But are these patterns typical? Table 5.13 presents summary information on the link between vote and policy preference for all of the more contentious issues initially presented in Table 5.11. It shows for both supporters and opponents of each policy the proportion who voted for one of the parties that agreed with their policy position. Given that these are all issues on which all or nearly all of the parties adopted a 95

Revolution or Evolution? position in their manifestos, if people’s attitudes on an issue were reflected in how they voted then the entry in both columns should be well above 50%. Table 5.13 The Link between Vote and Policy Preference % voted for an appropriate party Primarily Devolved Issues Cut business taxes to strengthen Scotland’s economy No student should have to pay fees Introduce local income tax Abolish all prescription charges Primarily Reserved Issues British government right to renew Trident No more nuclear power stations built in Scotland Britain was wrong to go to war with Iraq

Amongst those who support policy oppose policy 59

64

52 49 46

49 56 48

60

62

57

61

56

65

For sample sizes see Appendix to this chapter. Source: Scottish Social Attitudes, 2007.

This is true of only one of the primarily devolved issues in our table. Around three in five people voted for a party whose policy position was consistent with their attitude towards cutting business rates. For example, support for the SNP, who favoured the policy, was almost twice as high (42%) amongst those who agreed with the policy as it was amongst those who opposed it (22%). But on the other three devolved issues, the pattern we have already seen in the case of tuition fees is replicated, with only around half supporting a party that agreed with their view on a particular policy. In contrast, people’s attitudes towards those issues that we have labelled as primarily reserved were, to some degree at least, reflected in how they voted. Typically around three-fifths of voters voted for a party that upheld their view on a reserved issue. Note, however, that so far as Labour in particular was concerned, the differences in its level of support 96

Lost Labour Votes? between supporters and opponents of Trident and nuclear power stations were rather less than the 19 point gap we saw previously in the case of the Iraq war. Thus, supporters of Trident were only 11 points more likely than opponents to vote Labour, while in the case of nuclear power stations the equivalent gap was just six points. These were issues that seemed to make more of a difference to the chances of someone voting Conservative on one side of the argument, or the SNP on the other, than it did to Labour.9 So it seems that the outcome of the 2007 election was not for the most part a reflection of the popularity or otherwise of the policy positions put forward by the parties on what should be done by the Scottish Parliament over the next four years. Rather, reserved issues such as Trident and Iraq, over which the devolved institutions do not have any say at all, seem, if anything, to have had more resonance for voters. Not only is it not clear that devolved elections help hold the Scottish government accountable to the people of Scotland, but they also do not necessarily ensure that the views of parliamentarians reflect those of the Scottish public. Meanwhile, we have seen that whereas some of Labour’s positions on devolved issues were popular (though this was not true of the one devolved issue, business taxes, that did seem to make a difference to how people voted), its positions on all of the reserved issues that we considered were relatively unpopular. It seems as though the stance taken by the UK Labour government on those issues, and especially on Iraq, certainly did not do the Scottish party any favours in the 2007 Scottish parliamentary election. At least some of the blame for the party’s defeat seems to lie with decisions taken by the party in London.10 CONCLUSION We have pursued two interwoven themes in this chapter. First we have asked whether voters used the election to hold the Scottish government to account or to express their views about some of the principal policy issues facing the Scottish Parliament. Second, we have considered the implications of our findings for Labour’s performance in the election in particular. The experience of the 2007 parliamentary election casts considerable doubt on the claim that devolved elections necessarily help ensure the 97

Revolution or Evolution? government of Scotland is both accountable and representative. True, we have found some evidence that voters who were unhappy with Labour’s record in office or its leadership were less likely to vote for the party. But we also uncovered at least as much evidence – if not indeed more – that in so doing they were expressing their opinion more about the performance of the UK Labour government and its leadership than they were that of the Labour-led Scottish Executive and its leader. Meanwhile, it seems that in general the ways in which people voted in the election were more likely to reflect their views about those policy issues that were primarily the responsibility of Westminster rather than the Scottish Parliament to decide. Inevitably these findings would seem to suggest that the source of Labour’s electoral difficulties in 2007 lay more at Westminster than in Edinburgh. We should, though, bear in mind that although some voters may have been dissuaded from voting Labour by their evaluations of the UK Labour government or of the Prime Minister, Tony Blair, neither was on balance viewed negatively. Only in the case of some of the policies pursued by the UK government, most notably the Iraq war, does the evidence clearly suggest that Labour lost out because of dissatisfaction with the party at Westminster. Moreover, it is far from clear that either the Labour-led Scottish Executive as whole, or its leader Jack McConnell in particular, together with the devolved policy positions that the party adopted in 2007, generated much enthusiasm amongst the public. And while in 2003 this may not have mattered, in 2007 the party was facing a far more popular opponent in the SNP leader, Alex Salmond. Labour may not so much have lost the election as the SNP won it. In any event we now clearly need to examine what did indeed encourage people to vote for the SNP. NOTES 1. This includes those who said they did not know whether recent trends had been favourable or unfavourable. 2. As we noted earlier in this chapter, it is possible that in the case of some respondents their evaluations of Labour’s performance was a consequence of how they decided to vote rather than a cause. If so, this would mean that our analysis overestimates the impact of evaluations 98

Lost Labour Votes? of Labour’s performance on how people voted. However, our primary interest here is in comparing the relative importance of evaluations of Labour’s performance at the two levels of government. This will not be affected if we can assume that respondents were as likely to align their evaluations of the Scottish Executive with their vote as they were their evaluations of the UK government (and vice-versa). 3. As indicated by those who said that they would have voted Labour if the parliamentary election ballot paper had required them to express a single first-preference vote. Note that those who said they did not vote in 2007 are included in the denominator. Our main findings remain the same if non-voters are excluded from our analyses. 4. The equivalent analysis for 2003 Liberal Democrat voters shows much the same result. Amongst such voters who thought the Liberal Democrat ministers on the Scottish Executive had done a good job, 52% supported the party again in 2007. The equivalent figure for those who did not think Liberal Democrat ministers had done a good job was 40%, a gap of 12 percentage points. 5. This is not an unimportant consideration when reflecting upon how well or badly Labour performed in 2007. It might be pointed out that as the party’s share of the constituency vote only fell by 2.2% between 2003 and 2007, and hardly fell at all on the regional list vote, that it is mistaken to be focusing an analysis on why Labour lost votes. However, the scale of the party’s losses only looks so small because of the eight-point decline in support for ‘other’ parties on both ballots. 6. This was achieved by including an interaction term between thinking the Scottish Executive was either primarily or jointly responsible for recent trends in the NHS and evaluations of those trends. 7. So far as the Conservative and Liberal Democrat leaders were concerned, many voters were unable to give them a mark at all. This was true of no less than 36% in the case of Annabel Goldie, the Conservative leader, and no less than 56% in respect of the Liberal Democrat leader, Nicol Stephen. Just 15% gave Miss Goldie a good score, 17% a bad one, while the equivalent figures for Mr Stephen were only 7% and 10% respectively. Given their low and uncontroversial profile no further consideration is given here to ascertaining what impact evaluations of these two leaders had on party support. 99

Revolution or Evolution? 8. Attitudes towards devolution and independence have already been presented in Chapter 4. How far these attitudes were reflected in people’s voting behaviour is considered in the next chapter where we discuss the role of the issue in the success of the SNP. 9. Amongst supporters of Trident, 21% voted Conservative, compared with just 8% of opponents. Equally 21% of those who believed new nuclear power stations should be built in Scotland voted Conservative, but only 8% of this policy’s opponents. As many as 43% of those who opposed Trident backed the SNP, but only 26% of those who supported it; 41% of those who did not favour more nuclear power stations voted SNP, compared with 23% who favoured building more nuclear stations. 10. If we undertake a multivariate analysis of the same kind that we undertook earlier in the chapter, we find that the issues that were significantly associated with switching to or away from Labour were, in order of importance, Iraq, business taxes and nuclear power stations. REFERENCES Anderson, C. (2006), ‘Economic Voting and Multi-Level Governance: A comparative individual level analysis’, American Journal of Political Science, 50: 449–63. Anderson, C. (2007), ‘The End of Economic Voting? Contingency Dilemmas and the limits of democratic accountability’, Annual Review of Political Science, 10: 271–96. Bartels, L. (1996), ‘Uninformed Votes: Information effects in presidential elections’, American Journal of Political Science, 40: 194–230. Bean, C. and Mughan, A. (1989), ‘Leadership Effects in Parliamentary Elections in Australia and Britain’, American Political Science Review, 83: 1165–79. Bromley, C. (2006), ‘Holyrood 2003 – Where were the Voters?’, in Bromley, C., Curtice, J., McCrone D., and Park, A. (eds), Has Devolution Delivered?, Edinburgh: Edinburgh University Press. Butler, D. and Stokes, D. (1974), Political Change in Britain, 2nd edn, London: Macmillan. Clarke, H., Jenson, J., LeDuc, L. and Pammett, J. (1979), Political Choice in Canada, Toronto: McGraw-Hill, Ryerson. 100

Lost Labour Votes? Clarke, H., Sanders, D., Stewart, M, and Whiteley, P. (2004), Political Choice in Britain, Oxford: Oxford University Press. Curtice, J. (2006), ‘Is Holyrood Accountable and Representative?’, in Bromley, C., Curtice, J., McCrone D., and Park, A. (eds), Has Devolution Delivered?, Edinburgh: Edinburgh University Press. Curtice, J. (2009), ‘At the Ballot Box’, in Curtice, J. and Seyd, B. (eds), Has Devolution Worked?, Manchester: Manchester University Press. Delli Carpini, M. and Keeter, S. (1996), What Americans Know about Politics and Why it Matters, New Haven, CT: Yale University Press. Evans, G. and Andersen, R., (2006), ‘The Political Conditioning of Economic Perceptions’, Journal of Politics, 68: 194–207. Gélineau, F. and Bélanger, É. (2005). ‘Electoral Accountability in a Federal System: National and provincial economic voting in Canada’, Publius: The Journal of Federalism 35: 407–24. Given, L. and Ormston, R. (2006), Scottish Social Attitudes Survey 2005 – Core Module Report 1: Attitudes towards public services in Scotland, Edinburgh: Scottish Executive Social Research. Glaser, T. and Salmon, C. (1991), Public Opinion and the Communication of Consent, New York: Guilford. Graetz, B. and McAllister, I. (1987), ‘Popular Evaluations of Party Leaders in the Anglo-American Democracies’, in Clarke, H. and Czudnowski, H. (eds), Political Elites in Anglo-American Democracies, Dekalb, IL: Northern Illinois University Press. Katz, R. (1997), Democracy and Elections, New York: Oxford University Press. Kiewit, D., (1983), Macroeconomics and Micropolitics: The electoral effects of economic issues, Chicago: University of Chicago Press. Milner, H. (2002), Civic Literacy: How Informed Citizens Make Democracy Work, Hanover, NH: University Press of New England. Mughan, A. (1995), ‘Television and Presidentialism: Australian and US legislative elections compared’, Political Communication, 12: 327–42. Mughan, A, (2000), Media and the Presidentialization of Parliamentary Elections, Basingstoke: Palgrave. Nadeau, R. Niemi, R. and Yoshinaka, A. (2002), ‘A Cross-National Analysis of Economic Voting: Taking account of the political context across time and nations’, Electoral Studies, 21: 403–23. Ormston, R. and Sharp, C. (2007), Scottish Social Attitudes Survey 2006 – Core Module Report 1: Attitudes towards public services in Scotland, Edinburgh: Scottish Executive Social Research. 101

Revolution or Evolution? Page, B. and Shapiro, R. (1992), The Rational Public, Chicago: University of Chicago Press. Park A. and McCrone, D. (2006), ‘The Devolution Conundrum’, in Bromley, C., Curtice, J., McCrone, D. and Park, A. (eds), Has Devolution Delivered?, Edinburgh: Edinburgh University Press. Popkin, S. (1991), The Reasoning Voter: Communication and persuasion in presidential campaigns, Chicago: University of Chicago Press. Powell, G. Bingham, Jr (2000), Elections as Instruments of Democracy: Majoritarian and Proportional Visions, New Haven: Yale University Press. Powell, G. Bingham Jr. and Whitten, G. (1993), ‘A Cross-National Analysis of Economic Voting: Taking account of the political context’, American Journal of Political Science, 37: 391–414. Reif, K. (1984), ‘National Electoral Cycles and European Elections 1979 and 1984’, Electoral Studies, 3: 244–55. Reif, K. and Schmitt, H. (1980), ‘Nine Second-Order National Election Results: A conceptual framework for the analysis of European election results’, European Journal of Political Research, 8: 3–44. Schumpeter, J. (1976), Capitalism, Socialism and Democracy, 4th edn, London: Allen and Unwin. Stewart, M. and Clarke, H. (1992), ‘The (Un)importance of Party Leaders: Leader images and party choice in the 1987 British election’, Journal of Politics, 54: 447–70. Whitten, G. and Palmer, H. (1999), ‘Cross-National Analyses of Economic Voting’, Electoral Studies, 18: 49–67.

102

Lost Labour Votes? APPENDIX TO CHAPTER 5: SAMPLE SIZES FOR TABLE 5.13 The sample sizes on which the figures in Table 5.13 are based are shown below: Amongst those who support oppose policy policy Primarily Devolved Issues Cut business taxes to strengthen Scotland’s economy No student should have to pay fees Introduce local income tax Abolish all prescription charges Primarily Reserved Issues British government right to renew Trident No more nuclear power stations built in Scotland Britain was wrong to go to war with Iraq

103

472

136

244 717 375

592 99 392

272

391

426

264

573

165

CHAPTER 6 __________________________________________

HOW THE SNP WON

We appear to be left with a considerable puzzle. There was no surge of support for independence in 2007, no sudden swelling of discontent with devolution, and no new rejection of Britishness. Those who were unhappy with Labour’s performance in office were less likely to remain loyal to the party, but neither the party nor its leadership was markedly less popular than it had been four years previously. It seems that not only was the SNP’s victory historic, but it was achieved in a climate that was not particularly propitious for the party at all. So how could it have been possible? One possibility, perhaps, is that the party was able to persuade those who had previously stayed at home to back the party this time. The party has previously had some apparent difficulty at getting those who claimed to support it to go to the polls (Paterson, 2006). Maybe, too, the nationalists profited from the large fall in support for the smaller parties, most notably the Scottish Socialists? There were signs in 2003 that those whose discontent with Labour might have led them to back the SNP opted instead to vote for the socialists (Curtice, 2004). Now that the socialists had split into two – the Scottish Socialists and Solidarity – perhaps voting for the nationalists looked like the only effective way of voting against Labour. A second possibility, however, is that there was a change in the underpinnings of the SNP vote. Some social groups, most notably women, older people and Catholics have often been relatively reluctant to support the party (Paterson, 2006; Seawright and Curtice, 1995). Maybe that reluctance was finally overcome? But perhaps more importantly we should bear in mind that prior to 2007 those who supported independence, or who said they were ‘Scottish not British’, did not necessarily support the SNP (Bond, 2000; Bond and Rosie, 2002; McCrone, 2001; Paterson, 2006; Rosie and Bond, 2003). True, supporters of independence and those who rejected a British identity were more likely to back the party, but they were far from being unanimous in their support. Potentially, therefore, the SNP could advance not by  

104

How the SNP Won persuading more people of the merits of independence or by undermining their sense of Britishness, but rather by persuading more of those who already backed independence or who already did not feel British to support the party. In other words perhaps this time around the SNP were more successful in securing the support of those who might be thought to form its ‘natural’ constituency. In this chapter we consider each of these possibilities in turn. First, we consider how those who voted for the SNP in 2007 behaved in the previous Scottish election in 2003. Second, we compare the demographic profile of the nationalist vote in 2007 with its character four years previously. Finally, we assess the link between national identity, constitutional preference and voting for the SNP and how it has varied between elections. WHO SWITCHED TO THE SNP? We can in fact dismiss our first suggestion fairly easily. Those who (in our 2007 survey) said they did not vote in the 2003 election were least likely of all to switch to the SNP in 2007. Just 8% of those who recalled abstaining in 2003 turned out in 2007 and regarded the nationalists as their first preference party. Even those who voted for the Conservatives on the constituency vote in 2003 were no less likely to switch to the SNP this time around; nine per cent of this group did so. However, the SNP were relatively successful at attracting the support of those who backed one of the smaller parties in 2003; no less than one in five (20%) of this group switched to the nationalists this time around (see also Denver, 2007: 66). Indeed this group was more likely to switch to the SNP than any other, exceeding the 17% rate amongst 2003 Liberal Democrat supporters and the 12% figure for former Labour supporters. But evidently all three groups were more likely to switch to the SNP than those who had abstained. We should not, though, exaggerate the role that gaining votes from those who said they supported a smaller party in 2003 played in the SNP’s success. We have already seen (see for example Table 5.5) that Labour also profited from the decline in support for the smaller parties between 2003 and 2007. Moreover, those who voted for one of the smaller parties in 2003 only comprise 3% of all 2007 SNP voters in our survey. Those 105

Revolution or Evolution? who abstained in 2003, in contrast, constituted a much larger group and as a result, despite their relative reluctance to switch to the nationalists, they still accounted for no less than 14% of all SNP voters in 2007. In fact when we look at where the 2007 SNP vote came from we discover that the single biggest source of new nationalist voters in 2007 comprised those who voted Labour in 2003. No less than 21% of all SNP support came from this group. (Former Liberal Democrat and Conservative voters, in contrast, accounted for just eight and three per cent respectively.) Winning votes directly from the party’s principal competitor, Labour, was therefore an important component of the SNP’s success. KINDS OF VOTERS Our second suggestion was that perhaps the SNP was notably more successful in 2007 at winning support amongst those sections of Scottish society that had hitherto been relatively reluctant to back the party. To see whether or not that proved to be the case, Table 6.1 shows the degree to which those in different groups backed the party in 2007, and how this compared with the position at the previous Scottish election in 2003. In practice the kind of person who backed the SNP in 2007 was much the same as the kind of person who did so in 2003. Those in the oldest age group were least likely to have voted for the party in 2003; the same was true in 2007. Meanwhile there was little difference at all between the remaining age groups. Catholics remained somewhat less likely to vote for the party, as was the case in 2003. As it happens, our 2003 survey did not uncover any difference between men and women in their willingness to vote for the party, but in 2007 the traditionally higher level of nationalist support amongst men was somewhat in evidence once again. Meanwhile, employers and managers were a little less likely to vote for the party than those in routine, ‘working-class’ occupations, but this gap too was much the same as it was in 2003.

106

How the SNP Won Table 6.1 Demographic Profile of SNP Support, 2003 and 2007 % vote SNP amongst: Gender Men Women Age group 18–34 35–44 45–54 55-64 65+ Religion Protestant Catholic None Social class Employers and managers Routine and semi-routine workers

2003

2007

23 23

36 32

19 34 22 25 18

38 38 36 37 26

21 17 28

29 24 40

19 25

32 39

Social Class based on the NS-SEC classification scheme (see Appendix to this book for further details)). It is determined by the respondents’ occupation unless the respondent is economically inactive (other than retired) and has a partner who is economically active (or retired) in which case it is determined by partner’s occupation. Not all classes are shown. Those adhering to a non-Christian faith are not shown. Percentages are based on those who reported having voted. For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 2003, 2007.

Of course, what is also striking in Table 6.1 is the fact that such differences that do exist are mostly relatively small ones. There is nothing, for example, to match the large difference between the level of support for Labour amongst Catholics (61% of whom backed the party in 2007) and that amongst Protestants (33%) and those of no religion (32%). Socially, at least what continued to be the most striking feature of SNP support was its relatively even spread (see also Denver, 2007: 73–4).

107

Revolution or Evolution? NATIONAL IDENTITY, CONSTITUTIONAL PREFERENCE AND SNP SUPPORT We are left then with one final possible explanation of the SNP’s success – a strengthening of the link between SNP support and both national identity and constitutional preference. Table 6.2 gives us an initial glimpse of this possibility by showing the proportion of those with a given national identity (as measured by the ‘Moreno’ question introduced in Chapter 2) who voted for the SNP at each election since the first Scottish Parliament election in 1999. We will consider what it reveals in two parts. First we will compare what happened in the two Scottish Parliament elections of 1999 and 2003 with what occurred in the two UK general elections of 2001 and 2005. Thereafter we will examine how the pattern of SNP support in 2007 compares with what happened in 2003. Table 6.2 National Identity and Nationalist Support, 1999–2007 % voted SNP amongst those whose national identity is: Scottish not British More Scottish than British Equally Scottish and British More British than Scottish British not Scottish Gap

1999 2001 2003 2005 2007

Change 2003–7

43 27

24 16

40 27

25 19

58 42

+18 +15

18

8

6

9

15

+9

6 6 37

9 3 21

2 1 39

4 2 23

11 7 51

+9 +6

Gap: Difference between the percentage of those saying they are ‘British not Scottish’ voting SNP and the percentage of those saying they are ‘British not Scottish’ doing so. Non-voters excluded throughout. Details of the sample sizes on which each entry is based are to be found in the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

There is a striking contrast between the top and bottom halves of the table in the period between 1999 and 2005. In the lower half, that is amongst those who only feel British, or who at least feel British more than they do Scottish, the level of support is not only low but varies little from one election to the next. In contrast, in the top half of the table, that is 108

How the SNP Won amongst those who only feel Scottish or who feel Scottish more than they do British, there is considerable volatility, and especially so in the top row. Moreover, there is a distinct pattern to this volatility. The SNP did particularly well amongst those with a relatively strong Scottish identity in the two Scottish Parliament elections, and performed less well in the two UK contests. As a result the difference between the level of SNP support amongst those who only feel Scottish and that amongst those who only feel British was much greater in the two Scottish elections than it was in the two UK contests. So, it seems that those with a relatively strong Scottish identity were already more inclined to support the nationalists in Scottish Parliament elections even before the 2007 contest was under way. Here, perhaps, is a clue as to why, as we noted in Chapter 1, the SNP have consistently performed better in Scottish Parliament elections than they have in UK contests (Bromley, 2006).1 But then we can see that in 2007 the link between SNP support and national identity became stronger still. For the first time the SNP actually won a majority of the votes of those who said they were Scottish and not British. In contrast, support amongst those who said they were only British was no higher than it had been in 1999. Overall we can see from the last column in the table that the more Scottish and less British someone felt, the more the nationalist vote increased compared with 2003. The 2007 election witnessed, then, a marked strengthening of the link between national identity and support for the SNP. The party succeeded in capturing the support of those with a predominantly Scottish sense of identity in a hitherto unprecedented manner. It had already appeared to be the case that Scottish Parliament elections provided an environment in which the party was particularly able to secure the support of this group. In 2007, however, that apparent potential was realised to a much greater extent than before. But what of the link between nationalist support and people’s constitutional preferences? Do we find that those who support independence were particularly more inclined to support the SNP in Scottish Parliament elections? And was this particularly the case in 2007? Table 6.3 suggests that the answer to both questions is an affirmative one. Prior to 2007 support for the SNP varied most amongst those who supported independence, and barely at all amongst those who did not want any kind of Scottish parliament. Meanwhile between 2003 and 2007 109

Revolution or Evolution? SNP support rose most strongly amongst supporters of independence, less so amongst those who backed devolution, and hardly at all amongst those opposing any kind of parliament. Over three-quarters of those who support independence supported the SNP in 2007, making the link between constitutional preference and nationalist support a strong one by any standards. Table 6.3 Constitutional Preference and Nationalist Support, 1999–2007 % voted SNP amongst those favouring: Independence Devolution No parliament Gap

1999

2001

2003

2005

62 17 5 57

35 19 3 32

58 12 4 54

33 9 3 30

Change 2003–7 78 +20 22 +10 6 +2 72

2007

Gap: Difference between the percentage of those favouring independence who voted SNP and the percentage of those opposed to any kind of parliament doing so. Non-voters excluded throughout. Details of the sample sizes on which each entry is based are to be found in the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

In fact we can see much the same pattern if we look at other measures of constitutional preference.2 Table 6.4, for example, shows how SNP support has varied according to people’s attitudes towards strengthening the powers of the Scottish Parliament. As we would expect, the more strongly that someone agrees that those powers should be increased, the more likely they are to support the SNP. But this link has always been stronger in Scottish Parliament elections than it has been in UK elections, and it was particularly strong in 2007. It seems the SNP was not just particularly successful in 2007 in securing the support of those who favoured independence, but also the votes of those who wanted a stronger Scottish Parliament while remaining within the framework of the Union.3

110

How the SNP Won Table 6.4 Attitudes towards More Powers and Nationalist Support, 1999–2007 % voted SNP amongst those whose attitudes towards more powers is: Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree Gap

1999 2001 2003 2005 2007

Change 2003–7

58 33

29 17

54 27

34 18

67 34

+13 +13

19

7

12

9

15

+3

6 3 55

3 5 24

6 4 50

3 0 34

8 0 67

+2 -4

Gap: Difference between the percentage of those who strongly agree the Scottish Parliament should have more powers voting SNP and the percentage of those who strongly disagree doing so. Non-voters excluded throughout. Details of the sample sizes on which each entry is based are to be found in the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

So it seems that we may have found the vital ingredient of the SNP’s success in 2007. Previously many who might have been thought to be ‘natural’ SNP supporters had failed to back the party. But the introduction of Scottish Parliament elections created an environment in which it seems those ‘natural’ supporters were consistently more likely to give the party their vote, while that environment was exploited with particular success in 2007. Evidently, what we have to establish now is, first, why the link between national support and both national identity and constitutional preference always seems to be stronger in Scottish Parliament elections and, second, to explain why those links were particularly strong in 2007. WHAT MAKES SCOTTISH PARLIAMENT ELECTIONS DIFFERENT? There is one obvious difference between Scottish and UK elections. The former are about what is happening in Scotland alone; the latter about developments across the whole of the UK. So the electorate in Scotland are invited to view the world through a different lens in Scottish elections from the one they use in a UK election. They are invited to ask who 111

Revolution or Evolution? would provide the best government for Scotland in particular, rather than who might be best for the UK as a whole. We should perhaps not be surprised that, asked a different question, the electorate provide a different answer, and that a party like the SNP whose focus and concern is with Scotland alone (and which indeed is not a contender for government in a UK election) might be more popular when people are asked who might be best for Scotland in particular. Indeed, Scotland’s experience in this regard is replicated elsewhere where nationalist parties fight ‘devolved’ elections (Jeffrey and Hough, 2003). The SNP are certainly more likely than any of the other parties to be regarded as a party that closely looks after the interests of the people of Scotland. In our 2007 survey, for example, as many as 23% reckoned the SNP looked after those interests ‘very closely’, whereas just 7% said the same of the ‘Labour Party in Scotland’, 4% the Liberal Democrats and only 3% the Conservatives. The pattern was much the same when the question was asked in previous years.4 So if in devolved elections voters’ attention is focused more on who they think would provide the best government for Scotland, then there is good reason to anticipate that this might increase their propensity to back the SNP. Meanwhile voters do indeed seem to focus their attention in Scottish elections on what is going on in Scotland in particular. After each election held since 1999 the Scottish Social Attitudes survey has asked those who voted whether they had decided how to vote ‘mostly according to what was going on in Scotland’ or ‘mostly according to what was going on in Britain as a whole’. As Table 6.5 shows, in Scottish elections typically just over half say they have voted according to what was happening in Scotland while no more than a third say they considered what was going on throughout the whole of Britain. In contrast, in UK elections only around a third say they voted on the basis of what was going on in Scotland, while the percentage who indicate they voted according to what was happening across Britain as a whole is about ten points higher. In short, many voters do view the two types of contest through different lenses.

112

How the SNP Won Table 6.5 What Is Uppermost in Voters’ Minds? 1999 %

2001 %

2003 %

2005 %

2007 %

What mostly going on in Scotland 52 Britain as a whole 34 Sample size 1,032

34 44 1,084

54 27 908

32 43 1,079

56 29 890

Table based only on those claiming to have voted. Source: Scottish Social Attitudes, 1999–2007.

Moreover, we can also demonstrate that those who say they focused on what is happening in Scotland in particular are more likely to support the SNP. In 2003, for example, 33% of those who said they voted according to what was happening in Scotland backed the nationalists; in contrast, only 6% of those focusing on developments across Britain did so. The figures were much the same in 2005; as many as 32% of those focusing on Scotland in particular supported the SNP, compared with just 4% of those looking at developments across the whole of Britain. The lower level of support for the SNP in the 2005 UK election as compared with the devolved one in 2003 can thus be directly related to the lower proportion of people voting on the basis of what was happening in Scotland in particular. The different perspective that voters bring to different kinds of elections also helps to account for the weaker relationship between constitutional preference and vote choice in UK elections. Those who support independence are particularly more likely to back the SNP if they are primarily interested in developments in Scotland. In 2003, for example, no less than 66% of those supporters of independence who voted on the basis of what was going on in Scotland supported the SNP; only 14% of those focusing on the position across the whole of Britain did so. In 2005 the equivalent figures were 47% and 13% respectively. In contrast, for those supporting devolution or opposing any kind of parliament at all, the focus of their attention made rather less of a difference to their chances of voting SNP.5 Evidently devolved elections consistently provide the SNP with an environment in which it is particularly able to garner the support of those with a relatively strong sense of Scottish identity and thereby obtain a higher overall vote than it would do in a contemporaneous UK election. 113

Revolution or Evolution? WHAT HAPPENED IN 2007? But we still have to explain why the SNP did so well in 2007, and especially so amongst those who preferred independence or had a Scottish identity. Here it is useful to recall a couple of findings in the previous chapter. First, we noted that Alex Salmond was far more popular than his predecessor, John Swinney, had been in 2003. Second, those who thought that Mr Salmond would make a good First Minister were less likely to support Labour irrespective of what they thought of Tony Blair or Jack McConnell. Perhaps the particular success of the SNP amongst its ‘natural’ supporters in 2007 lay in the perception amongst this group that the party could provide good government, as exemplified by the popularity of its candidate for First Minister (Clarke et al., 2004). Table 6.6 provides a first vital piece of evidence in considering this possibility. It compares the popularity of Mr Swinney in 2003 with that of Mr Salmond in 2007 separately for supporters of independence, devolution and no parliament at all. The pattern it unveils is striking. The difference between the popularity of Mr Salmond and that of Mr Swinney was far greater amongst those favouring independence than it was amongst those who preferred devolution and, to an even greater extent, those who preferred no parliament at all. In short, it looks as though the reason why the SNP were so successful amongst supporters of independence in 2007 is that it was this group that was most convinced of Mr Salmond’s ability, and perhaps by implication that of his party too, to provide effective government. Table 6.6 Constitutional Preference and Evaluations of SNP Leaders, 2003–7 % giving SNP leader score of seven or more amongst those who support Independence Devolution No parliament

2003

2007

Change 2003–7

14 6 3

62 35 18

+48 +29 +15

For details of the sample sizes on which the entries to this table are based see the Appendix to this chapter. Source Scottish Social Attitudes, 2003, 2007.

114

How the SNP Won Much the same pattern is observed if we compare the popularity of Mr Salmond and Mr Swinney according to people’s attitudes towards the parliament having more powers. Amongst those who strongly agreed that the Scottish Parliament should have more powers, the proportion giving the SNP leader a mark of seven or more out of ten was no less than 46 points higher in 2007 than it was in 2003. Amongst those who just agreed with the proposition, the equivalent figure was 32 points. In contrast the comparable figure amongst those who disagreed was 18 points and those who strongly disagreed 13 points. In short, it was amongst those who wanted a stronger parliament that Mr Salmond made the biggest impression. Equally, Mr Salmond was also relatively popular amongst those who felt more of an attachment to Scotland than to Britain. The proportion giving him a good score was no less than 40 points higher amongst those who said they were Scottish and not British. It was only half that amongst those who stated they were only or predominantly British. There seems little doubt that the SNP was able to galvanise its ‘natural’ constituency by persuading its members in particular that the party could now provide the country with effective leadership. A SECOND LOOK So far in this chapter we have compared the results of separate surveys conducted in different years to demonstrate that SNP support was unusually high in 2007 amongst those who supported independence and feel Scottish more than they do British. But perhaps there is something unusual about the sample of respondents we interviewed in 2007. Perhaps, for example, the people in favour of independence in the sample of people interviewed that year consisted disproportionately of ‘hard-core’ nationalists who have always been strongly committed to the SNP. To assuage such doubts and to confirm or otherwise the robustness of our findings we can analyse the reported pattern of switching between 2003 and 2007 amongst our sample of 2007 respondents. For if indeed the SNP particularly gained ground amongst those with a strong sense of Scottish identity and those in favour of independence and/or a more powerful parliament, then we should be able to demonstrate that those with such an identity and constitutional preferences were more likely to have switched to the SNP since 2003. 115

Revolution or Evolution? Table 6.7 presents the results of multivariate analyses of voting for the SNP that are similar in form to the analyses we conducted in the previous chapter (see, for example, Table 5.5). Our first step in the analysis is to include how people said they voted in the 2003 Scottish Parliament election. (Note that all of the coefficients are positive, albeit not necessarily significant. This reflects the fact, as we noted earlier, that those who abstained in 2003 were relatively unlikely to switch to the nationalists in 2007.) This means that any other variable that enters the model and proves to be statistically significant has to be linked to the pattern of switching between 2003 and 2007. There are three sets of analyses in our table, but let us first of all confine our attention to the one on the left-hand side. Here we have added to our analysis measures of respondent’s constitutional preference, including their response to the possibility of the parliament having more powers as well their preference for independence, devolution or no parliament at all, together with their position on the Moreno national identity scale. We have scored the ‘more powers’ measure so that the higher a respondent’s score, the more strongly they favour the parliament having more powers. Equally, the higher someone’s score on the national identity scale, the closer they are to feeling Scottish and not British. If those with a strong sense of Scottish identity and those who want a more powerful parliament switched to the SNP between 2003 and 2007 then we should find that these measures are statistically significant in this first analysis. This proves to be true of all three measures. Those who support independence were significantly more likely to have switched to the SNP between 2003 and 2007 than were those who preferred no parliament at all. Equally the more strongly someone favoured more powers for the parliament, the more likely they were to have been won round by the SNP in 2007. Meanwhile, in addition, the more that someone felt Scottish rather than British, the more likely they were to back the SNP in 2007 after not having done so in 2003. In short, our claim that the SNP did well in 2007 by winning more support amongst its ‘natural’ constituency in particular is further supported by this alternative mode of analysis.

116

How the SNP Won Table 6.7 Logistic Regression of SNP Support, 2007 Voted SNP 2007 v. Did not Model (1) (2) (3) 2003 recall vote Conservative .87 Labour .55 Liberal Democrat 1.32 SNP 2.84 Other 1.06 (Abstain) Constitutional preference Independence 1.21 Devolution .62 (No parliament) More powers Scottish identity Evaluation Salmond Evaluation Blair Iraq war wrong Nagelkerke R2 Sample size

(.48) (.26)* (.34)* (.26)* (.55) (.61)* (.57)

.40 (.13)* .35 (.11)* 33% 1,085

.79 .54 1.14 2.46 1.04

(.49) (.27)* (.35)* (.27)* (.57)

.41 .81 1.05 2.05 .58

(.50) (.28)* (.36)* (.28)* (.59)

.97 (.61) .52 (.57)

.84 (.63) .59 (.59)

.26 (.13)* .32 (.11)* .32 (.06)* 44% 1,084

.21 (.14) .30 (.11)* .37 (.06)* .20 (.04)* .19 (.09)* 48% 1,083

Main cell entries are parameter coefficients; entries in brackets are associated standard errors. 2003 recall vote and constitutional preference entered as categorical variables. The coefficients for these variables are simple contrast coefficients that show the impact on the probability of voting SNP of being in the relevant category rather than in the baseline category. The baseline category is indicated in brackets. All other variables entered as interval level variables. Scottish identity is Moreno national identity, with ‘Scottish not British’ given the highest score. Evaluation of Salmond and Blair is mark out of ten, with those who said ‘don’t know’ recoded as 5. Iraq war wrong is five-point agree–disagree scale, with ‘strongly agree’ given the highest score. Those unable to say whether the Iraq war was wrong or not were recoded as ‘neither agree nor disagree’. * Significantly different from zero at the 5% level. Source: Scottish Social Attitudes, 2007.

117

Revolution or Evolution? The second analysis in our table, meanwhile, lends further weight to our suggestion that the reason why the SNP did so well amongst this group was because it was persuaded above all of the merits of Mr Salmond. In this analysis we have included an additional variable, the mark out of ten that respondents gave Mr Salmond. Note, first of all, that the positive coefficient for this measure confirms our presumption that those who thought highly of the SNP leader were particularly likely to switch to the SNP in 2007. But of equal interest is what happens to the other variables in the model. For if we are correct in arguing the SNP gained support in particular from those who have a strong sense of Scottish identity and those who want a stronger parliament because these groups were particularly likely to be impressed by Mr Salmond, the link between voting SNP and both constitutional preference and national identity should be weaker than it was in our first analysis. Of this there is a clear sign. The coefficient for those in favour of independence is no longer significant while that for strength of support for more powers only just remains significant. On the other hand, the coefficient for Scottish identity is relatively little affected by the inclusion of the rating given to Mr Salmond, so perhaps the intensification of support amongst those with a strong sense of Scottish identity was less clearly a consequence of Mr Salmond’s popularity. We can also use our multivariate analysis to check whether there are reasons other than those we have considered so far why some groups of voters were particularly likely to switch to the SNP in 2007. For example, were those in favour of certain policy positions particular likely to switch to the SNP? Did the SNP benefit from those who defected from Labour because they disliked Tony Blair or Jack McConnell or because they disapproved of the party’s record in office? The final right-hand column of Table 6.7 summarises what we uncovered when we undertook analyses designed to address these questions. In this case we have only added to the model those measures that proved to be statistically significant. This procedure suggests there were two further ingredients to the SNP’s success in 2007. First those who gave the Prime Minister, Tony Blair, a low mark were particularly likely to switch to the SNP. So also were those who opposed the Iraq war. Both findings are worth examining further. We perhaps might not be surprised that those who thought little of the Prime Minister’s performance were particularly likely to have switched to 118

How the SNP Won the SNP. If voters are moved in Scottish elections by their evaluations of the performance of the UK government and leaders – and as we have already seen in the previous chapter, they can be – then we might well anticipate that switching to the SNP would always be likely to be linked to evaluations of the Prime Minister. But our finding is not in fact simply a routine one. This becomes apparent in Table 6.8, which compares the link between SNP support and evaluations of Tony Blair in 2007 with that in 2003. Table 6.8 Support for SNP and Evaluations of Tony Blair, 2003–7 % vote SNP amongst those whose evaluation of Blair was: Good (7–10) Middle (4–6) Bad (0–3) Gap

2003

2007

15 27 35 20

19 39 49 30

Change 2003–7 +4 +12 +14

Gap: Difference between the percentage of those who gave Mr Blair a bad mark voting SNP and the percentage of those who gave him a good mark doing so. Non-voters excluded throughout. Details of the sample sizes on which each entry is based are to be found in the Appendix to this chapter. Source: Scottish Social Attitudes, 2003, 2007.

Once again we have evidence of a link that became stronger in 2007. The difference between the level of SNP support registered amongst those who thought little of Mr Blair and those who regarded him highly was ten percentage points bigger in 2007 than it was in 2003. In other words, the SNP profited more from Mr Blair’s unpopularity in 2007 than it did in 2003. So although we saw in the previous chapter that Mr Blair was not markedly less popular in 2007 than he had been in 2003, his unpopularity proved more telling this time around. There are signs that the same might be true also of attitudes towards Iraq, where we also asked the same question in 2007 as we had done on our 2003 survey. Support for the SNP was 14 points higher in 2007 than it was in 2003 amongst those who strongly agreed that Britain was wrong to have gone to war in Iraq, whereas it hardly increased at all amongst those who either simply disagreed or who neither agreed nor disagreed.6 Again it seems that the SNP were better placed to profit from the 119

Revolution or Evolution? unpopularity of Labour’s record, though of course on this issue there were also many more people who were unhappy with that record than had been the case in 2003. At the same time, however, there is an important contrast between the role that unpopularity with aspects of Labour‘s record played in the SNP’s success and that fulfilled by the popularity of Mr Salmond. For the introduction of evaluations of Mr Blair and attitudes towards Iraq into our analysis has little impact on those measures that were included already. In other words, they do not seem to have been involved in enabling the SNP to win votes amongst its ‘natural’ constituency in particular. This indeed is confirmed if we compare the link between constitutional preference and both evaluations of Blair and attitudes towards the Iraq war in 2003 and 2007. The change between 2003 and 2007 in the proportion of people who gave Blair a good rating was exactly the same amongst supporters of devolution as it was amongst supporters of independence. The same is true of the change in the proportion thinking that Britain was wrong to go to war in Iraq. CONCLUSION In the introduction to this book we noted that the SNP focused its election campaign in 2007 on presenting itself as an effective alternative government rather than as a movement campaigning for independence. It might be thought that in so doing it was trying to sideline the issue of Scotland’s constitutional future. In practice, however, what that campaign seems to have done is to persuade those who were already in favour of independence, or at least a more powerful parliament, to vote for the party in unprecedented numbers. Rather than sidelining the issue of Scotland’s constitutional future, the SNP’s tactics – and the popularity of its leader in particular – succeeded in drawing into its ranks more effectively than previously the popular demand for constitutional change.7 But if the SNP seized its opportunity in 2007, it was devolution that gave the party its chance. In establishing elections to a separate Scottish parliament the devolution settlement created a forum in which voters in Scotland were invited to focus on who could best represent the needs and wishes of their part of the UK in particular. In so doing it established an environment in which the SNP consistently finds it easier to prosper and 120

How the SNP Won galvanise the support of its ‘natural’ constituency, even when it is led by an unpopular leader, as was the case in 2003. So in many respects the 2007 election represented politics as usual, a seemingly natural evolution of the devolution settlement. Devolved elections consistently provide a favourable environment for the SNP (albeit perhaps an unintended consequence of creating the Scottish Parliament) and the party exploited that environment in 2007 by presenting itself as an effective alternative government. But this was not just a case of voters deciding to hold the incumbent government to account by replacing it with a seemingly more attractive opposition; rather, by presenting itself as a credible alternative government, the SNP were able to galvanise and unite the existing demand for revolutionary constitutional change. Yet there is a hidden paradox in all this. For while the SNP might have been unusually successful in persuading those who want Scotland to become independent to vote for it, this does not mean that those who voted for the party in 2007 were unusually united in their support for independence. Rather the opposite was true. Those who preferred a devolved parliament might have been less likely to have switched to the SNP between 2003 and 2007, but, as we saw in Chapter 4, they constituted a far larger group than those who supported independence, and indeed especially so in 2007. So although the gains the SNP made amongst this group were smaller, their impact was large. Indeed, in 2007 the proportion of SNP voters who supported independence was unusually low. Just half (50%) wanted their country to become independent; previously in the period since 1999 that proportion had never been less than 59%. Despite the party’s particular success in winning the support of those backing independence, the coalition that helped bring the SNP to power was still very much a variegated one. NOTES 1. Indeed, we should note that when we asked our respondents how they would have voted if a UK general election rather than a Scottish Parliament election had been held on 3 May 2007, the proportion who said they would vote SNP was ten percentage points lower than the proportion backing the SNP in the Scottish Parliament election. 121

Revolution or Evolution? 2. It might be thought that the strengthened relationship between voting SNP and support for independence was simply a consequence of the relatively low level of support for independence registered in our 2007 survey (see Chapter 4). Perhaps those who expressed support for independence in 2007 consisted of the ‘hard core’ of independence supporters who have always been more likely to vote SNP. But if this argument were correct we should not also find, as in practice we do, a strengthening of the link between support for more powers (whose popularity was much the same as it was in 2003) and voting SNP. 3. Even if we exclude those who said they supported independence, 41% of those who said they strongly agreed that the Scottish Parliament should have more powers voted SNP, an increase of 19 points on 2007. There was also a 12-point increase (to 28%) amongst those who simply agreed with the proposition. 4. Labour’s image did show signs of improvement in 2003 when 13% said that it looked very closely after the interests of people in Scotland. But the 2007 reading of 7% was exactly the same as the equivalent reading in 1999. 5. Thus, for example, in 2003 17% of those who supported devolution and were voting on the basis of what was happening in Scotland backed the SNP, while the equivalent figure for those voting according to what was happening in Britain as a whole was 5%. The parallel figures for those opposing any kind of parliament were 10% and 2% respectively. 6. The pattern is, however, broken by a 15-point increase amongst those who disagreed. By 2007, however, this had become a small group comprising only 4% of our sample; the statistic is therefore highly subject to sampling error. 7. This suggests that it is unwise to assume, as Johns et al. (2009) do, that ‘valence’ issues, such as the perceived competence of the parties, mattered to voters because voters saw few differences between the parties on ‘position’ issues, such as Scotland’s constitutional status. Rather, our analysis suggests the role of valence issues was to mobilise divisions within the electorate on position issues.

122

How the SNP Won REFERENCES Bond, R. (2000), ‘Squaring the Circles: Demonstrating and explaining the political “non-alignment” of Scottish national identity’, Scottish Affairs, 32: 15–35. Bond, R. and Rosie, M. (2002), ‘National Identities in Post-Devolution Scotland’, Scottish Affairs, 40: 34–53. Bromley, C, (2006), ‘Devolution and electoral politics in Scotland’, in Hough, D. and Jeffery, C. (eds), Devolution and Electoral Politics, Manchester: Manchester University Press. Clarke, H, Sanders, D., Stewart, M. and Whiteley, P. (2004), Political Choice in Britain, Oxford: Oxford University Press. Curtice, J. (2004), ‘Adding Up the Figures: Conclusions from 2003’, in Spicer, M. (ed.), The Scotsman Guide to Scottish Politics, Edinburgh: Scotsman Publications. Denver, D. (2007), ‘“A Historic Moment”? The results of the Scottish Parliament Elections 2007’, Scottish Affairs, 60: 61–79. Jeffery, C. and Hough, D. (2003), ‘Regional Elections in Multi-Level Systems’, European Urban and Regional Studies, 10: 199–212. Johns, R., Mitchell, J., Denver, D. and Pattie, C. (2009), ‘Valence Politics in Scotland: Towards an explanation of the 2007 election’, Political Studies, 57: 207–33. McCrone, D. (2001), Understanding Scotland: The sociology of a nation, 2nd edn, London: Routledge. Paterson, L. (2006), ‘Sources of Support for the SNP’, in Bromley, C., Curtice, J., McCrone D., and Park, A. (eds), Has Devolution Delivered? Edinburgh: Edinburgh University Press. Rosie, M. and Bond, R. (2003), ‘Identity Matters: The political and social significance of feeling Scottish’, in Bromley, C., Curtice, J., Hinds, K. and Park, A. (eds), Devolution – Scottish Answers to Scottish Questions?, Edinburgh: Edinburgh University Press. Seawright, D. and Curtice, J. (1995), ‘The Decline of the Scottish Conservative and Unionist Party 1950–92: Ideology, religion or economics?’, Contemporary Record, 9: 319–42.

123

Revolution or Evolution? APPENDIX TO CHAPTER 6: SAMPLE SIZES FOR TABLES Table 6.1 Gender Men Women Age group 18–34 35–44 45–54 55–64 65+ Religion Protestant Catholic None Social class Employers and managers Routine and semi-routine workers

2003

2007

366 447

365 483

113 148 154 164 234

98 157 166 174 252

416 89 299

280 94 372

334 294

211 288

Table 6.2 Scottish not British More Scottish than British Equally Scottish and British More British than Scottish British not Scottish

1999 318 363 236 35 36

124

2001 351 329 265 35 34

2003 246 292 200 42 25

2005 305 341 245 53 56

2007 211 270 239 48 45

How the SNP Won Table 6.3 Independence Devolution No parliament

1999 276 613 92

2001 280 642 96

2003 208 563 114

2005 328 493 161

2007 176 483 76

1999 170 422 174 190 32

2001 220 506 135 127 44

2003 124 353 119 169 54

2005 175 462 171 154 59

2007 187 373 126 121 27

Table 6.4 Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree Table 6.6 Independence Devolution No parliament

2003 208 563 114

2007 176 483 76

2003 360 302 148

2007 341 285 220

Table 6.8 Good (7–10) Middle (4–6) Bad (0–3)

125

CHAPTER 7 __________________________________________

DO VOTERS CARE ABOUT PARTIES ANY MORE? We now turn our attention in the third part of our book to the second major change heralded by the 2007 elections, use for the first time of the single transferable vote (STV) electoral system to elect members of Scotland’s 32 local councils. As we noted in Chapter 1 this system is radically different from the single-member plurality system that had previously been in use in Scottish local elections. Under that system voters were simply asked to place one ‘X’ against the name of an individual candidate and these votes were tallied to determine who had most votes and thus was elected as the one and only representative for that ward. The new system, in contrast, invited voters to place the candidates in rank order of preference and these preferences were then counted in order to elect as many as three or four councillors in each ward. Thus STV enables voters to express a more complicated set of choices should they so wish. Under the previous system voters could do no more than indicate that they liked party x or party y best of all. Or if they particularly liked an individual candidate who was standing for a different party they could vote for him or her but have to accept that this meant being unable to express support for their preferred party. Under STV, in contrast, voters can indicate that they like more than one party by giving high preferences to the candidates of more than party. If they like an individual candidate who is standing for a party other than the one they most prefer, they can give that candidate their first preference but then still go on to give relatively high preferences to the candidates of their preferred party. And in any event voters may well find that more than one candidate has been nominated by their preferred party (in the belief that they can win more than one of the three or four seats at stake), and that they have to decide which of those candidates to rank first. In short STV makes it easier for voters to say more than, ‘I like this party most’. But it is an invitation, not a command. Voters can if they want confine their preferences to candidates of the party they like most. And if that party has nominated more than one candidate locally voters  

126

Do Voters Care about Parties Any More? can follow the advice of that party as to the order in which those candidates should be placed. In these circumstances elections under STV are just as much about a uninominal choice of party as they are under single-member plurality. How STV is used by voters in practice depends on the motivations they bring to the ballot box. As we noted in Chapter 1, this is very apparent in international experience of STV. In Ireland voters often do give a preference to candidates from more than one party and express their views on the merits of individual candidates. In contrast in both Malta and in Australia such behaviour is relatively rare. It would seem then that the most important influence on how voters use STV is likely to be what feelings they have towards political parties. If voters feel a strong sense of attachment to a political party they can be expected to want to support that party irrespective of any other considerations. But if on the other hand voters do not have such a strong sense of commitment, they might be thought likely to express a more nuanced choice. They would certainly seem more likely to give preferences to the candidates of more than one party, even if they put all of the candidates of their preferred party first. They might well be more willing to decide for themselves the order in which they place the candidates of their preferred party. However, they might also be willing to go much further and vote for the candidates they like best irrespective of the party for which those candidates are standing – or at least take personal qualities as well as party into account in deciding the order in which to place the candidates. So before looking directly at how people in Scotland actually used the STV ballot paper, a task we undertake in the next chapter, first of all in this chapter we assess the strength of the link between voters and parties in Scotland both before and at the time of the 2007 elections. In part we do so by examining a much used measure of commitment to political parties known as party identification. But we also assess a key behavioural measure thrown up by the fact that the Additional Member System (AMS) used in Scottish Parliament elections also gives voters some opportunity to express a more nuanced choice than is possible under single-member plurality. In particular, as noted in Chapter 1, under AMS voters cast not one vote, but two. One of these is for a MSP for their individual constituency, the other for a regional party list. Crucially there is no obligation on voters to cast these two votes for the same party – and the less they do so the less we would expect voters to confine their preferences to one party under STV. We thus examine how voters have  

127

 

Revolution or Evolution? used their two votes in Scottish Parliament elections for further clues as to how they might have approached the task of completing a STV ballot. PARTY IDENTIFICATION The potential role that emotional attachments to political parties can play in influencing how people vote has long been recognised in academic analysis of voting behaviour. In their pioneering study of voting behaviour in the United States in the 1950s Campbell et al. (1960) argued that many voters held such an emotional attachment, which they labelled ‘party identification’. Voters, they suggested, come to regard themselves as a ‘Democrat’, ‘Republican’ or ‘Independent’ (that is someone who does not regard themselves as a supporter of any party). This identity, it is argued, is often formed in childhood and passed from generation to generation (Jennings and Niemi, 1968), while in the United States at least it is institutionalised in the requirement placed on voters in many states to register as a Democrat or Republican in order to determine in which party’s primary elections they can participate. Thought of in that way, party identification is much like any other social identity, such as national identity, that is a claim to membership of a social group whose members acknowledge each other as being part of a collective ‘us’ (Tajfel and Turner, 1979). This party identification influences voters in a number of ways. It gives them a motivation to turn out and vote in order to express their identity. Of course, on occasion they may be dissatisfied with the current performance of the party with which they identify, but their sense of partisanship helps ensure that they remain loyal. Or if on occasion voters do vote for the candidate of another party, they still retain their party identification with the result that voters soon return to their original party loyalty – as seemed to be the case in the 1950s when many a Democrat identifier voted for the Republican presidential candidate (and former war hero) Dwight Eisenhower. Although some have doubted whether voters in Europe, including Britain, draw as sharp a distinction as voters apparently do in the US between their current party preference and their supposedly longer-term party identification (Butler and Stokes, 1974; Thomasson, 1976) or indeed whether party identification really is an emotional attachment at all (Fiorina, 1981; Clarke et al., 2004), the two  

128

Do Voters Care about Parties Any More? are certainly not wholly synonymous. For example, when our 2007 survey asked respondents whether they generally regarded themselves as supporter of or as closer to one of the parties, Labour emerged with as much as a nine point lead over the SNP; in other words the SNP were only able to defeat Labour in 2007 because some voters did not vote for the party with whom they still identify. Our interest here, however, is in how strong this sense of party identification is amongst the public, irrespective of whichever is the particular party with which they identify. For it has long been claimed that, even if people do still acknowledge some sense of identity (though perhaps even the degree to which that is the case is exaggerated by the wording commonly used to capture party identification in surveys conducted in the UK – see Bartle, 2001; Bartle, 2003; Sanders, 2002), in Britain as elsewhere they are far less likely nowadays to do so strongly (Crewe and Thomson, 1999; Denver, 2006) – much indeed as it is often argued that processes of individualisation are undermining social identities more generally (Heath et al., 2007). This decline is evidenced by the fact that relatively few now say that they are a ‘strong’ supporter of whichever party with which they claim to identify. Whether such a decline has occurred in Scotland is examined in Table 7.1. Here we have available to us the evidence compiled by Scottish Election Studies conducted before the advent of devolution as well as of the more recent Scottish Social Attitudes (SSA) series. Both sets of surveys asked the following question of those who declared that they supported or identified with a particular party: Would you call yourself very strong [name of party], fairly strong, or not very strong? As well as showing the proportion choosing each of these three options, we also show how many indicated that they did not identify with a party at all.

 

129

 

Revolution or Evolution? Table 7.1 Strength of Party Identification, 1979–2007

Very strong Fairly strong Not very strong None Sample size

1979 % 18 46 28 5 661

1992 % 13 46 33 6 957

1997 % 14 43 35 7 882

1999 % 10 41 40 7 1,482

Very strong Fairly strong Not very strong None Sample size

2003 % 7 28 46 12 1,508

2004 % 6 29 43 13 1,637

2006 % 6 25 46 14 1,594

2007 % 7 25 44 13 1,508

2001 % 8 31 46 11 1,605

Note: This question followed a sequence of questions designed to identify the direction of someone’s party identification. This sequence was different between 1979 and 1999 from that used from 2001 onwards. Source: Scottish Election Study, 1979–97; Scottish Social Attitudes, 1999–2007.

However, although these surveys all ascertained strength of identity in the same way, we need to be aware of two potential methodological disturbances to the continuity of the time series in Table 7.1. First, those surveys conducted up to and including 1999 ascertained the particular party with which respondents identified using a different sequence of questions than was deployed in those surveys conducted from 2001 onwards.1 Second, surveys that are introduced to respondents as being primarily about elections may be less successful than those that were stated to be about social attitudes in general in securing the participation of those with less interest in politics, and thus a weaker sense of commitment to a party (Heath et al., 2007: 39). Again this was a characteristic that the 1999 SSA shared with the earlier election studies (as the survey in that year was almost wholly about devolution and the election), and distinguishes these surveys from subsequent ones in the SSA series. In short, care needs to be exercised in looking at trends over the whole time series as opposed to examining separately developments in the period  

130

Do Voters Care about Parties Any More? between 1979 and 1999, and again between 2001 and 2007. As it happens, however, the evidence within these two periods consistently suggests that a gradual but persistent decline in people’s strength of party identification has taken place in Scotland. In 1979 nearly two-thirds (64%) reported a ‘very’ or ‘fairly’ strong identity; by 1999 the figure had fallen to little more than half (51%). Meanwhile whereas in 2001 around two in five (39%) said they identified with a party strongly, by 2007 a little under one in three did so (32%). There seems little doubt that by the time that STV was introduced in Scotland in 2007 only a minority had a strong sense of party identification and that there was good reason to anticipate that some voters at least would use the new ballot paper to do more than express loyalty to just one party (Beck et al., 1992). SPLITTING THE VOTE If, however, party loyalties were relatively weak we would anticipate that voters would also use the opportunity provided to them under the AMS system to cast a vote for two different parties. Moreover, there are a couple of other behavioural indications of weak or transient loyalties that we can examine. First, not only were local elections held on the same day as the parliamentary election in 2007, but this also happened in 1999 and 2003 (when the local elections were conducted using single-member plurality). So voters could vote for a different party in the local elections from whichever one they supported in the parliamentary election. Second, we have evidence for each Scottish election of how voters said they would have voted if a UK general election had been held on the same day. Indeed in this case we also have evidence for the UK general elections held in 2001 and 2005 of how people said they would have voted if a Scottish election had been held at that time.

 

131

 

Revolution or Evolution? Table 7.2 Voters Voting for Different Parties across Different Kinds of Election/Ballot, 1999–2007

Constituency and List Vote Sample size Holyrood and Local Council Sample size Holyrood and Westminster Sample size

1999 %

2001 %

2003 %

2005 %

2007 %

20

-

28

-

20

1,016

817

25

-

985

26

821 -

801

29 651

18

14

22

20

25

967

913

791

880

745

The first row shows the proportion who voted for different parties on the two votes for the Scottish Parliament election. The second row shows the proportion who reported voting for a different party on the constituency vote of the Scottish Parliament election from the one they supported in the contemporaneous local election. In 2007 local election vote is the party of the candidate who received the respondent’s first-preference vote. The third row shows the proportion supporting a different party on the constituency vote of a Scottish Parliament election from the one they supported in a UK general election. In those years when a Scottish election was held the UK election vote is a hypothetical vote; in those years when a UK election was held the Scottish Parliament constituency vote is a hypothetical one. All figures are based on those voting and naming a party in both contests. Source: Scottish Social Attitudes, 1999–2007.

Ticket splitting, that is supporting different parties on the two votes, is not uncommon in Scottish parliamentary elections. According to voters’ reports of how they voted in the first three such elections, regularly between a fifth and just over a quarter of voters do so. In practice this is something of an underestimate. In contrast to the 1999 and 2003 elections, in 2007 the two votes were cast on a single ballot paper (of which more below), while the votes were counted electronically. As a result a record was made of both the constituency and the list vote on each ballot paper. Moreover, in an unusual move, the resulting tally of constituency vote by list vote was published for every polling district in Scotland (Scotland Office, 2008). Adding up these tallies across the whole of Scotland reveals that, amongst those voters who cast a valid vote on  

132

Do Voters Care about Parties Any More? both ballots, 24.4% voted for different parties with their two votes, somewhat higher than the 20% figure recorded by our 2007 survey. This level of ticket splitting is much as we might anticipate from the experience of other mature democracies that use AMS. Ticket splitting used to be relatively rare in German federal elections (Klingemann and Wessels, 2003), but more recently has become more common and by 2005 reached a figure of nearly 25% (Conradt, 2006: 16–17). Meanwhile, since New Zealand switched to AMS (known there as Mixed Member Proportional) in 1996, the level of ticket splitting has been even higher, between 28% and 39% (Vowles et al., 2006). So from this evidence party loyalties are not particularly weak in Scotland; rather they are simply as weak as we might anticipate given contemporary experience elsewhere. We might, though, anticipate that voters would be more likely to vote for different parties in different kinds of election than to do so in the same election. However, in 1999 and 2003 the proportion of people voting for a different party in the local elections from the one that they backed on the constituency vote of the parliamentary election2 was, at around a quarter, not dissimilar to the proportion that split their two votes in the parliamentary contest. Still, that figure does at least suggest that it was possible to exaggerate the concern that has been expressed by some that holding the two kinds of election on the same day meant that the outcome of the local elections was overwhelmed by the national popularity of the parties (Gould, 2007: 33–7; McIntosh, 1999; Scottish Government, 2008). Meanwhile, as we might have anticipated from the fact that the SNP systematically perform better in Scottish elections than in UK general elections, around one in five voters typically votes differently from the way that they would have done if a different kind of parliamentary election had been held that day. CONFUSION? However, we should be aware that words of caution are sometimes uttered about assuming that split ticket voting is a reflection of voters’ motivations, such as weak party identification. Rather, it is argued voters often split their ballots under AMS simply because they do not understand how the system operates (Jesse, 1988; Schoen, 1999). One potential  

133

 

Revolution or Evolution? source of confusion to which reference is often made is the use, as in Germany, of the terms ‘first vote’ and ‘second vote’ to refer to the constituency and list vote respectively. These terms soon became commonplace in informal discussion of the voting system in Scotland too and would seem to run the risk that voters believe that the list vote is a ‘second-choice’ vote, when in fact it is the more important of the two votes so far as the determination of the total number of seats won by each party is concerned (Arbuthnott, 2006). Indeed in 2003 only around a half of the adult Scottish public were able to recognise as false the statement that under AMS, ‘People are given two votes so that they can show their first and second preferences’. Voters themselves were willing to admit to considerable difficulty in understanding how the AMS system translates votes cast into seats won. In both 1999 and 2003 as many as 40% of those who voted said that this was either ‘very’ or ‘fairly’ difficult to understand, almost as many as the 45% and 44% respectively who said it was ‘not very’ or ‘not at all’ difficult. The equivalent figures in 2007 were much the same at 42% and 40% respectively. However, as Table 3 shows, voters were much less likely to admit that they found it difficult to complete the ballot papers themselves. That was typically regarded as a relatively straightforward exercise – until at least a new single ballot paper was introduced in 2007. Table 7.3 Perceptions of Difficulty of Completing Scottish Parliamentary Ballot Paper(s), 1999–2007

Very difficult Fairly difficult Not very difficult Not at all difficult Sample size

1999 % 1 8 38 51 873

2003 % 2 11 30 47 808

2007 % 6 25 34 34 831

Table based on those who reported voting in the parliamentary election. Source: Scottish Social Attitudes, 1999–2007.

In any event there was only limited evidence in either 1999 or 2003 that voters had been more likely to fail to vote for their first-choice party on the list vote than they were on the constituency ballot. After  

134

Do Voters Care about Parties Any More? each Scottish election respondents to the SSA survey were asked to state ‘which party would you have put as your first choice’ if ‘you had been given a different kind of ballot paper, that required you to give two votes, in order of preference’. As Table 7.4 shows, in 1999 voters were just five points more likely to vote on the list ballot than on the constituency vote for a party other than the one they said they would have indicated was their first choice. In 2003 the difference was only a little greater, at nine points. Nevertheless, a UK government-appointed commission (Arbuthnott, 2006) recommended that a single ballot paper be used in Scottish parliamentary elections with the list vote placed on the left-hand side (thus making it the vote more likely to be completed and regarded as ‘first’), a recommendation that the UK government, with whom responsibility for the conduct of Scottish parliamentary election lies, decided (unlike most of the Commission’s other proposals) to implement.3 Table 7.4 Voting for Other than First-Choice Party, 1999–2007

Constituency Sample size List Sample size

1999 % 12 1,012 17 1,010

2003 % 16 814 25 824

2007 % 16 806 10 841

Source: Scottish Social Attitudes, 1999–2007.

In the event the new ballot paper itself sowed the seeds of confusion. As can be seen from Table 7.3 above as many as 31% of those who voted in 2007 said that they regarded the paper as ‘very’ or ‘fairly’ difficult to complete, compared with just 13% in 2003. More importantly the change coincided with a dramatic increase in the number of invalid votes. No less than 4.1% of all ballot papers contained an invalid constituency vote that could not be included in the count, while 2.9% had an invalid list vote. In contrast in 2003 only a little under 0.7% of either vote was ruled invalid. The principal source of confusion seems to have been that voters thought they could only cast one vote on the single ballot paper; around half of the invalid votes were instances where no constituency vote was  

135

 

Revolution or Evolution? cast, while another quarter consisted of a failure to register a list vote (Gould, 2007).4 Such a high level of invalid votes in what had proven to be a closely contested election led some to wonder whether the voters’ evident confusion had affected the overall result. Indeed there is good reason to believe that Labour was particularly disadvantaged by the confusion. No less than 42% of those with no educational qualifications said that they found the ballot paper ‘very’ or ‘fairly’ difficult to complete, a group amongst whom support for Labour was as much as nine points higher than in our survey as a whole.5 Equally Carman et al. (2008) found that the incidence of invalid votes was higher in less affluent (and thus typically more strongly Labour) parts of the country. But perhaps most tellingly of all, no less than 4.3% of those who voted for Labour on the list vote cast an invalid constituency vote, compared with 3.0% of those who voted for the SNP (and just 2.0% of Conservative and 1.8% of Liberal Democrat voters).6 The difference between these Labour and SNP proportions amounts to just under 7,000 votes. However, as the SNP’s lead over Labour on the constituency vote was just over 24,000 votes, it cannot credibly be argued that the SNP only secured a nationwide lead on that part of the ballot paper because of Labour’s apparent disproportionate loss of votes. Still, despite the confusion that it generated, it appears that the new ballot paper did make it less likely that voters voted on the list vote for a party other than one they most preferred. Indeed just 10% reported doing so, while the proportion saying they behaved in that way on the constituency vote was, at 16%, actually the higher of the two figures this time around. But whatever its incidence, if we are to regard either failure to vote for the party most preferred or voting for different parties on the two ballots as the product of confusion, we should be able to show that such behaviours were more common amongst those who might be thought more likely to be confused. Of this, however, there is only limited and inconsistent evidence. One pattern that we might expect to find, if voters split their votes because they do not understand how the AMS system works, is that those who thought that it was difficult to complete the ballot paper or understand how votes are translated into seats would be more likely to split their votes. Table 7.5 examines both of these possibilities. For all three elections it shows those who reported difficulty in completing the  

136

Do Voters Care about Parties Any More? ballot paper exhibited only a slightly greater tendency to split their votes.7 Meanwhile, apart from the relatively small group who said it was not at all difficult to understand how votes were translated into seats, the perceived difficulty of the mechanics of the electoral system apparently was not related at all to the chances that someone would use the two ballots to vote for different parties. Table 7.5 Ticket Splitting by Perceived Difficulty of Ballot Paper and Votes/Seats Relationship, 1999–2007

% ticket split 1999 2003 2007 % ticket split 1999 2003 2007

Completing Ballot Paper Not at all Very/fairly Not very difficult difficult difficult 29 20 22 35 26 27 22 21 16 Understanding Votes into Seats Not at all Very/fairly Not very difficult difficult difficult 25 23 20 27 30 22 20 21 12

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

Equally, Table 7.6 suggests that failure to vote for one’s first choice party on either of the ballots is not consistently linked to the perceived difficulty of the electoral system. So far as the perceived difficulty of understanding the relationship between votes and seats is concerned, once again it is only amongst the small group who say the relationship is ‘not at all’ difficult to comprehend that the incidence of the behaviour in question has sometimes been a little lower. Meanwhile although there does seem to have been some link between failure to vote for the party they most preferred and the perceived difficulty of the ballot paper in 2003, the link seems to have been weak or non-existent in both 1999 and 2007.8

 

137

 

Revolution or Evolution? Table 7.6 Not Voting for First Choice by Perceived Difficulty of Ballot Paper and Votes/Seats Relationship, 1999–2007 % not vote for first-choice party 1999 Constituency List 2003 Constituency List 2007 Constituency List % not vote for first-choice party 1999 Constituency List 2003 Constituency List 2007 Constituency List

Completing Ballot Paper Very/fairly Not very Not at all difficult difficult difficult 15 13 13 16 18 17 23 17 13 35 21 24 19 16 14 10 12 7 Understanding Votes into Seats Very/fairly Not very Not at all difficult difficult difficult 13 15 7 19 18 14 18 14 8 26 27 20 17 17 15 10 12 4

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

A second pattern we might expect to obtain if ticket splitting and failure to vote for one’s first choice party were primarily the product of voter confusion, is that those with fewer educational qualifications would be more likely to exhibit these behaviours. In fact, the very opposite is true. In 2007, 30% of those with a university degree split their tickets, compared with 17% of non-graduates. In 1999 and 2003 the equivalent figures were 33% and 44% respectively amongst graduates, but only 17% and 24% amongst non-graduates. Meanwhile Table 7.7 shows graduates have also been more likely to fail to vote for the party they stated was their first choice for the Scottish Parliament election. This was particularly true of the list vote in 1999 and 2003, that is the very vote on which fears that voters were mistakenly expressing a second choice were focused.9

 

138

Do Voters Care about Parties Any More? Table 7.7. Not Voting for First Choice by Highest Educational Qualification, 1999–2007 Highest Qualification % not vote for first-choice party 1999 Constituency List 2003 Constituency List 2007 Constituency List

Degree 15 26 15 37 25 12

Less than degree 12 15 16 23 14 9

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

WEAK PARTISANSHIP? But if there is at most limited and far from consistent evidence that split ticket voting and voting for other than one’s first-choice party are the product of confusion, can we demonstrate that the incidence of these behaviours depended on voters’ motivations? There is certainly clear evidence that, as anticipated, such behaviour is more common amongst those with a weak sense of party identification (Gcshwend, 2007; Karp et al., 2002). For example, as Table 7.8 shows, in both 1999 and 2007 those who only had a weak party identity or else none at all were around three times as likely to split their votes as were those with a ‘very strong’ identity. Indeed much the same is true of the other kinds of voting for different parties that we introduced in Table 7.2. Thus for example, in 2003, as many as 30% of those with a weak identity or none at all voted for different parties in the local elections and on the constituency vote of the parliamentary election, whereas only 19% of those with a stronger identity did so. Meanwhile, the equivalent figures for voting differently from the way they would have done in a UK general election were 24% and 17% respectively.

 

139

 

Revolution or Evolution? Table 7.8 Ticket Splitting by Strength of Party Identification, 1999–2007

% ticket split 1999 2003 2007

Very strong 8 21 9

Strength of Party Identification Fairly Not very strong strong/none 18 25 20 33 13 25

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

Equally, failure to vote for one’s first-choice party is generally highest amongst those with a weak party identity or none at all (see Table 7.9).10 Moreover, this seems to be equally true of the two votes. So it seems likely that the decline in the strength of party identification in recent years means that STV was indeed introduced into local elections in Scotland at a time when the electorate was more likely to take the opportunity the system affords to express a more nuanced electoral choice.11 Table 7.9 Not Voting for First Choice by Party Identification, 1999–2007 % not vote for firstchoice party 1999 Constituency List 2003 Constituency List 2007 Constituency List

Strength of Party Identification Very Fairly Not very strong strong strong/none 6 9 18 7 16 19 13 7 21 22 17 31 8 10 21 1 9 12

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

 

140

Do Voters Care about Parties Any More? SOPHISTICATED VOTING? Still, while the fact that a voter lacks a strong party identity might indicate that they are potentially willing to vote for different parties on the two AMS ballots, and in so doing sometimes fail to vote for the party they most prefer, it does not explain why they should decide to take the opportunity to do so (Roberts, 1988). One possible explanation is tactical (otherwise known as strategic) voting. A voter is said to vote tactically if they decide not to vote for the party they most prefer because they feel voting for another party would be more likely to help prevent an outcome they would prefer to avoid. Under single-member plurality such voting occurs when a voter believes that the party they like most has no chance of winning in their constituency, and they opt to vote instead for a party they think has a chance of defeating a party the voter does not want to win. For example, in recent UK elections many Liberal Democrat supporters living in constituencies where their party lies third have decided to vote Labour in order to help defeat the Conservatives locally, a pattern that seems to have been reciprocated by Labour supporters where their party lay third locally (Curtice and Steed, 1997, 2001; Evans et al., 1998; Fisher and Curtice, 2006). The constituency vote in Scottish Parliament elections is, of course, cast in a single-member plurality election. So we might expect such tactical voting to be evident on this particular vote. Indeed, it might be thought to be more likely to occur than in an election where only a singlemember plurality contest is taking place. For in a Scottish Parliament election voters can vote tactically on the constituency vote and still express their support for their first-choice party on the more proportional list vote. In short they can both vote tactically and express their first preference rather than be forced to choose to take one course of action or the other. At the same time, the list vote is not devoid of potential tactical considerations either. Voting for a very small party might still prove to be a ‘wasted’ vote; in practice a party that is incapable of winning at least 5% of the list vote in a region is unlikely to succeed in winning one of the ‘additional’ party list seats to be allocated in that region. So, a voter who doubts the ability of their preferred party to reach that threshold may prefer to vote for a party they also find relatively attractive and does seem to have a chance of securing a seat. Second, a party might be so successful  

141

 

Revolution or Evolution? in winning constituency seats in a region that it is not entitled to any additional seats. In 2003, for example, this was the position in which the Labour party found itself in no less than six of the eight Scottish Parliament regions, while the same was true of the Liberal Democrats in three. Supporting on the list vote a party that does relatively well in the constituency contests thus runs the risk of having no impact on the outcome. To this there is another consideration to add (Gschwend, 2007). The use of a relatively proportional electoral system in Scottish Parliament elections makes it unlikely that any one party will succeed in winning an overall majority on its own. Thus a coalition government may well be formed, as Labour and the Liberal Democrats did between 1999 and 2007. Voters may well have views on whether it is desirable for a coalition to be formed and between whom. It is sometimes argued that AMS gives voters the chance to express such preferences by dividing their votes between the two potential coalition partners (Bogdanor, 1981). This might be thought to be particularly efficacious where casting a vote for one’s first-choice party would seem unlikely to help it win a seat, whereas voting for its potential coalition partner might help that party to do so (Cox, 1997). In short, split ticket voting and not voting for one’s first choice of party could reflect a relatively sophisticated understanding of how a more nuanced form of voting might be a more effective way of achieving a voter’s objectives under AMS (Bawn, 1999; Pappi and Thurner, 2002). But how far is such behaviour evident in practice? After each of the three Scottish elections, the SSA survey has asked its respondents the following question in respect of both their constituency vote and their list vote: Which one of the reasons on this card comes closest to the main reason you voted for the party you chose? I always vote that way I thought it was the best party I really preferred another party but it had no chance of winning in this constituency/any party list seats in this region Those who gave the third answer are deemed to have voted tactically. In practice, however, this group proves to be relatively small.  

142

Do Voters Care about Parties Any More? Consistently across all three elections and the two votes, no more than between 4% and 6% say they voted the way they did for that reason. Moreover, despite our speculation that perhaps such voting would be more common on the constituency vote than in a stand-alone singlemember plurality election, it seems that in practice it is less so. When much the same question was asked of voters in Scotland after each of the last three UK general elections, consistently between 9% and 10% indicated that they voted tactically.12 Still, such tactical voting as does occur is associated with ticket splitting. For example, amongst those who in 1999 said that they had voted tactically on the constituency vote, 56% voted for a different party on the list vote; the equivalent figures for 2003 and 2007 were 46% and 55% respectively. But perhaps there is greater evidence of the potentially more significant behaviour of voters splitting their tickets in order to promote the prospects of a particular coalition. First of all we might expect voters who do this to prefer coalition government in principle. However, consistently only a half or so say they prefer having ‘a government in Edinburgh formed by two political parties together – in coalition’, while two in five back having ‘a government in Edinburgh formed by one political party on its own’.13 Still, as Table 7.10 shows, the half or so of voters who say they prefer coalition government are rather more likely to vote for different parties on the two ballots. Table 7.10 Ticket Splitting by Preference for Single-Party or Coalition Government, 1999–2007 % ticket split 1999 Sample size 2003 Sample size 2007 Sample size

Preferred Government Single-party Coalition 14 24 416 538 26 29 353 427 15 22 359 390

Source: Scottish Social Attitudes, 1999–2007.

Crucial, however, is whether such vote splitting conforms to voters’ preferences as to the parties they would like to see form a coalition.  

143

 

Revolution or Evolution? Irrespective of whether or not they said they preferred a coalition in principle, respondents to the SSA have also been asked after each Scottish election which parties they would like to see form a coalition. With four larger parties together with a varying number of smaller ones capable of securing representation in the parliament, the number of possible coalitions is not inconsiderable. However, given a reluctance on the part of the Conservatives to enter any kind of coalition (a reluctance reciprocated by the other parties), in practice only two possible coalitions have seemed at all likely, that is between the Liberal Democrats and Labour or between the Liberal Democrats and the SNP. Our survey question thus asked respondents to choose between these two possibilities. As we might anticipate given the outcome of those elections, the prospect of a Labour/Liberal Democrat coalition was attractive to more voters in 1999 and 2003 than was the possibility of a deal between the SNP and the Liberal Democrats. In 1999, 53% preferred a Labour/ Liberal Democrat coalition, as did 50% in 2003; in contrast, just 32% wanted to see a SNP/Liberal Democrat coalition as did no more than 36% in 2003. In 2007, however, a SNP/Liberal Democrat coalition was slightly the more popular of the two options; that possibility was supported by 44% whereas only 37% now backed a Labour/Liberal Democrat partnership. Meanwhile in 1999 there was some evidence that those who backed a Labour/Liberal Democrat coalition were more likely to split their ticket between those two parties. Amongst those who said they preferred a coalition rather than a single-party government, who backed the idea of Labour/Liberal Democrat coalition in particular, and who voted for one of those two parties on the constituency vote, 10% split their tickets by voting for the other of the two parties on the list vote. In contrast, the equivalent figure amongst those who voted for one of those two parties on the constituency vote, but who said they preferred single-party government, was just 3%. However, no such clear difference was evident in the behaviour of these two groups in 2003 or 2007. On both those occasions 5% of those who voted Labour or Liberal Democrat on the constituency vote and who preferred single-party government shared their two votes between the two parties, only a little less than the 6% (in 2003) or 7% (in 2007) who did so amongst supporters of a Labour/Liberal Democrat coalition.  

144

Do Voters Care about Parties Any More? There is perhaps a hint too that in 2007 supporters of a SNP/Liberal Democrat coalition were more likely to split their ticket between the two parties. Just 1% of those who voted for one of those two parties on the constituency vote and who preferred single-party government split their ticket by voting for the other party on the list vote. In contrast 4% of those SNP and Liberal Democrat constituency voters who backed a coalition between those two parties did so. Still, this latter figure is hardly a large one. Meanwhile, in 1999 and 2003 there was no sign at all of those who backed the idea of a coalition between Labour and the SNP being more willing to divide their votes between the two parties.14 Of course, one possible reason is that in contrast to the position in which Labour – and to a lesser extent the Liberal Democrats – found themselves, there was nowhere in Scotland where the nationalists were likely to do so well in the constituency part of the election that the party would not be eligible for at least one list seat. CANDIDATE VOTING? It seems, then, that there is only limited evidence of voters using the opportunities provided by AMS to engage in various forms of sophisticated voting. But we still have to consider one further possible explanation of why voters might split their tickets and fail to vote or their first-choice party. Moreover, it is perhaps the one of greatest interest to us in assessing the likelihood that voters might have taken the opportunity provided by STV to vote on the basis of the perceived merits of individual candidates as well as (or even instead of) the attractiveness of the various parties. This is that voters might use their constituency vote to vote for the candidate they most prefer rather than the party they most like. These constituency votes are, after all, formally at least cast for individual candidates rather than parties. Meanwhile, under AMS voters can vote for the candidate they most like in their local constituency contest and still express their support for their first-choice party on the list vote. Indeed, it is only the latter vote that should matter in determining the overall strength of the parties; if, by voting for a candidate of a different party, a voter enables that candidate to win, the only consequence should be that that candidate’s party secures one list seat fewer.15 So AMS would seem to provide a strategic environment that makes it easy for voters to vote for  

145

 

Revolution or Evolution? candidates rather than parties on the constituency vote (Bawn, 1999; Pappi and Thurner, 2002). Table 7.11 shows how voters themselves responded when asked after each of the three Scottish elections which of three possible reasons best described why they voted as they did on the constituency vote. These reasons were: I voted for a party, regardless of the candidate I voted for a party only if I approved of the candidate I voted for a candidate, regardless of his or her party It seems that party mattered far more than candidate to the way that most people voted. Typically well over half have said that they voted for a party, regardless of candidate, whereas only around one in ten said they did the opposite. Nevertheless, that still suggests that voters are twice as likely to cast a personal vote for an individual candidate than they are to vote tactically on the constituency ballot. Meanwhile, typically around a quarter of voters indicate that while they voted primarily for a party, their vote was conditional on what they thought of that party’s candidate. So while the qualities of individual candidates may have been secondary in most people’s minds in deciding how to vote, they were not necessarily entirely absent. Table 7.11 Reported Motivation for Constituency Vote, 1999– 2007

Voted for party regardless of candidate Voted for party if approved of candidate Voted for candidate regardless of party Sample size Source: Scottish Social Attitudes, 1999–2007.

 

146

1999 %

2003 %

2007 %

62

52

53

23

27

27

9

13

10

1,060

908

889

Do Voters Care about Parties Any More? But are those who claim to be voting for the candidate rather than the party they like most suitably distinctive in the manner in which they vote? Table 7.12 suggests that they are indeed far less likely than other voters to have voted on the constituency ballot for the party they liked most. Some voters evidently really did take the opportunity to cast a purely personal vote. Even so, less than half of those who said they voted for a candidate regardless of party ended up as a result voting for other than their firstchoice party. Evidently many of these voters did not find themselves having to choose between the best candidate and the best party; rather, the two were synonymous. Table 7.12 Not Voting for First Choice by Motivation for Constituency Vote, 1999–2007 % not vote for first choice on constituency vote 1999 2003 2007

Party 7 9 10

Motivation for Vote Depended Candidate 13 39 16 38 17 44

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

Those voters who said they voted for a candidate, regardless of party, are also more likely than other voters to split their ticket (see Table 7.13). Typically around half report doing so, similar to the equivalent proportion amongst those who said they voted tactically. In contrast, in 1999 and 2007 only around one in eight voters who said they had voted for a party regardless of candidate split their tickets, though this figure temporarily crept up to one in five in 2003.

 

147

 

Revolution or Evolution? Table 7.13 Ticket Splitting by Motivation for Constituency Vote, 1999–2007 % ticket split 1999 2003 2007

Motivation for Vote Depended Candidate 22 42 29 48 23 51

Party 13 21 12

For details of the sample sizes see the Appendix to this chapter. Source: Scottish Social Attitudes, 1999–2007.

So it seems some voters did take the opportunity afforded by AMS to cast a personal vote for an individual candidate rather than for a party, and this in part helps explain the incidence of ticket splitting. Nevertheless, their numbers should not be exaggerated. On this evidence at least STV was not launched upon an electorate that readily focused on the merits of individual candidates rather than parties. CONCLUSION Parties do indeed to seem to matter less to voters in Scotland than they once did. Considerably fewer now feel a strong emotional attachment to a party than they did twenty or thirty years ago. Meanwhile, those who do not feel a strong sense of attachment to a political party are indeed less likely to remain consistently loyal to one party, as evidenced by the way in which they vote under the AMS voting system used in Scottish Parliament elections. Yet while there is relatively little evidence that people divide their votes between two parties and fail to vote for the party they most like because of confusion, there is also only limited evidence of voters using the system to express sophisticated nuanced choices. Equally, while some apparently pay some consideration to the merits of the individual candidates in deciding how to cast their constituency vote, few seem to go so far as to put candidate before party. So it seems that STV was introduced in the local elections on an electorate whose ties to party may have been weakened, but were still far from broken. While some might be expected to have used the opportunity afforded by the new system to express a more nuanced choice  

148

Do Voters Care about Parties Any More? that paid attention to the merits of candidates rather than parties, it is far from clear that sufficient will have done so to revolutionise the character of local elections in Scotland. In the next chapter we examine exactly how Scotland’s voters did indeed use the new system. NOTES 1. The question sequence up to and including 1999 ran as follows. First respondents were asked, ‘Generally speaking, do you think of yourself as Conservative, Labour, Liberal/Liberal Democrat, Nationalist or what?’ Those who did not name a party in response to this question were further asked, ‘Do you generally think of yourself as a little closer to one of the parties than the others?’, and if they said, ‘Yes’, were asked, ‘Which party?’. From 2001 onwards the initial question was, ‘Generally speaking, do you think of yourself as a supporter of any one political party?’ If respondents did not name a party in response to thus they were then asked, ‘Do you think of yourself as a little closer to one political party than to the others?’, while those who did not name a party at this point were finally asked, ‘If there were a general election tomorrow, which political party do you think you would be most likely to support?’. 2. We use the constituency rather than the list vote to illustrate the level of vote splitting because the strategic context of the constituency vote is closer to that in which a Westminster vote is cast, as indeed it was to the context in which a local vote was cast prior to the introduction of STV 2007. 3. We should note, however, that the final form of the ballot paper (shown in Carman et al., 2008) departed considerably from the New Zealand ballot paper that had appealed to Arbuthnott. 4. This pattern did, however, mean that just 0.7% of ballot papers contained neither a valid constituency nor a valid list vote (figure calculated from Scotland Office, 2008). Thus the number of voters who were wholly disenfranchised by the confusion was relatively small. 5. Though we should note that the figure was also relatively high amongst graduates (33%), perhaps because members of this group were particularly likely to be aware of the controversy that subsequently surrounded the high incidence of invalid votes.  

149

 

Revolution or Evolution? 6. Figures calculated from Scotland Office (2008). 7. We should perhaps also bear in mind the possibility that voters were more likely to regard the ballot papers as difficult to complete because they wanted to vote for two different parties. 8. We should also note that in so far as there was a link in 2003, it was apparent on both the constituency and the list vote. Such voter confusion that may have led some people on that occasion to fail to vote for their preferred party was evidently not confined to the ‘second’, list vote. 9. Meanwhile, intriguingly, graduates were particularly likely to report having failed to vote for their first-choice party on the constituency vote in 2007, when that vote was placed on the right-hand side of the single ballot paper. 10. Much of the failure to vote for one’s first choice is in fact also specifically associated with a decision to vote for one of the smaller parties (that is other than the Conservatives, Labour, the Liberal Democrats or the SNP). For example, in 1999 and 2003 no less than three-quarters of those who voted for a smaller party on the list vote said that the party was not their first choice; such behaviour accounted for a quarter of all the list votes not cast for one’s first-preference party in 1999 and no less than nearly half (48%) in 2003. In 2007 the proportion of the list vote for smaller parties accounted for by nonfirst- choice voting fell back to 42%, though this still constituted 29% of all non-first-choice list votes. In so far as the decision to place the list vote on the left-hand side of the ballot paper did discourage voters from using that vote to vote for other than their first-preference party (whether out of confusion or otherwise), it evidently also served to discourage people from voting for one of the smaller parties. Meanwhile, according to the tabulations subsequently released by the Scotland Office (2008), in 2007 those voting for one of the smaller parties on the list vote constituted no less than 55% of all those who voted for different parties on the two votes. 11. We should, however, note that the smaller parties (and especially the Greens) often failed to contest the constituency ballot and so those voters who regarded them as their first-preference party were unable to support their first preference on that ballot, and thus were either obliged to split their votes or abstain on the constituency vote. To that extent some voters’ decisions to split their tickets and vote for a party other  

150

Do Voters Care about Parties Any More? than their first choice was a consequence of the choice they faced rather than their motivation. However, even at the height of their popularity in 2003 only 7% of voters said that one of the smaller parties was their firstpreference party, while in 1999 and 2007 only 2% and 5% respectively did do. Further, not all of these voters would have been unable to vote for their first-preference party on the constituency vote. 12. The figure for 1997 is 10% and comes from the Scottish Election Study, which asked the same question as detailed in the main text. The figure for 2001 and 2005 is 9% and comes from the Scottish booster sample of the British Election Study for those years (Clarke et al., 2004; 2009). This asked a somewhat differently worded question, which read as follows: ‘People give different reasons why they vote for one party rather than another. Which of the following best describes your reasons? The party has the best policies. The party has the best leader. I really preferred another party, but it stood no chance of winning in my constituency.’ 13. In 1999, 52% (of all respondents) said they preferred coalition government, while in 2003 and 2007 the figure was 50% and 51% respectively. Meanwhile in 1999, 40% supported single-party government, a figure that was unchanged in 2003 and increased by just one point in 2007. Note that there is no evidence that the experience of coalition government for eight years had any impact on attitudes towards its desirability. 14. In 1999, 6% of SNP and Liberal Democrat constituency voters who preferred single-party government voted for the other party on the list vote, while 7% of those who preferred a coalition between the two parties did so. In 2003 the equivalent figures were 8% and just 2% respectively. 15. However, it is possible for a party to win more constituency seats in a region than its proportional entitlement, in which case the result in an individual constituency can make a difference to the overall outcome. This happened in a number of regions at all three elections, primarily to the benefit of the Labour party. For details in respect of the 1999 and 2003 elections see Curtice (2006). In 2007 Labour won three extra seats in two regions (Glasgow and the West of Scotland), two at the expense of the SNP and one the Conservatives. The Liberal Democrats also secured an extra seat at the expense of the SNP in the Highlands and Islands. The net effect was thus to cut the SNP’s majority from seven seats to just one.  

151

 

Revolution or Evolution? REFERENCES Arbuthnott, J. (chmn) (2006), Putting Citizens First: Boundaries, voting and representation in Scotland, Edinburgh: The Stationery Office. Bartle, J. (2001), ‘The measurement of party identification in Britain: Where do we stand now?’, British Elections and Parties Review, 11: 9–22. Bartle, J. (2003), ‘Measuring Party Identification: An exploratory study with focus groups’, Electoral Studies, 22: 217–37. Bawn, K. (1999), ‘Voter Responses to Electoral Complexity: Ticket splitting, rational voters and representation in the Federal Republic of Germany’, British Journal of Political Science, 29: 487–505. Beck, P., Baulm, L., Clausen, A., and Smith, Jr, E. (1992), ‘Patterns and Sources of Ticket Splitting in Subpresidential Voting’, American Political Science Review, 86: 916–28. Butler, D. and Stokes, D. (1974), Political Change in Britain, 2nd edn, London: Macmillan. Campbell, A., Converse, P., Miller, W. and Stokes, D. (1960), The American Voter, New York: John Wiley and Sons. Carman, C., Mitchell, J. and Johns, R. (2008), ‘The Unfortunate Natural Experiment in Ballot Design: The Scottish parliamentary elections of 2007’, Electoral Studies, 27: 442–59. Clarke, H., Sanders, D., Stewart, M. and Whiteley, P. (2004), Political Choice in Britain, Oxford: Oxford University Press. Conradt, D. (2006), ‘The Tipping Point: The 2005 election and the deconsolidation of the German party system?’, in Langenbacher, E. (ed.), Launching the Grand Coalition: The 2005 Bundestag election and the future of German politics, Oxford: Berghahan. Cox, G. (1997), Making Votes Count: Strategic co-ordination in the world’s electoral systems, Cambridge: Cambridge University Press. Clarke, H., Sanders, D., Stewart, M. and Whiteley, P. (2004), Political Choice in Britain, Oxford: Oxford University Press. Clarke, H., Sanders, D., Stewart, M. and Whiteley, P. (2009), Performance Politics and the British Voter, Cambridge: Cambridge University Press. Crewe, I. and Thomson, K. (1999), ‘Party Loyalties: Dealignment or realignment’, in Evans, G. and Norris, P. (eds), Critical Elections: British parties and elections in long-term perspective, London: Sage.  

152

Do Voters Care about Parties Any More? Curtice, J. (2006), ‘Forecasting and Evaluating the Consequences of Political Change: Scotland and Wales’, Acta Politica, 41: 300–14. Curtice, J. and Steed, M. (1997), ‘Appendix 2: The Results Analysed’, in D. Butler and D. Kavanagh, The British General Election of 1997, London: Macmillan. Denver, D. (2006), Elections and Voters in Britain, 2nd edn, Basingstoke: Palgrave Macmillan. Evans, G., Curtice, J. and Norris, P. (1998), ‘New Labour, New Tactical Voting?’, in Denver, D., Fisher, J., Cowley, P. and Pattie, C. (eds), British Elections and Parties Review Volume 8, London: Frank Cass. Fiorina, M. (1981), Retrospective Voting in American National Elections, New Haven: Yale University Press. Fisher, S. and Curtice, J. (2006), ‘Tactical Unwind? Changes in Party Preference Structure and Tactical Voting between 2001 and 2005’, Journal of Elections, Public Opinion and Parties, 16: 55–76. Gould, R. (2007), Scottish Elections 2007: The Independent Review of the Scottish Parliamentary and Local Government Elections 3 May 2007, London: Electoral Commission. Gschwend, T. (2007), ‘Ticket-Splitting and Strategic Voting under Mixed Electoral Rules: Evidence from Germany’, European Journal of Political Research, 46: 1–23. Heath, A., Martin, J. and Elgenius, G. (2007), ‘Who Do We Think We Are? The Decline of Traditional Social Identities’, in Park, A., Curtice, J., Thomson, K., Phillips, M., and Johnson, M. (eds), British Social Attitudes: The 23rd report – perspectives on a changing society, London: Sage. Jennings, K. and Niemi, R. (1968), The Transmission of Political Values from Parent to Child’, American Political Science Review, 62: 169–84. Jesse, E. (1988), ‘Split-Voting in the Federal Republic of Germany: An analysis of federal elections from 1953 to 1987’, Electoral Studies, 7: 109–24. Karp, J., Vowles, J., Banducci, S. and Donovan, T. (2002), ‘Strategic Voting, Party Activity and Candidate Effects: Testing explanations for split voting in New Zealand’s new mixed system, Electoral Studies, 21: 1–22. Klingemann, H.-D. and Wessels, B. (2003), ‘Political Consequences of Germany’s Mixed-Member System: Personalisation at the grassroots?’, in Shugart, M. and Wattenburg (eds), Mixed-Member Electoral Systems: The best of both worlds?, Oxford: Oxford University Press.  

153

 

Revolution or Evolution? McIntosh, N. (chmn) (1999), The Report of the Commission on Local Government and the Scottish Parliament, Edinburgh: Scottish Office. Pappi, F. and Thurner, W. (2002), ‘Electoral Behaviour in a Two-Vote System: Incentives for ticket splitting in German Bundestag elections, European Journal of Political Research, 41: 207–32. Roberts, G. (1988), ‘The ‘Second-Vote’ Campaign Strategy of the West German Free Democratic Party’, European Journal of Political Research, 16: 317–37. Sanders, D., Burton, J. and Kneeshaw, J. (2002), ‘Identifying the True Party Identifiers: A question wording experiment’, Party Politics, 8: 193–205. Schoen, H. (1999), ‘Split-Ticket Voting in German Federal Elections, 1953–90: An example of sophisticated balloting?’, Electoral Studies, 18: 473–96. Scotland Office (2008), Sub-Constituency Election Results for the 2007 Scottish Parliament Elections. Available at http://www.scotlandoffice.gov.uk/ scotlandoffice/10801.134.html. Scottish Government (2008), Scottish Elections 2007: Scottish government response to the Independent Review of the Scottish Parliament and Local Government Elections 3 May 2007, Edinburgh: Scottish Government. Tajfel, H. and Turner, J. (1979), ‘An Integrative Theory of Intergroup Conflict’, in Austin, W. and Worchel, S. (eds), The Social Psychology of Intergroup Relations, Philadelphia: Taylor and Francis. Thomasson, J. (1976), ‘Party Identification as a Cross-Cultural Concept: Its meaning in the Netherlands’, in Budge, I., Crewe, I. and Farlie, D. (eds), Party Identification and Beyond, London: John Wiley. Vowles, J., Banducci, S. and Karp, J. (2006), ‘Forecasting and Evaluating the Consequences of Electoral Change in New Zealand’, Acta Politica, 41: 267–84.

 

154

Do Voters Care about Parties Any More? APPENDIX TO CHAPTER 7: SAMPLE SIZES FOR TABLES Table 7.5

1999 2003 2007

1999 2003 2007

Completing Ballot Paper Not at all Very/fairly Not very difficult difficult difficult 82 311 429 106 243 379 199 262 265 Understanding Votes into Seats Not at all Very/fairly Not very difficult difficult difficult 360 303 110 323 276 96 328 254 102

Table 7.6

1999 Constituency List 2003 Constituency List 2007 Constituency List

1999 Constituency List 2003 Constituency List 2007 Constituency List  

Completing Ballot Paper Very/fairly Not very Not at all difficult difficult difficult 82 309 427 82 307 427 105 239 373 107 241 378 198 257 264 212 266 270 Understanding Votes into Seats Very/fairly Not very Not at all difficult difficult difficult 360 301 109 359 300 110 319 271 95 325 273 96 325 253 102 339 261 102 155

 

Revolution or Evolution? Table 7.7

1999 Constituency List 2003 Constituency List 2007 Constituency List

Highest Qualification Less than Degree degree 153 854 154 851 126 675 128 683 166 635 169 666

Table 7.8

1999 2003 2007

Strength of Party Identification Very Fairly Not very strong strong strong/none 129 476 401 79 291 417 84 301 394

Table 7.9

1999 Constituency List 2003 Constituency List 2007 Constituency List

 

Strength of Party Identification Very Fairly Not very strong strong strong/none 130 474 398 129 474 397 78 288 412 78 288 422 81 300 388 83 310 406

156

Do Voters Care about Parties Any More? Table 7.12 1999 2003 2007

Party 668 447 443

Motivation for Vote Depended Candidate 239 87 235 111 244 91

Party 668 449 450

Motivation for Vote Depended Candidate 240 90 239 115 248 96

Table 7.13 1999 2003 2007

 

157

 

  CHAPTER 8 __________________________________________

A PERSONAL VOTE? HOW VOTERS USED THE STV BALLOT We established in the previous chapter that while the bonds of party appear to be looser than they once were, it is uncertain whether they are sufficiently weak to ensure that voters will have exploited the opportunities to express a more nuanced and more candidate-centred vote provided by the introduction of the single transferable vote (STV) in 2007. At the same time, we have learnt to be aware of the possibility, however remote, that voters’ ability to use a more complex voting system may be compromised by lack of knowledge or confusion about how to complete the ballot paper or how the electoral system works. In this chapter, therefore, we examine the apparent incidence of more nuanced and candidate-centred behaviour in the way voters completed the STV ballot paper in the 2007 local elections. In so doing we try to establish how far such behaviour seems to be the product of a relative disregard for party politics rather than ignorance or confusion about the system, while also taking into account the impact of the circumstances in which voters were asked to cast their votes – that is, the choices voters were offered. We also consider the partisan consequences of the patterns we uncover. In doing so we will compare our findings with similar research on how people behave in parliamentary elections in the Republic of Ireland (Marsh et al., 2008), where, as we noted earlier, candidate-centred voting appears to be relatively commonplace. We are particularly interested in behaviour that indicates an apparent willingness to exploit the opportunities for more nuanced and more candidate-centred behaviour (Marsh, 2007). Hence we start our examination below by looking at the extent to which voters express multiple preferences. Did voters take the most obvious opportunity to cast a more nuanced vote provided by STV, that is, to express a preference for more than one candidate? If voters simply indicated which candidate they liked most and no more, they would be voting in almost exactly the same manner as they were obliged to do under single-member plurality. The introduction of STV could hardly be regarded as a success if this were all that most voters opted to do.  

158

How Voters Used the STV Ballot Even if they do express a preference for more than one candidate, however, voters could still be confining their choice to the candidates of one party. A party may well decide to nominate more than one candidate in a ward where it believes it has a reasonable prospect of winning more than one seat. In such circumstances even the most partisan of voters will have good reason to vote for more than one candidate. Yet in so doing they might still confine their preferences to candidates standing for their preferred party. Only if voters gave preferences to candidates from more than one party will we clearly be able to argue that they have expressed a more nuanced choice than is possible under single-member plurality. We thus also examine how far this kind of behaviour was in evidence. However, even those who express a preference for candidates of more than one party may still be motivated primarily by their attitudes towards the parties rather than their views of the merits of the individual candidates. They might, for example, have a clear preference for one party above all others. As a result they may rank all of the candidates of the party they like most above those of any other party, and only thereafter give a preference to a candidate standing for any other party. Such behaviour is hardly indicative of significant candidate-centred voting. Only if the order in which voters place candidates is not aligned with the parties for which those candidates are standing can we begin to argue that voters used STV to express a preference for individual candidates rather than just their parties. The second part of our analysis looks at this and other indicators of the relative importance of party and candidate in determining how people voted. Initially, our analysis of these various behaviours focuses on the extent to which they are affected by voters’ sense of partisanship and by the pattern of choices with which they were presented locally. However, as we have already noted, they could also be the result of confusion. Perhaps voters did not vote for more than one candidate or more than one party because they did not understand the new voting system? This possibility is considered briefly in the third part of our analysis. Meanwhile, we have to bear in mind that if a party nominates more than one candidate then even the most partisan of its supporters has to decide the order in which to place those candidates. There is no guarantee that voters will determine that order on the basis of what they think of the individual candidates. As part of an effort to maximise the number of seats that it wins, the political party in question may well issue guidance to  

159

Revolution or Evolution? voters as to how they would like them to order their candidates (Gallagher, 1992; Farrell et al., 1996). Voters might decide simply to follow that guidance. Alternatively voters may be indifferent between the various candidates who are standing for their preferred party and thus take the easy option of placing those candidates in the order in which they appear on the ballot paper (Darcy and McAllister, 1990; Miller and Krosnick, 1998; Brockington, 2003; Koppell and Steen, 2004; King and Leigh, 2009). Such alphabetic voting is clearly neither nuanced nor candidate centred, and its incidence and importance in STV elections has been the subject of considerable dispute (Robson and Walsh, 1974; Marsh, 1981; Ortega Villodres and de la Puerta, 2004; Ortega Villodres, 2008). In a further section we therefore look for signs of alphabetic voting in the 2007 local elections. Each of these analyses are important for any assessment of the degree to which the introduction of STV in Scotland’s local elections might be considered to be a ‘success’, and thereby to have enhanced the personal link between voters and their local representatives. But of course the way that voters behaved under the system could also have had partisan consequences. Perhaps some parties were more successful than others in securing lower preferences from voters who did not back them on the first-preference vote? Maybe the popularity of individual candidates made more of a difference to the ability of some parties’ candidates to secure votes than it did to those standing for other parties? The answers to such questions are clearly of potential importance in assessing the impact of the introduction of STV on Scottish politics, and so the final section of our analysis considers what those answers might be. DATA Before embarking on our analysis, however, we need to describe the specific data used in this chapter. We can only establish both how voters used the ballot paper and why they behaved as they did if detailed information on how individuals completed their STV ballot paper is available in combination with other data about those individuals, such as their party identification, attitudes towards individual candidates and their understanding of the electoral system. To ascertain how voters did complete their ballot, the 2007 Scottish Social Attitudes survey undertook  

160

How Voters Used the STV Ballot an elaborate exercise to collect full details of how its respondents voted in the local elections. This involved giving to each respondent who said they had voted a mock ballot paper that included the names and parties of all the candidates who had stood in their ward, and inviting them to complete that mock ballot in exactly the same way as they had marked their real paper. In order to ensure that each respondent was presented with the correct mock ballot, an exercise was conducted before interviewing commenced to identify the ward in which each address selected for interview was located.1 Altogether some 703 respondents, or 80% of those who said they had voted in the local elections, completed the mock ballot. Between them these respondents lived in 137 of the 353 multimember wards in which the local elections were contested. The data gathered in this way have their disadvantages. In particular, voters may not recall precisely how they voted, whether through forgetfulness or fatigue. The incidence of lower preferences may particularly be underestimated as a result. Meanwhile, some of the information about use of the ballot paper in which we are interested, most notably the number of preferences expressed by the typical voter, can be discerned from the detailed records of the electronic counts that have been made publicly available by most local authorities (Electoral Reform Society, 2008). However, none of the information that can be gleaned from the published details of the counts can be linked to other pieces of information about individual voters, such as their party identification or understanding of the electoral system. In contrast, this is possible with our survey data, and this capacity is essential if we are to ascertain why voters behaved as they did. PREFERENCES AND PARTIES The first matter to be explored is whether voters used the opportunity afforded by STV to express more than one preference and whether in so doing they included candidates from more than one party amongst their preferences. In Table 8.1 we show the number of preferences our respondents indicated on their mock ballot paper, and across how many parties those preferences were spread. It shows that nearly four in five voters cast more than one preference, and that the average respondent cast 2.5 preferences. This latter figure is a little below what we might  

161

Revolution or Evolution? have anticipated from analysis of the data published from the electronic counts, which shows that the median voter expressed three preferences (Electoral Reform Society, 2008: 69). However, at 22%, our estimate of the proportion of voters who cast only one preference is very close to the 20% figure indicated by the electronic count data. Our data thus do not appear to be seriously in error. Meanwhile it seems we can conclude that most voters did indeed take the opportunity afforded by STV to make a more nuanced choice by indicating more than one preference. Table 8.1 Number of Candidates and Parties Voted for

One Two Three More than three

No. of candidates % 22 34 30 14

Mean

2.5

No. of parties % 41 36 14 9 1.9

Sample size 703. Note: Independents counted as a party. Source: Scottish Social Attitudes, 2007.

However, many of those who cast more than one vote evidently still confined their preferences to the candidates of one party. At 41% the proportion of people who expressed support for candidates of only one party was nearly twice the proportion that cast only one preference. Even so, evidently a majority of voters did something that they would be unable to do under single-member plurality, that is to indicate support for candidates of more than one party on the same ballot paper. Doing so was certainly a far more common practice than was splitting support between two parties on the two votes provided by the Additional Member System in the Holyrood election, a practice that we saw in the previous chapter was engaged in by just under a quarter of voters. Still, expressing more than one preference or backing candidates from more than one party was less common than might have been expected from Ireland’s experience. There, equivalent mock ballot data for the 2002 general election show that just 6% cast just one preference, while only 19% confined their preferences to candidates of one party. It seems that, as we anticipated  

162

How Voters Used the STV Ballot might be the case, the commitment to parties in Scotland was sufficient for voters to be markedly less likely than their counterparts in Ireland to take up the opportunity afforded by STV to express a more nuanced choice. This presumption is examined in Table 8.2. This compares the level of multi-preference and multiple party voting amongst those with a stronger and a weaker sense of party identification, and according to the mark out of ten that respondents gave to the party of their first-preference vote when asked how much they liked (or disliked) them. This analysis suggests that in fact the number of preferences voters expressed had relatively little to do with the strength of their attachment to or their liking for a political party. On the other hand the more a voter liked the party of their firstpreference vote the less likely they were to give a preference to any other party. To that extent at least, a strong sense of partisanship does seem to have inhibited the expression of a more nuanced choice. Table 8.2 Voting for Multiple Candidates and Parties by Partisanship

Strength of party identification Fairly/Very Strong Not Strong/None Likeability Score of first-preference party 9/10 7/8 0–6

Cast more than one preference %

Voted for more than one party %

79 76

56 61

329 374

75 82 79

45 61 72

131 228 146

Sample size

Data for Likeability Score of first-preference party is based only on those who cast their first-preference vote for a major party candidate. Source: Scottish Social Attitudes, 2007.

In any event, we should bear in mind that how voters behave under STV may depend not only on the motivations that they bring to the ballot  

163

Revolution or Evolution? box, but also on nature of the choice with which they were presented on the ballot paper. In this respect there were some important differences between the choice faced by voters in the Scottish local elections and the circumstances that pertain in Ireland. First of all, we should remember that even the most partisan of voters has more reason to express more than one preference if the party they support nominates more than one candidate. However, none of the wards used in Scotland elected more than four members, thereby reducing the likelihood of multiple candidatures. In elections to the Dáil in 2002, in contrast, a dozen of the 42 constituencies elected five members. Coupled with a tendency by the parties in Scotland to pursue a rather conservative strategy in deciding how many candidates to nominate (Electoral Reform Society, 2008), no less than 56% of voters in the 2007 local elections found that the party of their first-choice candidate only nominated one candidate in their ward.2 Second, voters have more reason to express a preference for more than one party if it seems possible that the party they like most might fail to win a seat locally. Otherwise they run the risk that their vote will not contribute to the election of any candidate. However, because the party system is less fractured in Scotland than in Ireland, Scottish voters were less likely to find themselves in that situation than were their counterparts in Ireland. As can be seen in Table 1.2 in Chapter 1 above, all four of the parties that dominate representation in Scotland won an eighth of the local vote or more, while Independent candidates were often popular locally. In contrast, in the 2002 Irish election only two of the six parties that dominated representation won more than one eighth of the vote. Just how far the presence of these two sets of circumstances discouraged voters in Scotland from expressing more than one preference or supporting more than one party is shown in Table 8.3. First, in the upper half of the table we can see that those whose first-preference vote went to a candidate of a party that put forward more than one nominee were more likely to express at least two preferences. At the same time, however, they were less likely to give a preference to more than one party. Meanwhile, in the lower half of the table we demonstrate that those living in wards where it was quite possible that the party of their firstpreference candidate would fail to secure a seat (defined as a ward where that party’s share of the first-preference vote constituted less than 86% of the vote required to be certain of winning a seat) were far more likely to give a preference to more than one party, and especially so as compared  

164

How Voters Used the STV Ballot with voters whose party of their first-preference vote looked potentially capable of winning more than one seat locally. Table 8.3 Number of Candidates and Parties Voted for by Characteristics of First-Preference Party

Voted for more than one candidate Voted for more than one party Sample size

Voted for more than one candidate Voted for more than one party Sample size

No. of candidates standing for first-preference party More than One one % % 67 86 67

52

396

307

Share of first-preference vote in ward of first-preference party < 0.86 quota 0.86–1.86 > 1.86 quotas % % % 76 75 84 75

59

44

162

347

194

1.86 quotas. The quota is calculated as 1+(Valid votes/seats+1). Source: Scottish Social Attitudes, 2007.

Some voters, then, did exploit the opportunity afforded by STV to express a more nuanced choice, but the degree to which voters as a whole did so was rather less than we might have anticipated, given experience of the system in Ireland. So far as willingness to express a preference for candidates of more than one party is concerned, this would appear to reflect the fact that many voters in Scotland still feel a degree of exclusive partisanship. In addition, the way in which people voted reflected the  

165

Revolution or Evolution? circumstances in which many voters found themselves. In combination the two sets of influences served to produce a rather more partisan looking pattern of voting than perhaps some advocates of the switch to STV might have been expecting. SOLID SUPPORT? We have seen how many preferences voters cast and how many parties they voted for, and shown how this behaviour is related to partisanship and the choices offered to each voter. A more direct measure of how far voters’ use of the ballot paper was determined by their sense of partisanship can be obtained by looking more closely at the behaviour of those whose party of their first-preference vote nominated more than one candidate locally. If voters are truly partisan, we can expect them to place all of the candidates of the party they like most above the candidates of any other party for whom they might be willing to express some support. In the event just over half (57%) of those voters whose first-preference candidate had at least one running mate placed all of those running mates above the candidates of any other party – if indeed they gave any kind of preference to a candidate from another party at all. This compares with an equivalent figure of just 45% in the 2002 Irish general election. Clearly party was the principal cue for many voters in the Scottish local elections, and especially so, as we would anticipate, amongst those who strongly liked the party of their first-preference candidate (see Table 8.4).3

 

166

How Voters Used the STV Ballot Table 8.4 Loyal Party Voting by Partisanship Voted for all candidates of one party above any others % Strength of party identification Fairly/Very Strong Not Strong/None Likeability Score of first-preference party 9/10 7/8 0–6

Sample size

60 53

163 144

73 61 32

75 132 65

Table confined to those whose first preference was given to a candidate of a party that nominated two or more candidates. Source: Scottish Social Attitudes, 2007.

Nevertheless, many voters evidently did not simply follow the dictates of party loyalty, implying that for some at least what they thought of one or more candidates did make a difference to how they voted. Additional evidence that the use of STV did encourage some voters to vote for candidates, rather than just parties, comes from the answers that our survey respondents gave to two further questions we asked about their voting behaviour in the local elections. First, in referring to the candidate to whom respondents gave their first preference vote, the survey asked: If this candidate had been running for any of the other parties would you still have given a first-preference vote to him/her? As many as 59% said they would not have done so. However, just over a quarter (27%) indicated that they would still have voted for the same candidate, while just over one in ten (11%) said that it would have depended on which was the other party for which they were standing. Unsurprisingly, those who voted for an Independent candidate were particularly likely to say they would still have given their first-preference vote to the same candidate; as many as 64% said they would have done so.  

167

Revolution or Evolution? But even amongst those who did vote for a candidate that stood for a party, just under a quarter (23%) still said they would have made the same choice. A further piece of evidence comes from a question to which an introduction has already been given in the previous chapter. In Table 7.11 in that chapter we showed that 53% of people said they voted on the constituency ballot of the parliamentary election ‘for a party, regardless of the candidate’. In contrast, 10% said that they ‘voted for a candidate, regardless of his or her party’, while another 27% reported a mixture of the two motives, that is they responded that they ‘voted for a party only if I approved of the candidate’. When asked the same question about their first-preference vote in the local elections, just 47% said they voted ‘for a party, regardless of the candidate’, while as many as 17% indicated they voted for a candidate, regardless of party. (At 31%, the proportion reporting a mixture of motives was also a little higher than in the parliamentary election.) Of course the greater incidence of candidatebased voting in the local elections as compared with the parliamentary election could be accounted for by the different nature of the two contests; voters might be more concerned about the qualities of the individual candidates in local elections anyway. But the contrast is at least consistent with the proposition that the introduction of STV encouraged some voters who might not otherwise have done so to vote on the basis of candidate rather than party. CONFUSION? But while the introduction of STV may have enabled some voters to express a choice that reflected their views about the individual candidates, rather than just the parties, evidently there were plenty for whom this was not the case. Perhaps the failure of some voters to exploit the opportunities afforded by STV may have reflected a lack of understanding of the new system, if not indeed a degree of confusion, rather than reflecting a strong sense of partisanship. It is to this possibility that we now turn. Our approach to this issue here is much the same as it was in the previous chapter. One of the ways in which we looked for signs of possible confusion was by comparing the incidence of vote splitting  

168

How Voters Used the STV Ballot amongst those holding a degree with that amongst those without such a qualification. We also considered the possibility that confusion might have been more common amongst those who reported difficulty in completing the ballot paper or in understanding how votes were converted into seats. Analogous questions about the difficulty of the local election ballot paper and about the allocation of seats under STV were also included in our survey (see Table 8.5). Table 8.5 Perceptions of Difficulty of STV Ballot Paper and Understanding Allocation of Seats

Very difficult Fairly difficult Not very difficult Not at all difficult Sample size

Ballot Paper % 6 22 33 36 787

Seat Allocation % 8 35 35 12 787

Table based on those who reported voting in the parliamentary election. Source: Scottish Social Attitudes, 2007.

As in the case of the parliamentary election, the ballot paper was less likely to be regarded as difficult to complete than the method of allocating seats was thought to be difficult to understand. Indeed, as a comparison with the results in Table 7.3 above shows, voters were actually slightly less likely to regard the STV ballot as difficult to complete than they were the new parliamentary ballot. Moreover, they were no more likely to regard the STV method of allocating seats as difficult to understand than they were the method used in the parliamentary election.

 

169

Revolution or Evolution? Table 8.6 Use of STV Ballot Paper by Highest Educational Qualification and Perceived Difficulty of Electoral System % within each category who Only Sample Voted for all voted size candidates of for one party one above any Party other Highest Educational Qualification Degree 20 35 142 63 Professional 21 39 108 53 Higher 16 38 99 60 Standard Grade 23 43 109 58 1–3 Standard Grade 23 40 65 40 4–6 None/foreign 30 52 180 61 Perceived Difficulty of Completing Ballot Paper Very difficult 45 65 32 54 Fairly difficult 23 38 124 52 Not very 21 39 224 56 difficult Not at all 20 42 252 64 difficult Perceived Difficulty of Understanding Seat Allocation Very difficult 33 51 54 63 Fairly difficult 24 43 232 54 Not very 21 38 222 52 difficult Not at all 14 36 89 65 difficult Only cast one preference

Sample size

55 51 43 55 31 72 14 54 95 110 24 86 95 49

Source: Scottish Social Attitudes, 2007.

But while the STV ballot paper and system for allocating seats were no more likely to be regarded as difficult than their equivalents in the parliamentary election, there is some sign that those who felt the paper or the system was difficult were rather less likely to use their ballot paper to  

170

How Voters Used the STV Ballot express a more nuanced choice. As Table 8.6 shows, those who thought the ballot paper was ‘very difficult’ to complete were more likely to express only one preference and to vote for only one party. Much the same is true of those who thought the method of allocating seats was ‘very difficult’ to understand. Of course, it should be borne in mind that only 6% and 8% of voters respectively fall into these two groups. Even so, there is also some sign that those with no educational qualifications at all were most likely to confine themselves to a single candidate or party. This group does seem to have been less willing to exploit the opportunity to express a more nuanced choice afforded by the STV ballot paper, though, given the other evidence in Table 8.6, perhaps this reflected a lack of motivation to do so rather than widespread confusion amongst its members. However, there is no clear link between the perceived difficulty of the ballot paper or the system of allocating seats and voting for all of the candidates of one party above those of any other. Equally there seems to be no association between the pattern of voting in that way and educational qualification. Those who put the candidates of one party above those of all others seem to have been reflecting their commitment to that party rather than any misunderstanding of the electoral system. Overall, then, it seems that confusion or misunderstanding did not play a major role in inhibiting voters from expressing a more nuanced or more candidate-based choice. ALPHABETIC VOTING Still, as we noted earlier, even the most partisan of voters has a choice to make when faced with two or more candidates who have all been nominated by the party they support. They have to decide which of those candidates they should put first. In this respect at least it would seem that they still have to take the perceived qualities of the candidates into consideration. However, this need not be the case. As we noted earlier, such voters might follow advice disseminated by the party itself as to the order in which its supporters should place its candidates, advice designed to ensure that the distribution of first preferences maximises that party’s chances of winning seats. Alternatively, loyal party voters could simply take a short cut and vote for the candidates in the order in which they appear on the ballot paper.  

171

Revolution or Evolution? In fact less than one in five voters (18%) living in a ward where the party of their first-preference vote nominated more than one candidate remembered receiving advice from that party about the order in which they should place its candidates. This compares with an equivalent figure of 30% in the 2002 Irish election. It seems the parties in Scotland were not particularly active in trying to manage the vote in their strongholds. Even so, those voters who did receive such advice appear to have been ready to follow it. No less than 62% claimed to have done so, far higher than the equivalent figure of 34% in the 2002 Irish general election. It would seem there is scope for the parties in Scotland to exert more influence on the distribution of the vote between their candidates than they managed to achieve in 2007. So in the absence of guidance from the parties, did voters simply follow the order of the candidates on the ballot paper? Overall, amongst all those voters who cast more than one preference, 60% gave their second preference to a candidate placed lower down the ballot paper than the candidate to whom they gave their first preference. If voters had not taken any account of the order of the candidates on the ballot paper, we would have expected this figure would have been close to 50%, as indeed it was in Ireland in 2002 (51%). It is not immediately obvious, however, that such alphabetic voting was the product of confusion or misunderstanding about the electoral system. As we see in Table 8.7, those who said they thought either the ballot paper was ‘very difficult’ to complete or the allocation of seats was similarly difficult to understand were in fact particularly unlikely to have put their first two preferences in the order in which the candidates appeared on the ballot paper. Meanwhile those without any educational qualifications were also, if anything, relatively unlikely to put their first two preferences in alphabetical order.

 

172

How Voters Used the STV Ballot Table 8.7 Incidence of Alphabetic Voting in Scotland by Indicators of Capacity Voted down ballot Sample size paper % Highest Educational Qualification Degree 58 112 Professional 63 87 Higher 60 82 Standard Grade 1–3 71 84 Standard Grade 4–6 61 48 None/foreign 52 122 Perceived Difficulty of Completing Ballot Paper Very difficult 34 20 Fairly difficult 58 94 Not very difficult 63 172 Not at all difficult 57 202 Perceived Difficulty of Understanding Seat Allocation Very difficult 42 36 Fairly difficult 54 176 Not very difficult 60 172 Not at all difficult 62 75 Source: Scottish Social Attitudes, 2007.

One group that was very likely to put their top two preferences in the order in which they appeared on the ballot paper consisted of those partisan voters who gave a preference to all of the candidates of one party above those of any other. No less than 70% of this group behaved in that way. In contrast, amongst all other voters just 56% did so, only a little above the 50% figure that would be expected if the order of the candidates on the ballot paper did not make any difference to the way that voters behaved. We have already seen (Table 8.4 above) that those who loyally put all the candidates of one party above those of any other tended to be those who especially liked the party of their first-preference vote. Meanwhile, when asked how much they liked the candidates to whom they gave their first and second preference, those loyal party voters who placed their top  

173

Revolution or Evolution? two candidates in alphabetical order on average only gave their firstpreference candidate a mark 0.7 higher than the score they gave to their second-preference candidate. This is little more than half the equivalent difference (1.3) amongst those who placed those candidates in reverse order. It seems that for many loyal party voters the qualities of the particular candidates mattered so little that, often lacking any advice from the relevant party as to the order in which to place its candidates, they simply ranked them in the order in which they appeared on the ballot paper. As a result, as has been noted elsewhere, where a party nominated two or more candidates, the one who appeared lower down the ballot paper was unlikely to be elected (Denver, 2007). Evidently, if some voters did use the opportunity provided by STV to express a candidatebased choice then apparently for others it was an opportunity in which they appeared to have no interest at all. PATTERN OF TRANSFERS We have established that some voters did take their evaluations of the candidates into account in deciding how to vote, while a majority certainly gave a preference to candidates from more than one party. True, these behaviours might not have been as common as in Ireland, but they appear to have occurred on a sufficient scale for it to be worth investigating the possible partisan consequences of the way in which people used their ballot paper. Did any party particularly benefit from the willingness of some voters to take into account the perceived qualities of the candidates? And did any party particularly gain or lose from the pattern of second and subsequent preferences? Of course the first ambition of any party is to try and ensure that its supporters are loyal to its candidates, giving all of them a preference and preferably above the candidates of any other party. Table 8.8 shows how successful each of the four main parties were in that regard in those wards where they nominated more than one candidate. As we would anticipate from our earlier analysis, it shows that in most cases the parties were relatively successful; typically more than half of those voters who gave a first-preference vote to one of their candidates went on to vote for the remainder. However, this was not true of those voters who put a Conservative candidate first. Only a third of them gave any kind of preference to that candidate’s Conservative colleagues.  

174

How Voters Used the STV Ballot Table 8.8 Solidity of Party Support by Party of First-Preference Vote Party of firstpreference vote Conservative Labour Liberal Democrat SNP

% voting for all candidates of that party Above any other As any Sample size party’s candidate preference 31 33 27 57 60 146 58 70 40 60 69 94

Table includes only voters whose first-choice party nominated multiple candidates. Source: Scottish Social Attitudes, 2007.

However, we may need to be cautious in presuming that this trend was wholly disadvantageous for the Conservatives. This is apparent if we look at the pattern revealed in Table 8.9. In the first column of this table we show for each set of party supporters the difference between the mark (out of ten) they gave when asked how much they liked the party of their first-preference vote and the mark they indicated when asked how much they liked the candidate to which they gave their first preference. The equivalent information is also shown for each voter’s second-preference vote. Those who gave a first preference to a Conservative candidate clearly stand out. On average they gave a higher score to the candidate for whom they had voted than they gave to the Conservative party. In contrast, those whose first-preference vote was for a Liberal Democrat candidate liked the candidate as much as the party, while those who backed either a Labour or a SNP candidate typically rated the party for which they voted more highly than the individual candidate.

 

175

Revolution or Evolution? Table 8.9 Relative Liking of Parties and Candidates

Party supported on that preference Conservative Labour Liberal Democrat SNP All main party voters

Candidate like/dislike score – party like/dislike score for candidate and party supported on First preference Second preference Score Sample size Score Sample size 0.57 95 -0.54 48 -0.61 166 -1.30 104 0.09 81 0.03 37 163 -1.28 91 -1.29 -0.56

505

-0.96

280

Source: Scottish Social Attitudes, 2007.

This difference suggests that the reason why those who gave their first-preference vote to a Conservative candidate did not necessarily support any running mate of that candidate is that Conservative candidates were unusually reliant in winning first-preference votes on personal support for themselves as individuals. In other words, many a Conservative candidate in the local elections won support despite, rather than because of, their party. In the absence of such popular candidates, the Conservatives would probably have fared even less well than they actually did. Personal support may also have been relatively important to the ability of Conservative candidates to secure support on the secondpreference vote too. Unsurprisingly, perhaps, amongst voters in general, liking for a party seemed to matter relatively more in determining how they cast this vote than it did on the first-preference vote – although intriguingly this does not seem to be true of those who gave a second preference to a Liberal Democrat candidate. But how much a voter liked a Conservative candidate relative to his or her party still seemed to matter rather more than did how much they liked a Labour or SNP one. While all parties clearly have an interest in trying to persuade their supporters to rank their candidates above those of any other party, the stance they might adopt towards the possibility that their supporters might thereafter give a preference to one or more candidates from other parties is less obvious. There is certainly no reason why parties should discourage  

176

How Voters Used the STV Ballot such behaviour; such preferences cannot reduce the chances of their own candidates securing election. A party might, though, have good reason to encourage its supporters to give their lower preferences to candidates of one particular party rather than to those of any other if there is an understanding that a coalition might be formed with that party once the election is over. This kind of behaviour could well increase the chances of that coalition coming to power (Bogdanor, 1981: 245). Understandings of this kind were, however, not a feature of the 2007 local election campaign. Table 8.10 Number of Parties Voted for by First-Preference Vote First-Preference Vote No. of parties voted for

Conservative

Labour

Liberal Democrat

SNP

% 32 50 12

% 48 32 12

% 29 38 25

% 42 36 12

6

8

8

10

Mean

1.9

1.8

2.2

2.0

Sample size

112

184

94

188

One Two Three More than three

Note: Independents counted as a party. Source: Scottish Social Attitudes, 2007.

In these circumstances, it should perhaps come as little surprise to discover in Table 8.10 that those who gave their first-preference vote to a Labour or SNP candidate were less likely than those who backed a standard bearer for the Conservatives or the Liberal Democrats to give any kind of preference to a candidate from another party. Such voters were after all more likely to be presented with more than one candidate from the party of their first-preference vote. Equally they were less likely to be voting in a ward where the party of their first-preference vote might well fail to secure the election of a candidate. And as we saw in Table 8.3,  

177

Revolution or Evolution? these were precisely the circumstances in which voters in general were less likely to express support for candidates from more than one party. But what happened when voters did cross the party divide and give a lower preference to a candidate from a party other than the one they backed with their first-preference vote? Table 8.11 addresses this question by showing the party to which each group of first-preference supporters transferred most immediately after backing candidates of the party of their first preference. This reveals two striking patterns.4 First, those who backed a Labour, SNP or Liberal Democrat candidate with their firstpreference vote were all relatively reluctant to express support thereafter for a Conservative candidate. In each case only around one in twenty did so. Second, those who gave their first preference to a Labour, SNP or Conservative candidate were all relatively likely to express a subsequent preference for a Liberal Democrat candidate. Table 8.11 Transfer Structure of Voters’ Preferences

Labour

Liberal Democrat

SNP

Independent

Other

None

Sample size

Party of firstpreference vote Conservative Labour Liberal Democrat SNP

Conservative

% second-preference party

5 6

15 19

27 22 -

11 15 23

9 1 8

6 9 13

32 46 31

81 119 80

4

21

21

-

9

8

37

132

A voter’s second preference party is the party for one of whose candidates that voter gave the highest preference after voting for one or more of the candidates of the party of their first-preference vote. Table confined to respondents in wards where all four main parties standing. Source: Scottish Social Attitudes, 2007.

We have already seen that Conservative candidates appeared to have been unusually reliant on their personal popularity to secure votes. Now we have uncovered further evidence of the relative unpopularity of the Conservative party in Scotland. Indeed, when asked how much they liked or disliked the party, no less than 13% of our respondents gave the party a  

178

How Voters Used the STV Ballot score of zero. No other party came close to this figure, the next highest proportion being the 8% who gave the SNP that score. Such a ‘pariah’ status is a distinct disadvantage indeed when it comes to securing transfers from other parties at later stages of an STV count. In contrast, only 4% said they felt that way about the Liberal Democrats, no more than said the same about Labour despite the latter’s much higher level of support. So while few voters may have been enthusiastic about the Liberal Democrats, not many were hostile towards them. They thus were relatively willing to give a lower preference to any Liberal Democrat candidate who might have made a favourable impression. It was this willingness that in part at least explains why the Liberal Democrats emerged from the election with more seats than the Conservatives even though they won a lower share of the first-preference vote (Electoral Reform Society 2007).5 CONCLUSION Voters in the 2007 local elections were much less likely to use their STV ballot paper to express a more nuanced and more candidate-centred vote than we might have anticipated from the experience of Ireland. In particular, only around three in five took the opportunity to express a preference for candidates from more than one party, while some of those who confined their preferences to a single party had so little interest in that party’s individual candidates that they simply ranked them in the order in which they appeared on the ballot paper. Attachment to parties may be weaker than it once was in Scotland, but for some at least it evidently still matters rather a lot. Yet attachment to party was not so strong that the introduction of STV made no difference at all to the way that voters behaved. A majority of voters did express a degree of support for more than one party, something that it was impossible for any voter to do under the previous singlemember plurality system and which was also far less common under the Additional Member System used in the parliamentary election. Meanwhile it seems that, for a significant minority of voters, the way in which they voted did not simply reflect what they thought of the parties, but also their views about the individual candidates. Indeed if this were not the case many a Conservative candidate in particular would have struggled even more than they did to escape the consequences of their party’s  

179

Revolution or Evolution? relative unpopularity. Personal popularity may well not be sufficient to ensure electoral success under STV in Scotland, but it evidently is a potential asset that local councillors and candidates cannot afford to ignore. NOTES 1. This would not have been possible but for the generous assistance of the Local Government Boundary Commission for Scotland, who on our behalf matched the addresses of our issued sample to the correct ward (each of which had been newly created by the commission for this election). Once this linkage had been made, details of the candidates standing in each ward being covered by the survey were obtained from published sources, most notably the election web site maintained by The Herald newspaper. See http://election.theherald.co.uk/nomination2007/. 2. Though this figure falls to 46% if having two or more Independent candidates standing locally is regarded as analogous to a party nominating two or more candidates. 3. Though once again those with a relatively strong party identification were perhaps not as distinctly partisan in their behaviour as we might have anticipated. 4. These patterns are equally present if we look at the party to which each group of voters were willing to give any preference rather than just the party to whom they gave their next highest preference. See Curtice and Marsh (2009). 5. The parties’ respective fortunes were, however, also the product of differences in the average size of the wards in which they performed well. REFERENCES Bochel, H. and Denver, D. (2007), ‘A Quiet Revolution: STV and the Scottish council elections of 2007’, Scottish Affairs, 61: 1–17. Bogdanor, V. (1981), The People and the Party System, Cambridge: Cambridge University Press. Darcy, R. and McAllister, I. (1990), ‘Ballot Position Effects’, Electoral Studies, 9: 5–17.  

180

How Voters Used the STV Ballot Electoral Reform Society (2007), Local Authority Elections in Scotland: 3 May 2007: Report and analysis, London: Electoral Reform Society. Farrell, D., MacKerras, M. and McAllister, I. (1996), ‘Designing Electoral Institutions: STV systems and their consequences’, Political Studies, 45: 24–43. Gallagher, M. (1992), ‘Comparing Proportional Representation Electoral Systems: Quotas, thresholds, paradoxes and majorities’, British Journal of Political Science, 22: 469–96. King, A. and Leigh, A. (2009), ‘Are Ballot Order Effects Heterogeneous?’, Social Science Quarterly, 90: 71–87. Koppell, J. and Steen, J. (2004), ‘The Effects of Ballot Position in Election Outcomes’, Journal of Politics, 66: 267–81. Marsh, M. (1981), ‘Electoral Preferences in Irish Recruitment: The 1977 Irish election’, European Journal of Political Research, 9: 61–74. Marsh, M. (2007), ‘Candidates or parties? Objects of electoral choice in Ireland’, Party Politics, 13: 500–27. Miller, J. and Krosnick, J. (1998), ‘The Impact of Candidate Name Order on Election Outcomes’, Public Opinion Quarterly, 62: 291–330. Ortega Villodres, C. and de la Puerta, B. M. G. (2004), ‘Position Effects under STV: Ireland and Malta’, Representation, 41: 3–14. Ortega Villodres, C. (2008), ‘Gender and Party Duopoly in a Small State: Ballot position effects under the single transferable vote in Malta, 1947–2008’, Southern European Society and Politics, 13: 435–56. Robson, C. and Walsh, B. (1974), ‘The Importance of Positional Voting Bias in the Irish General Election of 1973’, Political Studies, 22: 191– 203.

 

181

  CHAPTER 9 __________________________________________

CONCLUSION

Throughout this book we have been engaged in a simple but important task – to assess the significance of the 2007 Scottish parliamentary and local elections. Did the electoral success of the SNP in the parliamentary election and the first use of the single transferable vote in the local elections constitute a revolutionary break from the past? Or are these developments better regarded as evolutionary in character, leaving much of the existing structure of Scottish politics still in place? In this concluding chapter we bring together the threads of our analyses to consider in each case whether the revolutionary or the evolutionary perspective is the more appropriate, and to examine the apparent implications of our interpretation. THE SNP’S SUCCESS The success of the SNP in the parliamentary election certainly had all the hallmarks of a revolution. Never before had the SNP won a nationwide contest; moreover, the party came to power on the back of a promise to bring to an end Scotland’s 300-year membership of the Union. Not only was the outcome of the election unprecedented; it appeared that it might pave the way for a revolutionary change in Scotland’s constitutional status. We seem, however, to have uncovered a curious paradox at the heart of the SNP’s success. At the very moment when it finally secured the reins of office, support for independence appears, if anything, to have been weaker than it had been at any time since the advent of devolution a decade earlier. The party’s success in the ballot box did not seem to signify an unprecedentedly high level of public discontent with the Union or a dramatic erosion of popular support for its maintenance. Instead it seems to have rested on hopes that devolution could be made to deliver more for Scotland than it had apparently done so far. In truth, the absence of any groundswell for independence was recognised by the party in the manifesto on which it campaigned (Scottish  

182

Conclusion National Party, 2007). Instead of arguing that Scotland should make a quick dash towards independence, the party proposed only holding a referendum on the subject during the last of its four years in office. It believed the Scottish public would embrace independence only after it had seen that the SNP could provide Scotland with effective government. As its manifesto acknowledged, the party needed to ‘earn the trust and support of the people of Scotland’. Yet if the SNP’s success was not an indication of a fundamental shift in favour of independence amongst the Scottish public, it represented more than a commonplace example of the regular ebb and flow of democratic life. Voters were not simply expressing their discontent with the incumbent Labour/Liberal Democrat coalition and looking to the opposition for more competent everyday management of the country’s affairs (for an alternative view see Johns et al., 2009). We have found little evidence that Labour’s defeat was due to the unpopularity of its performance in office in Edinburgh, or to a mismatch between its policy proposals and public preferences. Equally, while the perceived record and stance of the UK government in London hardly did the party any favours, Labour’s defeat cannot simply be blamed on the existence of a widespread protest vote against an unpopular UK government. Rather, there appears to have been a deeper sense of restlessness amongst the Scottish public. Many were still wondering what material benefit devolution had brought to their country. Many too had a persistent feeling that the devolved institutions were not as powerful or as influential as they had expected them to be. While most people may not have been ready to embrace independence, they did feel Holyrood was the place where most decisions about their country’s domestic affairs should be taken; and the impression they had formed was that in practice it was London that often still had the larger say. But above all, this sense of restlessness concerned the style of devolved government. For Labour, the dominant party in Edinburgh and in London during the first eight years, devolution was about a ‘partnership’ between the Scottish administration in Edinburgh and its UK counterpart in London. When necessary, Scotland’s case could be promoted discreetly within the corridors of Whitehall. When beneficial, the devolved institutions might pursue a somewhat different policy path from the rest of the UK. But open disputes and disagreements between the two administrations were best avoided, not least because of the embarrassment it might cause when the party was in power on both sides of the border.  

183

Revolution or Evolution? Such discretion, however, did little to persuade people that the devolved institutions were promoting Scotland’s cause as effectively as they might. As a result, devolution had failed to achieve what many people regarded as one of its key purposes – to ensure that Scotland’s distinctive identity, needs and interests are properly promoted and defended within the United Kingdom. Our research shows that no less than nine out of ten people in Scotland believe the devolved institutions should be willing to air their disagreements with London in public. We have also found that devolved elections in general are occasions when voters focus on developments within Scotland in particular rather than across the UK as a whole. Equally, they are occasions when people’s sense of Scottish identity is particularly likely to be reflected in how they vote. Indeed, it is this distinctive character of voting behaviour in devolved elections that provided the crucial foundation for the SNP’s success in the 2007 parliamentary election. When asked in devolved elections to consider who and what they think would be best for Scotland, some people persistently give a different answer than when they are asked, as in UK general elections, who can best govern the UK as a whole. In particular, the answer they give to the question posed by devolved elections has persistently been more likely to be ‘the SNP’, both as a way of expressing their distinctive Scottish identity and of ensuring that they have a devolved government willing to stand up for Scotland’s particular interests. Scottish parliamentary elections are far from unique in displaying this character. Nationalist parties tend in general to perform relatively well in sub-state elections where a distinctive sense of national identity exists, as, for example, in Catalonia (Jeffrey and Hough, 2003). This broader international experience suggests there was never much prospect that devolution would kill political nationalism ‘stone dead’. Rather, it was always likely to create an environment in which the SNP could flourish more easily. Even so, the more propitious environment created by devolved elections had still not been sufficient to enable the nationalists to displace Labour as the leading party in the country’s affections either in 1999 or 2003. Only in 2007 was the SNP able sufficiently to exploit the more favourable climate created by devolved elections. It was able to do so this time because the party was successful to an unparalleled degree in mobilising its potential vote, by persuading those who already supported  

184

Conclusion independence (or at least more powerful devolved institutions) and who already had a strong sense of Scottish national identity to support the party in hitherto unprecedented numbers. In 2007 people’s sense of identity and their constitutional preferences were reflected in the ballot box as never before. In this, the regular ebbs and flows of party politics did play a role. The re-emergence in 2004 of the highly charismatic Alex Salmond as the SNP’s leader was probably crucial to the nationalists’ success. Mr Salmond gave his party an air of assurance that his Labour opponent, Jack McConnell, could not match, not least because the latter found it difficult to emerge from under the shadow of the party’s UK leadership. Meanwhile the SNP focused its election campaign not on the case for independence, but rather on selling itself as a potential alternative government. Although the SNP’s efforts did not impress everyone, they apparently did register with those who were already broadly sympathetic to the party’s outlook and image. Circumstances thus conspired to give the SNP its chance to turn that sympathy into real support in the ballot boxes. If this analysis is correct, and we believe it is, the SNP’s success in 2007, historic as it was, marked a further evolution of the existing pattern of Scottish politics rather than a revolutionary break with the past. Scotland’s distinctive sense of national identity and wish for looser ties to London had been in place long before the 2007 election. The first two devolved elections had already demonstrated that national identity and constitutional preference had a greater effect on how people voted in the new electoral arena. In 2007 that last tendency was simply in evidence to an even greater degree – thanks to a circumstantial, commonplace battle over who could provide Scotland with more effective leadership. However, even if one accepts that the SNP’s electoral success was evolutionary in character, might not its accession to office still prove revolutionary in its consequences? Might not the SNP be correct in its presumption that the Scottish public will come to embrace independence once the party has demonstrated that it can provide Scotland with effective government? The party has certainly used the opportunity of office to put the issue of Scotland’s constitutional future back on to the agenda, though there is little evidence so far that this has aroused the public’s imagination. Of course, any speculation about the future course of events on the basis of data collected at the beginning of a government’s time in office must be  

185

Revolution or Evolution? treated with caution, but our analysis suggests no rising tide of opinion favouring the SNP’s ultimate goal of independence. After all, we have seen that the SNP’s success was based on a degree of frustration with the operation of devolution rather than a loss of faith in the principle of relative autonomy within the framework of the Union. Free of any ties of political loyalty to the incumbent administration in London, the devolved SNP government is able to air its disagreements with the UK government in public, and it has fewer inhibitions, either, in promoting Scottish identity and culture. In short, the party has adopted a style of government that could persuade people in Scotland that their country’s distinctive interests and identity can be promoted and defended effectively within the framework of the Union. As we have seen, the initial reaction of the public to the advent of the nationalist government in power was increased satisfaction with Scotland’s place in the Union. In contrast to previous surveys, a clear majority now felt that the creation of the Scottish Parliament had strengthened Scotland’s voice in the Union. For the first time, almost as many people felt that Scotland benefited economically more from the Union than England as believed the reverse was true. It would seem that in providing effective devolved government, the SNP might, paradoxically, strengthen support for the Union rather than undermine it. Still, meeting the public’s apparent dissatisfaction with the practice of devolution might be felt to require more than a change of governing style. Perhaps the apparent mismatch between the powers currently wielded by the devolved institutions and what the public regard as the appropriate balance of power between London and Edinburgh can be resolved satisfactorily only if the devolution settlement is rewritten in some way. Certainly, that seems to have been the conclusion drawn by all three main ‘unionist’ parties when they agreed in December 2007 to establish the Commission on Scottish Devolution chaired by Sir Kenneth Calman. In its final report, the Calman Commission recommended a modest extension of the powers of the Scottish Parliament, including control of the national speed limit and the regulation of airguns. More importantly, it also recommended that the parliament be given responsibility for raising a substantial part of its budget by setting a ‘Scottish rate’ of income tax, and through being given control of some additional minor tax powers (Commission on Scottish Devolution, 2009).1 The introduction of any such change will depend on decisions made by  

186

Conclusion a future UK government rather than by politicians in Edinburgh. Nevertheless, the commission would probably not have been created in the first place if the SNP had not won in 2007, while its proposals, which have been welcome by all three ‘unionist’ parties, appear to have the potential to prove a catalyst for some modest rewriting of the original devolution settlement. So in its consequences too, the parliamentary election could well prove to be evolutionary rather than revolutionary. WHITHER POLITICAL PARTIES? THE IMPACT OF STV While it was the future of the Union that appeared to be in doubt as a result of the electoral success of the SNP in the parliamentary election, it was the role of Scotland’s political parties in local government that appeared to be thrown into question as a result of the introduction of the single transferable vote for local elections – with potentially revolutionary consequences. Although usually regarded as a roughly proportional system, STV invites voters to think about the merits of individual candidates and not just the parties for whom those candidates are standing. If voters were to take up that invitation, the incentive for individual councillors and candidates to promote their personal popularity would be considerable, perhaps even to the point that, when faced with a choice between following the policy of their party and the demands of their constituents, they might be reluctant to follow party discipline. Parties could come to count for little in the country’s council chambers. In many respects voters did take up the opportunity presented by STV to express a more nuanced and thus potentially more candidatecentred vote. Around four in five voters took the opportunity to mark more than one preference, while nearly three in five included candidates from more than one party amongst their preferences. Those preferences suggested (as anticipated by Curtice and Herbert, 2005) a particular dislike of the Conservatives amongst supporters of all other parties. It is notable too that voters were more likely to vote for the candidate they liked best (rather than simply the party they most preferred) in the local elections than they were in the constituency contests for Holyrood, a phenomenon that seems to have been particularly important in enabling local Conservative candidates to overcome the widespread dislike of their party.  

187

Revolution or Evolution? Yet, alongside these signs of candidate-centred voting, party still mattered most to the way in which people voted, and certainly candidates mattered much less than might have been expected on the basis of the experience of the Republic of Ireland. In part, this was a consequence of the relatively conservative nomination strategies pursued by the political parties; rarely were they willing to take the risk of losing a seat by nominating too many candidates. Less than half of voters found themselves presented with a choice of candidates by the party to whom they gave their first-preference vote. As a result many voters could only take into account the merits of the individual candidates in deciding how to cast their first-preference vote if they were potentially willing to give it to a candidate who was not standing for the party they most preferred. Even so, some voters apparently wanted to use the STV ballot paper to do little more than express support for a particular party. When they were presented with more than one candidate from the party of their firstpreference vote, a majority of voters loyally ranked that party’s remaining candidates above those of any other party. Moreover, such voters appeared willing to follow the advice of that party in determining the order in which to place its candidates – while in the absence of that advice many imitated the order on the ballot paper, such that candidates placed lower down the ballot paper were consistently less likely to be elected (Denver and Bochel, 2007). It seems there were sections of the Scottish electorate for whom the personal merits of individual candidates mattered little at all. As a result, the introduction of STV must also be regarded as an evolutionary step rather than a revolutionary one. Introducing the same electoral system as in Ireland did not suddenly create the mirror image of that country’s candidate-centred politics. The perceived merits of individual candidates did matter to some extent to the way in which voters used their ballot paper, thereby generating an incentive for councillors and candidates to promote their personal popularity along with that of their party. The most important determinant of the success or otherwise of individual candidates, however, was still the popularity of the parties they represented. Scottish local government may have become a little more candidate-centred, but it is still very much wrapped in party labels. It is possible, of course, that the full consequences of this change of  

188

Conclusion electoral system may not have been evident on the occasion of its first use. Voters may still have been learning how to exploit the new system. Moreover, the decision of the SNP government to end in future the practice of holding local elections on the same day as elections for the Scottish Parliament will mean that future local contests will not be held under the shadow of the Holyrood contest. Perhaps when they only have one election on which to focus, significantly more voters will pay attention to the merits of individual candidates in future, though that is to presume that a stand-alone local election contest succeeds in generating sufficient interest in the first place. In the meantime, however, it would seem that Scotland’s initial experience of STV gives reason to be cautious about assertions, on both sides of the debate, that the introduction of STV in elections to other bodies, such as the Scottish Parliament or the UK House of Commons would radically change the relationship between elected representatives on the one hand, and both their electorates and their parties on the other. CONCLUSION In summary, neither the electoral success of the SNP in the 2007 Scottish parliamentary election, nor the introduction of the single transferable vote in that year’s local elections heralded a revolution in Scottish politics. Both developments are better regarded as evolutionary in nature, but that is not to argue that either development was of little significance. Scottish politics passed some important milestones in 2007. For the first time, power changed hands at Holyrood. It can no longer be thought that in establishing devolution Labour had created for itself an institutional arrangement that it would be able perpetually to dominate. It has now been demonstrated that power in the devolved institutions can alternate between different parties, albeit between varieties of coalition and minority government rather than between single-party majority governments as is the norm at Westminster. Ironically, given the independence ambitions of the SNP, the relative ease with which power has transferred from one party to another may actually have strengthened devolution and confirmed its ability to ensure that distinctive Scottish preferences can be accommodated within the Union. The outcome of the  

189

Revolution or Evolution? 2007 parliamentary elections also means that different parties lead government in London and Edinburgh. Politicians in the two capitals have had to learn to accommodate their partisan differences. At the same time, the local elections provided the first empirical evidence of how Scotland reacted to the use of a form of proportional representation that is more candidate-centred than any system hitherto used in any part of the UK outside of Northern Ireland. That its introduction did not undermine the importance of party in how people voted to the extent that might have been anticipated suggests that political parties have rather less reason to fear its use in other elections than previously thought. In the meantime, so far as Scottish local government is concerned, the change certainly did help to ensure that, as at Holyrood, power is now fragmented. After the 2007 elections devolved Scotland must now be regarded as quite clearly a multi-party state – and in that respect at least it has indeed become a rather different country than England. NOTE 1. The Commission proposed that the basic and higher rates of income tax collected by the UK government in Scotland should be reduced by 10p in the pound north of the border, and replaced by a flat rate tax set by the Scottish Parliament. The parliament would be required to decide whether this tax should be 10p in the pound, or be set at a higher or lower rate. In addition, it recommended that air passenger duty, landfill tax, aggregates tax and stamp duty land tax be wholly devolved to the Scottish Parliament. The block grant that currently funds the devolved institutions would then be reduced accordingly, with the calculations based on the assumption that the Scottish Parliament had set the same tax rates as in the rest of the UK. REFERENCES Commission on Scottish Devolution (2009), Serving Scotland Better: Scotland and the United Kingdom in the 21st century, Edinburgh: Commission on Scottish Devolution.  

190

Conclusion Curtice, J. and Herbert, S. (2005), STV in Local Government Elections: Modelling the 2003 results, SPICE Briefing 05/31, Edinburgh: Scottish Parliament Information Centre. Denver, D. and Bochel, H. (2007), ‘A Quiet Revolution: STV and the Scottish council elections of 2007’, Scottish Affairs, 61: 1–17. Jeffery, C. and Hough, D. (2003), ‘Regional Elections in Multi-Level Systems’, European Urban and Regional Studies, 10: 199–212. Johns, R., Mitchell, J., Denver, D. and Pattie, C. (2009), ‘Valence Politics in Scotland: Towards an explanation of the 2007 Election’, Political Studies, 57: 207–33. Scottish National Party (2007), Manifesto 2007, Edinburgh: Scottish National Party.

 

191

TECHNICAL APPENDIX __________________________________________

DETAILS OF THE SURVEYS

The data presented in this book on the attitudes and behaviour of the resident population in Scotland at the time of the 2007 Scottish Parliament election are drawn from the 2007 Scottish Social Attitudes Survey. The survey was conducted by the Scottish Centre for Social Research (part of the National Centre for Social Research). This was the ninth year of the study, which started in 1999 with the Scottish Parliamentary Election Survey/1999 Scottish Social Attitudes survey (Paterson et al., 2001). The series is parallel to the long-established British Social Attitudes survey (Park et al., 2009). THE 2007 SCOTTISH SOCIAL ATTITUDES SURVEY The questionnaire for the 2007 survey consisted of a number of separately funded modules of questions. Most of the data used in this book come from a module on the parliamentary and local elections funded jointly by the Leverhulme Trust and the Economic and Social Research Council (ESRC). The Leverhulme Trust provided support for questions on the parliamentary election and on attitudes towards devolution. The ESRC (grant number RES-062-23-0513 and contract number T0063.GH.P&S), provided support for questions on the local elections and for a number of additional questions on constitutional preference, the powers of the parliament and the instrumental impact of devolution. We also make use of the data from two other modules included on the 2007 survey. These are a module on attitudes to government and public services in Scotland funded by the Scottish Government Office of the Chief Researcher (the results of which are reported in detail in Ormston, 2008) and a module of questions on attitudes towards the funding, provision and delivery of public services that was funded by the ESRC as part of its Public Services Research Programme (grant number RES-16625-0043). The survey also contained modules on subjective perceptions of wellbeing in Scotland (funded by the Scottish Government Office of the 193

Revolution or Evolution? Chief Researcher – see Given and Webster, 2008, for findings) and attitudes to drinking alcohol and the role of alcohol in Scottish culture (also funded by the Scottish Government – see Ormston and Webster, 2008, for findings). Copies of the full questionnaires are available from the Scottish Centre for Social Research on request (e-mail: [email protected]). Sample Design The Scottish Social Attitudes survey is designed to yield a representative sample of adults aged 18 or over in Scotland. People were eligible for inclusion in the 2007 survey if they were aged 18 when the interviewer first made contact with them. The sampling frame for the survey was the Postcode Address File (PAF), a list of addresses (or postal delivery points) compiled by Royal Mail. For practical reasons, the sample was confined to those living in private households. People living in institutions (such as nursing homes or hospitals) were excluded, as were any households whose addresses were not on the Postcode Address File. The sampling method involved a multistage design, with three separate stages of selection: selection of sectors, addresses and individuals. A total of 134 postcode sectors were selected at random from a list of all the sectors in Scotland. Sectors in urban areas were selected with probability proportional to the number of addresses in each sector, while in rural areas they were selected with a probability of twice the address count. (Sectors in rural areas were those that belonged to one of the three more rural categories in the Scottish Government’s six-fold urban–rural classification of settlements; see Scottish Executive, 2006.) Thereafter, half of each selected postcode sector was picked at random (to reduce travel time for interviewers working in very large postcode sector areas). Prior to selection all the postcode sectors in Scotland were stratified by region, percentage of household heads recorded as being in non-manual occupations (defined as those in socio-economic groups 1–6 and 13 and taken from the 2001 Census) and the Scottish Government classification of urban and rural settlements. Addresses were selected at random within each selected postcode half sector. The number of addresses selected in each case varied, based on 194

Technical Appendix previous experience of the level of non-response in that type of sector. This practice was designed to try and ensure that (i) the achieved sample matched as closely as possible the geographic spread of the population (after taking account of the over-sampling of rural areas), and (ii) that interviewers working in different types of areas were able to achieve similar numbers of interviews from their batches of selected addresses. In total the selected sample comprised 3,055 addresses. Interviewers called at each selected address and identified whether it was the address of one or more private residential dwelling units. Where more than one dwelling unit was present at an address, all the dwelling units were listed systematically, and one was selected at random using a computer generated random selection table. In all selected dwelling units in which more than one adult aged 18 or over was resident, interviewers selected one person to interview at random, using a similar procedure. Fieldwork and Data Collection Interviews were conducted between late May and early November 2007 (with 84% completed by the end of August). An advance letter was sent to all addresses and was followed up by a personal visit from a Scottish Centre for Social Research interviewer. Interviewers were required to make a minimum of four calls at different times of the day (including at least one evening and one weekend call) in order to try and contact respondents, although in practice interviewers often made many more calls than this. All interviewers attended a one day briefing on the content and administration of the survey prior to starting work on the study. The interviewing was undertaken using face-to-face computer-assisted interviewing. The questions that interviewers were to ask appeared on the screen of a laptop computer, while respondents’ answers were entered into that same computer. All respondents were also asked to fill in a selfcompletion paper questionnaire that, whenever possible, was collected by the interviewer, but in some cases was posted back by the respondent. Up to three postal reminders were sent to maximise the return of selfcompletion questionnaires. 195

Revolution or Evolution? Response The face-to-face interview, the median length of which was 57 minutes, was completed by 1,508 respondents. This represented an overall response rate of 55% (for further details see Table A.1). The selfcompletion questionnaire was completed by 1,315 respondents (that is, 87% of those who undertook the face to face interview). Table A.1 2007 Scottish Social Attitudes Survey Response

Addresses issued Vacant, derelict and other out of scope 1 Unknown eligibility 2 In scope Interview achieved Self-completion returned Interview not achieved Refused 3 Non-contacted 4 Other non-response 5

Lower Estimate No. % 3055 326 10.7

Upper Estimate No. % 3055 326 10.7

121 2729 1508 1315 1221 824 144 132

121 2608 1508 1315 1100 824 144 132

4.4 55.3 48.2 44.7 30.2 4.4 4.8

4.0 57.8 50.4 42.2 31.6 5.5 5.1

The table shows a ‘lower’ and an ‘upper’ response rate. The former is calculated on the assumption that all addresses whose eligibility to participate was unknown were in fact eligible to take part. The latter is calculated on the assumption that they were all ineligible (because they were empty/derelict, non-residential, etc). The ‘true’ response is likely to lie somewhere between the two, since some addresses whose eligibility was unknown are likely to have been ‘deadwood’, while others may have been eligible. See Lynn et al. (2001) for a discussion of treatment of unknown eligibility in calculating response rates. 1 This includes empty/derelict addresses, holiday homes, businesses and institutions. 2 ‘Unknown eligibility’ includes cases where the address could not be located, where it could not be determined if an address was a residence and where it could not be determined if an address was occupied or not. For the lower response rate, this is shown as a percentage of ‘in scope’ addresses. For the upper response rate, it is shown as a percentage of issued addresses, since these addresses are excluded from ‘in scope’ for the purposes of calculating the upper response rate.

196

Technical Appendix 3 Refusals include refusals prior to selection of an individual, refusals to the office, refusal by the selected person, ‘proxy’ refusals made by someone on behalf of the respondent and broken appointments after which a respondent could not be recontacted. 4 Non-contacts comprise households where no one was contacted after at least four calls and those where the selected person could not be contacted. 5 ‘Other non-response’ includes people who were ill at home or in hospital during the survey period, people who were physically or mentally unable to participate and people who had insufficient English to participate.

Weighting Weights were applied to the data to correct for three potential sources of bias in the sample: 1. Differential selection probabilities. Adults living in large households had a lower probability of being selected for interview than adults who lived alone. 2. Deliberate over-sampling of those living in rural areas. 3. Non-response. Differences between the character of responding and non-responding households were modelled using information from the census about the area in which each selected address was located, as well as from interviewer observations of both participating and non-participating addresses. These weights were then further adjusted to ensure that the weighted data matched the age and sex profile of the Scottish population (based on 2006 mid-year estimates; see General Register Office for Scotland, 2007). The final weights were scaled so that the total number of weighted cases exactly equalled the number of unweighted cases. As indicated above, a total of 193 respondents (13% of those interviewed) did not return their self-completion questionnaire. Analysis by the National Centre for Social Research’s Survey Methods Unit indicated that it was not necessary to apply additional weights to correct for this non-response. All the percentages presented in this book are based on the weighted data; the sample sizes shown in the tables represent the number of unweighted cases on which a statistic is based. 197

Revolution or Evolution? OTHER SURVEYS USED IN THE BOOK Previous Scottish Social Attitudes Surveys The book also makes extensive use of previous Scottish Social Attitudes surveys, conducted annually between 1999 (when it was also an election survey) and 2006. These earlier surveys were funded by a variety of sources, including the Scottish Executive, the Economic and Social Research Council, the Leverhulme Trust and the Nuffield Foundation. Further details of these surveys can be found in Paterson et al. (2001), Curtice et al. (2001), Bromley et al. (2003), Bromley and Curtice (2003), Bromley et al. (2006), Bromley and Given (2005), Given and Ormston (2006) and Ormston and Sharp (2007). The achieved sample sizes of these surveys were as follows: 1999 2000 2001 2002 2003 2004 2005 2006

1,482 (response rate, 59%), of whom 1,165 also returned the self-completion questionnaire 1,663 (response rate, 65%), of whom 1,506 also returned the self-completion questionnaire 1,605 (response rate, 60%), of whom 1,383 also returned the self-completion questionnaire 1,665 (response rate, 62%), of whom 1,507 also returned the self-completion questionnaire 1,508 (response rate, 57%), of whom 1,324 also returned the self-completion questionnaire) 1,637 (response rate, 61%), of whom 1,514 also returned the self-completion questionnaire 1,549 (response rate, 57%), of whom 1,409 also returned the self-completion questionnaire 1,594 (response rate, 56%), of whom 1,437 also returned the self-completion questionnaire.

The sample design of these surveys was very similar to that of the 2007 survey, with the following exceptions: 198

Technical Appendix o rural and remote areas were not over-sampled prior to the 2002 survey, o before 2007, a smaller number of postcode sectors was selected, while these were not split further into half sectors, and o before 2007, the number of addresses selected within each sector did not vary in line with the anticipated level of nonresponse. Meanwhile, weights designed to correct for the possible impact of differential non-response have only been applied since 2005. Scottish Referendum Survey The Scottish Referendum Survey was undertaken in September and October 1997. It was funded by the ESRC. The interviewing was conducted by the National Centre for Social Research and began immediately after the referendum on 11 September 1997. The sample was designed to be representative of the adult population living in private households in Scotland and eligible to vote in the referendum. It was drawn from the Postcode Address File. The random sampling design involved both stratification and clustering. The data are weighted to correct for differences in selection probabilities. A total of 676 interviews were conducted face to face using a paper questionnaire. This represented a response rate of 68%. In addition a selfcompletion questionnaire was completed by 657 respondents (97% of those interviewed). Further details can be found in Taylor and Thomson (1999). Scottish Election Study Scottish Election Studies were conducted as part of a series of British Election Study series in October 1974, 1979, 1992 and 1997. In each case interviewing took place immediately after the relevant UK general election. The samples for each study were chosen using a random selection procedure that was modified by stratification and clustering. 199

Revolution or Evolution? Until and including 1992, the sampling frame was the electoral register; in 1997 it was the Postcode Address File. The achieved sample sizes were: 1974 1979 1992 1997

1,170 729 957 882

(response rate, 76%) (response rate, 61%) (response rate, 74%) (response rate, 62%)

Further details can be found in Miller (1981), Bennie et al. (1997) and Brown et al. (1998). British Social Attitudes Survey Data on attitudes in England are extracted from the British Social Attitudes survey, on which the Scottish Social Attitudes survey is modelled. The British survey has been running annually since 1983. It aims to yield a representative sample of adults aged 18 and over living in Britain. The sample is selected using methods similar to those used by the Scottish Social Attitudes survey. Generally, the sample size (of respondents living across Britain as a whole) has been between 3,000 and 3,500. Further information is to be found in a series of the annual reports based on the survey, the most recent of which is Park et al. (2009). Irish National Election Study The 2002 Irish National Election Study, the first of its kind, interviewed a representative sample of 2,663 adults aged 18 and over living in the Republic of Ireland. A three-stage clustered random sampling design was deployed. In the first instance, a random sample of primary sampling units was selected. At the second stage, a random sample of households was selected within each sampling unit. Finally, a random person was selected within each household. The response rate was 60%. For further details see Marsh et al. (2008: Appendix I).

200

Technical Appendix CLASSIFICATIONS USED IN ANALYSIS Most of the variables that are used in the analysis of our data are either self-explanatory or are described in detail in the main text. Here we provide additional information in respect of two more complex variables. National Statistics Socio-Economic Classification (NS-SEC) The National Statistics Socio-Economic Classification (NS-SEC) classifies occupations on the basis of the employment relations that typically accompany that occupation together with its associated employment status (that is whether someone is engaged in an occupation as self-employed, a manager, etc.). The classification has seven categories: Employers in large organisations, higher managerial and professional Lower professional and managerial; higher technical and supervisory Intermediate occupations Small employers and own account workers Lower supervisory and technical occupations Semi-routine occupations Routine occupations In this book those in the first two categories have been combined into a single group, as have those in the last two. Those respondents who were not currently in employment were classified on the basis of their last job, thereby leaving only a small group who have never had a job that could not be classified. For further details see Office for National Statistics (2005). National Identity National identity is obtained in two ways on the Scottish Social Attitudes survey. First of all respondents are asked: 201

Revolution or Evolution? Please say which, if any, of the words on this card describes the way you think of yourself. Please choose as many or as few as apply. The card in question contains the following identities: British, English, European, Irish, Northern Irish, Scottish, Ulster, Welsh, None of these. We refer to the answers to this question as ‘free choice’ national identity. Meanwhile, those respondents who choose more than one identity are asked further: If you had to choose, which one best describes the way you think of yourself? The resulting single choices (together with the choices of those who initially only chose one identity) are referred to as ‘forced’ choice national identity. Secondly, all respondents are asked a question, known as the Moreno scale (Moreno, 1988) that invites them to choose what balance of Scottish and British identity best describes themselves. The wording is as follows: Which, if any, of the following best describes how you see yourself? Scottish not British More Scottish than British Equally Scottish and British More British than Scottish British not Scottish Other (None of these)

202

Technical Appendix DATA INTERPRETATION Statistical Significance Nearly all the data in this book come from samples of the population, and as a consequence are subject to sampling error. However, it is possible to calculate confidence intervals for any figure obtained from a sample. These intervals describe a range within which it is likely (for any given level of probability) that the true population figure lies. Table A.2 gives an indication of the confidence intervals that apply to various percentages obtained from different sample sizes. The confidence intervals shown are the 95% intervals; they indicate that we can be 95% sure that the true answer does not lie above or below the relevant sample percentage figure by more than the interval shown. For example, there is a 95% chance that if the result obtained from a sample of 500 is 50%, the true results lies within +/- 4 percentage points of that figure (thus, between 46% and 54%). Table A.2 Confidence Intervals for Survey Findings

Sample size 50 100 250 500 1,000 2,000

Approximate 95% confidence limits for a percentage of: 10% or 30% or 90% 70% 50% +/+/+/8 13 14 6 9 10 4 6 6 4 4 3 3 3 2 1 2 2

The confidence intervals in table A.2 assume simple random sampling with no adjustment for the possible effect of clustering the sample into a number of sample points, as is the practice on the Scottish Social Attitudes survey. Such an adjustment would usually increase the confidence intervals slightly, but in most cases they would not differ notably from those shown on the table (see, for example, Paterson, 2000: Appendix). 203

Revolution or Evolution? Certain types of variables, that is those most associated with the area a person lives in, are affected to a greater degree by clustering than are others. For example, Labour identifiers and local authority tenants tend to be concentrated in certain areas and so, when the effect of the sample’s clustered design are taken into account, the confidence intervals around such variables are particularly likely to increase. Typically, however, this is not a feature of most attitudinal variables (see Park et al., 2009 for examples of these kinds of design effects). Tests of statistical significance take account of the confidence intervals attached to survey findings, and can be used to assess the likelihood that any difference between the results of two samples (including between two or more sub-groups within a survey) could have arisen by chance as a result of sampling error. Whenever comments on such differences are made in this book, these differences have usually been tested and found to be statistically significant at the 95% level or above. Statistical Modelling Chapters 5 and 6 contain the results of logistic regression analyses. Regression analysis aims to summarise the relationship between a ‘dependent’ variable and one or more ‘independent’ explanatory variables. It shows how well we can estimate a respondent’s score on the dependent variable from knowledge of their scores on the independent variables. The technique takes into account the relationships between the different independent variables (for example, between education and income, or social class and housing tenure) and thereby identifies the degree to which the association between any one independent variable and the dependent variable is separate from (or independent of) any association between the dependent variable and the other independent variables included in the analysis. Regression is often undertaken to support a claim that the phenomena measured by the independent variables cause the phenomenon measured by the dependent variable. However, the causal ordering, if any, between the variables cannot be verified or falsified by the technique. Causality is inferred through assumptions made by the analyst. Logistic regression is a particular form of regression that is commonly used when the dependent variable is binary, that is it either has the value 204

Technical Appendix zero or one. It assumes that the relationship between the dependent and the independent variables takes the form of an S-curve, whereby the impact on the dependent variable of a one-point increase in an independent variable becomes progressively less, the closer the value of the dependent variable approaches zero or one. This avoids the possibility that the value of the dependent variable estimated for any individual case is either less than zero or greater than one. In modelling the data, regression analysis also identifies whether the relationship or differences in question are statistically significant. Full technical details of regression can be found in any textbook on social statistics, such as Bryman and Cramer (1997). REFERENCES Bennie, L., Brand, J. and Mitchell, J. (1997), How Scotland Votes, Manchester: Manchester University Press. Bromley, C., Curtice, J., McCrone, D. and Park, A. (2006), Has Devolution Delivered?, Edinburgh: Edinburgh University Press. Bromley, C. and Given, L. (2005), Public Perceptions of Scotland after Devolution: Findings from the 2004 Scottish Social Attitudes Survey, Edinburgh: Scottish Executive Social Research. Available at http://www.scotland.gov.uk/Publications/2005/08/18151621/ 16219. Bromley, C. and Curtice, J. (2003), Attitudes to Discrimination in Scotland, Edinburgh: Scottish Executive. Bromley, C., Curtice, J., Hinds, K. and Park, A. (2003), Devolution – Scottish Answers to Scottish Questions?, Edinburgh: Edinburgh University Press. Brown, A., McCrone, D., Paterson, L. and Surridge, P. (1998), The Scottish Electorate: The 1997 election and beyond, London: Macmillan. Bryman, A. and Cramer, D. (1997), Quantitative Data Analysis, London: Routledge. Curtice, J., McCrone, D., Park, A. and Paterson, L. (eds) (2001), New Scotland, New Society?, Edinburgh: Polygon. General Register Office for Scotland (2007), Mid-2006 Population Estimates Scotland, Edinburgh: General Register Office for Scotland. 205

Revolution or Evolution? Given, L. and Ormston, R. (2007), Scottish Social Attitudes Survey 2005: Scottish Executive core module technical report, Edinburgh: Scottish Executive Social Research. Available at http://www.scotland.gov.uk/ Publications/2006/12/05122049/0. Given, L. and Webster, C. (2008), Scottish Social Attitudes Survey 2007 – Core Module Report 2: Subjective perceptions of well-being in Scotland, Scottish Government Social Research. Lynn, P., Beerten, R., Laiho, J. and Martin, J. (2001), Recommended Standard Final Outcome Categories and Standard Definitions of Response Rates for Social Surveys, Colchester: Institute for Social and Economic Research. Marsh, M., Sinnott, R., Garry, J. and Kennedy, F. (2008), The Irish Voter: The nature of electoral competition in the Republic of Ireland, Manchester: Manchester University Press. Miller, W. (1981), The End of British Politics? Scots and English political behaviour in the 1970s, Oxford: Clarendon Press. Moreno, L. (1988), ‘Scotland and Catalonia: The Path to Home Rule’, in McCrone, D. and Brown, A. (eds), The Scottish Government Yearbook 1988, Edinburgh: Unit for the Study of Government in Scotland. Office for National Statistics (2005), The National Statistics Socio-economic Classification: User manual, Basingstoke: Palgrave Macmillan. Ormston, R. (2008), Scottish Social Attitudes Survey 2007 – Core Module Report 1: Attitudes to government in Scotland, Scottish Government Social Research. Available at http://www.scotland.gov.uk/Publications/2008/ 2008/05/16095134/13. Ormston, R. and Sharp, C. (2007), Scottish Social Attitudes Survey 2006 – Core Module Report 2: Perceptions of government in Scotland, Edinburgh: Scottish Government Social Research. Available at http://www.scotland.gov.uk/ Publications/2007/11/19102513/0. Ormston, R. and Webster, C. (2008), Scottish Social Attitudes survey 2007 – Something to be Ashamed of or Part of our Way of Life? Attitudes towards alcohol in Scotland, Scottish Government Social Research. Available at http://www.scotland.gov.uk/Publications/2008/08/01112431/0. Park, A., Curtice, J., Thomson, K., Phillips, M. and Clery, E. (eds) (2009), British Social Attitudes: The 25th report, London: Sage. Paterson, L. (2000), ‘The Social Class of Catholics in Scotland’ Journal of the Royal Statistical Society (Series A), 163: 363–79. 206

Technical Appendix Paterson, L., Brown, A., Curtice, J., Hinds, K., McCrone, D., Park, A., Sproston, K. and Surridge, P. (2001), New Scotland, New Politics?, Edinburgh: Edinburgh University Press. Scottish Executive (2006), Scottish Executive Urban Rural Classification 2005–6, Edinburgh: Scottish Executive, 2006. Taylor, B. and Thomson, K. (1999), Scotland and Wales: Nations again?, Cardiff: University of Wales Press.

207

  __________________________________________

INDEX

Conservative party, 5, 6, 17, 91–2, 97, 99, 100, 105,106, 112, 136, 141, 144, 152, 175–9, 187–8 constituency vote, 3–5, 80–1, 89, 99, 105, 132–51 council tax, 91–3 curfew, 92–3

accountability, 75–85, 97–8 and political leaders, 86–90 Additional Member System (AMS), 3, 8, 13, 127,162,179 voting behaviour under, 131–49 age group, 104, 106, 197 Alexander, Wendy, 62–3 alphabetic voting, 171–4 Arbuthnott commission, 135, 149 Australia, 10, 127

Dáil, 164 defence, 63 Democrat party, 128 devolution, 1–2, 4, 7, 11–12, 17–19, 21, 23–7, 29–37, 42–3, 48, 53, 56–7, 62, 64, 68–9, 72–3, 78, 120–1, 129, 130, 182–4, 186–7, 189, 193 attitudes towards, 43–50, 52–3, 56–62, 65–9, 100, 104, 122 and attitudes towards SNP leaders, 114 and attitudes towards Tony Blair, 120 and voting SNP, 110–11, 113, 115–17 devolved elections, 2–7, 73–5, 90, 97, 112–13, 184–5; see also Scottish Parliament, elections to Dewar, Donald, 42

ballot papers in Scottish Parliament elections, 132 in single transferable vote election, 10,11, 160–1 voter understanding of, 134–8, 149–50, 168–71 voters’ use of, 162–8, 171–4, 179–80, 188 Barnett formula, 64 Blair, Tony, 6, 86–90, 98, 114, 117–20 British Election Study, 151, 199 British identity see national identity British Social Attitudes survey, 193, 200 Brown Gordon, 6 business taxes, 91–3, 96–7, 100

Economic and Social Research Council (ESRC), 193, 198, 199 economy and the Union, 42, 51, 59, 70 evaluations of and voting, 76–9, 82–3 education, 30, 33–4, 42, 44–6, 76–9 and devolution, 44–6, 63 educational qualifications, 136, 38–9, 149–50, 168–73 evaluations of, 76–9

Calman, Sir Kenneth, 186 candidate-centred voting, 158–60, 166–8, 179 Catalonia, 184 Catholic, Roman, 104, 106–7 Commons, House of, 3–5, 8, 12, 189; see also Westminster coalition government, 2–8, 13, 72–6, 91, 142–5, 151, 177, 183, 189 congestion charge, 92–3

 

209

Revolution or Evolution? Ireland, Republic of, 10, 14, 127, 158, 162–5, 166, 172, 174, 179, 188, 200 Irish Election Study, 14, 164, 172, 200

Eisenhower, Dwight, 128 Electoral Reform Society, 8 England, 12, 17–18, 30, 37, 42, 51, 59, 64, 70, 72, 73, 186, 190 attitudes in, 25–9, 31–8, 200 equality, 25–8, 34–6 European election, 5

Labour party, 1–11, 13, 30, 52, 56, 63, 66, 69, 72–3, 75–7, 80–100, 104–7, 112, 114, 117–20, 122, 129, 136, 141–5, 149–51, 175–9, 183–5, 189, 204 left–right scale, 25–8 Leverhulme Trust, 193, 198 Liberal Democrats, 3–8, 13, 62, 72, 76, 91–2, 105–6, 112, 136, 141–5, 149, 151, 175–9, 183 libertarian–authoritarian scale, 28–9 list vote, 3–6, 99, 132–45, 149–50 Local Governance (Scotland) Act, 8 local government, 1–2, 9, 91, 180 elections to, 1, 7–11, 126, 131–3, 139–40, 148, 150–78, 182, 187–90 Local Government Boundary Commission for Scotland, 8, 180 local income tax, 91–3, 96 Luckhurst, Tim, 28

First Minister, 5–6, 86–90, 114 first-preference vote, 99, 126, 163–7, 171–9, 188 fiscal autonomy, 65 foreign affairs, 63 gender, 104, 106–7, 197 Germany, 133–4 Goldie, Annabel, 99 government spending, 34–6 Scotland’s share of, 51–2, 64–5 Greater London Assembly, 8 Green party, 5, 6, 92, 150 health service, 30–1, 33, 36, 44–6, 63, 99 evaluations of and voting, 76–9, 82–6 Herald, The, 180 homosexuality, 28

McConnell, Jack, 6, 86–90, 98, 114, 118, 185 McLeish, Henry, 42 Malta, 10, 127 minority government, 6, 189 Moreno scale, 21–2, 27–8, 108, 116–17, 202 Muslim, 24

ICM, 4 independence, 1, 5–7, 46, 62, 120, 182, 185, 189 and attitudes towards SNP leaders, 114–15 and attitudes towards Tony Blair, 120 support for, 11–13, 53, 56–60, 63, 65–6, 68–70, 72, 100, 104, 183, 185–5 and voting SNP, 104–5, 109–10, 113, 115–18, 120–2 Independent candidates, 164, 167, 180 invalid vote, 135–6, 150 Ipsos–MORI, 5 Iraq war, 92–8, 100, 117–20 Ireland, Northern, 8, 190

 

National Centre for Social Research, 193, 197, 199 national identity, 2, 12, 14, 37, 72, 128, 184–6, 201–2 and attitudes to devolution, 67–8 British v. Scottish, 17–24, 33–6, 56 and other attitudes, 24–5, 33, 37 and voting SNP, 104–5, 108–9, 111, 113–18, 184–5

210

Index 43, 46–53, 56–63, 66–8, 72–3, 88–92, 96–8, 100, 104–22, 129, 136, 144–5, 150–2, 175– 9, 182–7, 189 Scottish Office, 7, 42 Scottish Parliament, 2, 12, 17–19, 36, 42–7, 56, 59–68, 90–1, 97–8, 109–10, 115, 122, 183, 186, 189–90 elections to, 3–7, 8, 13, 73–4, 80, 83, 85, 92, 108–13, 116, 121, 127–8, 131–48, 162, 182, 184, 187, 190, 193 see also devolved elections Scottish Referendum Survey, 199 Scottish Social Attitudes (SSA) survey, 11, 57–8, 130, 135, 142, 144, 193–8 Scottish Socialist Party, 5, 104 second-order elections, 74–5 single-member plurality, 1, 3–4, 8–9, 127, 131, 141, 158–9, 162 single transferable vote (STV), 1, 7–11, 126, 131, 148, 149, 182 voting behaviour under, 126–8, 140, 145, 158–80, 187–9 social class, 106–7, 201, 204 Solidarity, 104 standard of living, 76–8, 83 Steel, Lord, 62–4 Stephen, Nicol, 99 Sturgeon, Nicola, 30 Sutherland Commission, 30 Swinney, John, 5, 87–8, 114–15

National Statistics Socio-Economic Classification, 201 New Zealand, 133, 149 nuclear power, 90–4, 96–7, 100 Nuffield Foundation, 198 oil, 51, 62 opinion polls, 4, 69, 86 party identification, 127–31, 133, 139–41, 160–1, 163, 180 personal care, 30, 32–4 polling district, 132 Postcode Address File, 194, 199 prescription charges, 30, 32–4, 36, 38, 91, 93, 96 Prime Minister, 6, 86–90, 118–19 Protestants, 107 public services, provision of, 30–4 public transport, evaluations of, 76–9 referendum on devolution, 12, 21, 24, 44–5, 49, 60, 199 on independence, 6, 56, 58, 183 Republican party, 128 Robertson, George, 2, 4 Royal Mail, 194 Salmond, Alex, 5–6, 53, 86–90, 114–18, 120, 185 Scotland Office, 150 Scotland on Sunday, 4 Scotsman, The, 28 Scottish Centre for Social Research, 11, 193–4 Scottish Election Study, 129, 151, 199 Scottish Executive, 48, 60–1, 67, 75– 86, 98, 99, 198 Scottish government(s), 30, 36, 66, 74, 90–1, 97, 183 Scottish Government, 193–4 Scottish identity see national identity Scottish National Party (SNP), 1–7, 12–13, 18, 24, 30, 32, 34, 37,

 

tactical voting, 141–7 Tasmania, 10 taxation, 6, 34–5, 62–3, 64–6, 69, 73, 91–2, 96–7, 100 powers of Scottish Parliament, 57, 62–6, 186, 190 Thatcher, Margaret, 17 ticket splitting, 131–3, 137–8, 140, 143, 148–9, 168 Trident, 92–7, 100 tuition fees, 30–2, 34, 36–7, 91–6

211

Revolution or Evolution? United States, 92, 128

UK general elections, 108–13, 121, 131–3, 139, 141, 143, 184, 199; see also Westminster UK government, 2, 18, 24, 34, 36, 42, 48–9, 52, 56, 63–8, 74–86, 90–1, 98–9, 119, 135, 183–7, 190; see also Westminster unemployment benefit, 34–5 Union, 2, 12, 17–18, 36–7, 43, 49–53, 56–60, 70, 110, 182, 186–7, 189

 

Wales, 30, 38, 91 National Assembly of, 8 welfare benefit(s), 63, 65, 91; see also unemployment benefit Westminster, 5, 18, 43, 68, 74–5, 81, 83, 86, 98, 149, 189; see also Commons, House of; UK government; UK general elections

212