CREST CENTRE FOR RESEARCH INTO ELECTIONS AND SOCIAL TRENDS

Transcription

1 CREST CENTRE FOR RESEARCH INTO ELECTIONS AND SOCIAL TRENDS Working Paper Number 80 September 2000 Fond Hearts and Wise Heads? Attitudes to Proportional Representation following the 1999 Scottish and Welsh Elections By John Curtice, Ben Seyd, Alison Park and Katarina Thomson The Centre for Research into Elections and Social Trends is an ESRC Research Centre based jointly at the National Centre for Social Research (formerly SCPR) and the Department of Sociology, University of Oxford

2 by John Curtice, Ben Seyd, Alison Park and Katarina Thomson Fond Hearts and Wise Heads? Attitudes to Proportional Representation following the 1999 by John Curtice, Ben Seyd, Alison Park and Katarina Thomson. September 2000 Paper 80

3 Abstract There is considerable debate in Britain about how the public are likely to react to a referendum on electoral reform for the House of Commons. On the one hand it is argued that survey evidence shows there is a stable majority for reform and that this is likely to be bolstered when voters become better informed about the subject during the course of a referendum campaign. On the other hand it is argued that respondents' answers to survey questions about electoral reform are sensitive to question wording, and that this suggests that many voters do not have clear views on the topic and might be swayed in either direction during a referendum campaign. This article examines how voters in Scotland and Wales reacted to having the opportunity in the devolution elections in 1999 to use a proportional representation for the first time. Doing so gives us insight as to whether an electorate that becomes better acquainted with proportional representation is more likely to support it. The results suggest that this is not necessarily so and that as a result it cannot be assumed that any future referendum on electoral reform would secure a majority for change. Of the arguments commonly put up against reform, the one that has most resonance with the public is the claim it would mean the loss of their local MP. 2

4 Introduction The 6 th May 1999 was a historic day in British politics. For the first time ever, people in Scotland and Wales had the opportunity to vote for their own devolved institutions. But not only were the institutions new; so also was the electoral system. For the first time since 1950, voters living in Great Britain were asked to cast a ballot under a system of proportional representation. And there are at least three reasons why their reaction that experience matters. First, the perceived success of these elections could have important implications for the legitimacy and standing of the new devolved institutions. One of the main challenges of these institutions in their early lives is to become as accepted a part of the political landscape as, traditionally at least, the Westminster parliament has been. It was already the case in Wales at least that the narrowness of the vote in the September 1997 referendum in favour of creating a new Assembly had left a substantial question mark over whether there was even majority support for that body (Wyn Jones and Trystan, 1999). If now people in Scotland and/or Wales were to regard the way in which their new institutions were elected on their first outing as unacceptable, those institutions might find themselves under a cloud from which they would find it difficult to recover. It would also likely mean that the recent recommendation of the Kerley Committee (2000), that Scottish local elections should also be conducted using a form of proportional representation, would not find public favour. 1 Second, the elections took place at a time when the issue of electoral reform for the UK House of Commons was also on the policy agenda. In its 1997 manifesto the Labour government promised to set an independent commission the task of identifying what might be the best alternative to the existing single member plurality system, and then to give the 3

5 electorate the chance to choose between the two systems in a referendum. Although, far from perfectly proportional, the alternative system proposed by the Independent Commission on the Voting System under the chairmanship of Lord Jenkins (a system known as 'Alternative Vote Plus') does hold out the promise of being somewhat more proportional than the existing system (Jenkins Commission, 1998). Yet to date the government has proved reluctant to put even this relatively limited reform forward in a referendum. An adverse reaction amongst people in Scotland and Wales to the experience of proportional representation in action would doubtless only reinforce the government's apparent doubts about introducing change for Westminster. Third, how voters reacted to these elections provides us with an important opportunity to assess what might influence the British public's attitudes towards electoral reform should they eventually be given the opportunity to express their views about proportional representation. As we detail further below, how much support there is for electoral reform amongst the British public and what arguments might prove influential in any referendum campaign has been the subject of considerable dispute. Not least of the reasons for this is that until 1999 research has had to be conducted with in effect one hand tied behind its back. As no election had been held in Great Britain since 1950 under any system other than the plurality rule, testing voters attitudes to alternative systems involved seeking their views on a subject with which few were familiar. 2 The 1999 Scottish and Welsh elections however enable us to examine the views of an electorate that has experienced proportional representation in a relatively high profile election, as well as having had previous experience of single member plurality. We thus might look to them give us an understanding of how the British public in general might react in the wake of a referendum 1 The Committee reported in June 2000, and recommended that local councillors in Scotland be elected by the Single Transferable Vote method. 2 There is of course a considerable debate about how far knowledge affects people's decision-making in referendums and elections (Delli Carpini and Keeter, 1996; Luppia, 1994). Our assumption here is simply that greater knowledge may have an influence, particularly on a subject on which previous research has suggested the public is not well informed (see below), and that the 1999 Scottish and Welsh elections give us the opportunity to assess its possible impact. 4

6 campaign that would attempt to make them better informed about the subject. In this paper we use survey evidence collected at the time of the Scottish and Welsh elections in order to undertake that task. Questions about reform Previous research suggests that there are two key questions that we might ask about the public's reaction to the 1999 Scottish and Welsh elections. The first is what impact on attitudes did the experience of proportional representation have on the level of support for proportional representation? The second is how much hold do some of the allegedly weak links in the pro-reform case have when an electorate supposedly becomes relatively well informed? Questions about electoral reform have been posed to the British public in opinion polls on an occasional basis ever since the issue first became a topic of significant public debate in the mid-1970s following the two 1974 general elections (Weir, 1992). The first of those elections saw the most serious third party challenge yet in post-war British politics. At the same time, Labour won more seats than the Conservatives despite winning fewer votes, a position they were then able to consolidate in the second election, despite still having less than 40 per cent of the vote. But what those polls tell us about public opinion on the subject has been the subject of considerable - and sometimes heated - debate. What is agreed amongst the protagonists is that the distribution of responses to questions about electoral reform is sensitive to the wording of the question that is asked (Smith, 1991; Curtice, 1993; Dunleavy and Margetts, 1999: 30-31; Curtice and Jowell, 1998: 67-69). What is in dispute is what conclusions should be drawn from this. On the one hand it has been argued that all survey research uncovers some instability of response and that if one looks at the results produced by those polls that used a relatively neutral wording, there has been a stable majority for reform ever since the 1970s (Weir, 1992; Dunleavy et 5

7 al, 1993). On the other hand it has been argued that the sensitivity of responses to question wording indicates that, lacking much knowledge 3, understanding or interest in the subject, many voters do not have fixed views on the subject. Rather than showing that electoral reform is a popular measure that could be expected to win support in a referendum, the polling evidence indicates that the outcome would depend on the quality of the arguments marshalled by the two sides (Curtice and Jowell, 1997; Dunleavy and Margetts, 1997: ). Indeed, there was even a swing against proportional representation during the 1992 general election when the topic became a major campaign issue for the first time and when the opponents of reform appeared to put forward the stronger case (Curtice, 1993: Kellner, 1992). Recently, an important contribution to this debate has been made by Farrell and Gallagher (1999) who, instead of conducting survey research, implemented a number of focus groups on the subject. At the beginning of the discussions, the members of their focus groups proved to be relatively unknowledgeable about the subject. Unlike a survey interview, however, a focus group provided an opportunity to impart information to the participants. And Farrell and Gallagher found that this information affected participants' views. Initial adherence to first past the post (which was the only way many knew to run an election) gave way to support for electoral reform. As a result, Farrell and Gallagher (1999: 313) argued that there is good reason to believe that, once voters are made aware of alternative [electoral] systems and their consequences, a large proportion of them (perhaps a majority) would favour a change to a new system. 4 In other words, while accepting that there might not be stable majority for reform, Farrell and Gallagher appear to be suggesting that it is the pro-reform camp that can be expected to win any referendum debate. The survey evidence we present here brings new evidence to this debate. Having experienced proportional representation for themselves, the Scottish and Welsh electorates 3 In a MORI poll in 1990, 43 per cent of respondents said they had never heard of, or didn t know about, proportional representation. 6

8 can be expected to have become relatively knowledgeable about the subject. They will certainly have been subjected to arguments about its merits and demerits. So, we look to see whether experience was accompanied by an increase in support for reform. Moreover, we also examine whether, better appraised of the subject, voters also became more consistent in their answers. In other words, did experience of new voting arrangements make the heart fonder and the head wiser? The claim that the pro-reform camp cannot be sure of winning a referendum on proportional representation makes of course one important assumption - that its opponents have some good tunes. Indeed, we can identify at least four arguments that are commonly deployed by the opponents of electoral reform (Jenkins, 1998; Vowles, 1999). The first is that proportional representation systems are simply too complex for voters. Under first past the post it is very easy to grasp how to vote and how it is decided who wins. Under proportional representation in contrast it is all far more difficult. And if large sections of the public fail to understand how their electoral system works this could undermine the effectiveness of elections as a means of registering the popular will. However, although many voters may confess to knowing little, if anything, about proportional representation, it has been argued that ignorance does not constitute an insurmountable hurdle to changing the system. Farrell and Gallagher's qualitative research suggests that, while British voters may be concerned about the complexity of some proportional representation systems, they also appear capable of adapting to new electoral arrangements (Farrell and Gallagher, 1999: ). They conclude that a distinction needs to be drawn between the level of knowledge needed to use any new system (i.e. how to complete the ballot paper) and that required to understand how it actually works (e.g. how votes translate into seats). 4 Indeed, previous survey research has found that those who claim to know a fair amount about alternative systems are more likely to be in favour of reform (Weir, 1992: , 219). 7

9 A second claim made against proportional representation is reflected in Disraeli's famous aphorism, England does not love coalitions - and of course proportional representation appears more likely to result in (unstable) coalitions. (For a flavour of the elite debate on this subject see Sartori, 1997: 59-61, ; Budge 1998; Lijphart, 1999: ) However, survey evidence suggests that matters may not be so clear cut. In his review of past research, Weir (1992: 215) suggests that while some surveys have identified majorities against multi-party rule, others have found a majority to be in favour. A question on the desirability of coalition government that has often been asked on the British Social Attitudes and British Election Study surveys since 1983 has often found the public to be evenly divided on the subject (Curtice and Jowell, 1997: 103-4). In the event, proportional representation may or may not lead to coalition government. But it certainly requires that some legislators at least be elected in multi-member constituencies. Proponents of the existing system argue however that only single member constituencies ensure that all MPs are clearly accountable to a body of voters. Moreover such constituencies encourage MPs to intervene on behalf of their constituents when either individually or collectively they have difficulties with some action taken by the state (Curtice and Shively, 2000). And indeed, surveys have consistently found a strong attachment to single member constituencies, even among voters who simultaneously express support for the principle of proportional representation (Crewe, 1985; Young, 1985: 12; Smith, 1991: 435; Dunleavy et al, 1997: 30). However, Farrell and Gallagher (1999: 312) suggest that this attachment might be the result of voters misapprehension about the role of MPs in multi-member constituencies, or their ignorance of the potential benefits arising from such a feature. Not least of the reasons why advocates of the existing system believe it is better able to hold individual MPs accountable is that many alternative systems use closed lists. A closed list system is one where voters can only vote for a party without expressing any preference amongst the individual candidates nominated by that party. Which candidates are elected is simply determined by the order in which their party decides to place them on 8

10 their list. (Farrell: 1997: 71-77). As a result, it is argued, it is the parties, not the electorate, to whom individual legislators are beholden. However, only a limited amount of research has been conducted on public reactions to different forms of party lists. Qualitative analysis (e.g. Home Office, 1998; Farrell and Gallagher, 1999: ) suggests that closed lists are unpopular, with voters reacting against the perceived loss of control over which candidates get elected. However, the State of the Nation poll carried out in February 1998 suggested that only a thin majority preferred open to closed lists, with the DE social groups markedly less enthusiastic about open lists than their AB counterparts (Dunleavy, Margetts and Hix, 1998). 5 There is then reason to doubt the force in the public mind of each of these four criticisms of proportional representation. However, previous survey research on this subject has had to confront the fact that if voters are not very well informed about the subject, there is a limit to how far they can be expected to answer questions on the specific claims made for and against electoral reform. But after experiencing proportional representation for themselves, we could expect voters in Scotland and Wales to be able to respond to a more extended battery of questions that did focus on particular aspects of the reform debate. Moreover, the additional member system deployed in both countries used closed rather than open lists in allocating top-up seats, while in Scotland, at least, the election resulted in the formation of a coalition government. As a result, our survey results can provide important clues as to the likely force of these four criticisms during the course of any referendum campaign. Data and methodology 5 Under an open list system voters can express a preference for at least one individual candidate and which candidates are elected for each party is determined by which of them gets the most personal votes. 9

11 Our data were collected as part of the 1999 Scottish Parliament Election Study (SPES) and Welsh Assembly Election Study (WAES). 6 In the weeks immediately after the election, these studies interviewed a random sample of people designed to be representative of the adult population in the two countries. The interviews in Scotland (n=1,482) were conducted face to face in people's homes, while those in Wales (n=1,256) comprised, as part of a methodological experiment, a mixture of face to face and telephone interviews. Nearly all of the questions about electoral reform were asked as part of a self-completion booklet that respondents were asked to complete after they had finished the main interview. Inevitably some respondents did not complete this task, while in Wales we were only able to ask those who had undertaken a face to face interview to complete the booklet. Thus, most of the figures reported here are based on the answers of 1,169 people in Scotland and 440 in Wales. This latter figure means we have to be particularly careful about drawing conclusions from any small differences we might find within Wales. The first of the two key questions that we propose to address however requires us to be able to examine how opinion in Scotland and Wales might have changed after experiencing proportional representation for the first time. We cannot of course easily do so if we only have access to a survey undertaken at a single point in time. Fortunately a number of the questions that we asked in our survey were also included in the 1997 Scottish Election Study undertaken after the May 1997 general election, and the 1997 Scottish and Welsh Referendum Studies conducted after the September 1997 referendum on devolution, thereby giving us vital measures of opinion in the two countries prior to the devolved elections. The election study interviewed 882 people, while the two referendum studies interviewed 659 in Scotland and 632 in Wales (for further details see Brown et al, 1999; Taylor and Thomson, 1999). 6 The SPES was directed by Alison Park, John Curtice, David McCrone, Lindsay Paterson and Paula Surridge; the WAES was directed by Richard Wyn Jones, Anthony Heath and Katarina Thomson. Both surveys are funded by the Economic and Social Research Council. 10

12 The fruits of experience So, what did voters in Scotland and Wales make of proportional representation? Did they approve of their new electoral system? And, more importantly, what impact did experience apparently have on their attitudes? Was there a swing for or against proportional representation? And did their responses to differently worded questions become more consistent? We asked our respondents two questions, as detailed in Table 1, designed to measure their overall view about the use of proportional representation in the devolved elections. Both were modelled on questions that had previously been asked about House of Commons reform on the British Social Attitudes and British Election Study series (Curtice and Jowell, 1997). The first, shown in part a of the table, has consistently recorded a higher pro-reform response than the second (in part b). Doubtless, one of the reasons for this is that some respondents are reluctant to say they disagree with a single statement, a problem that is avoided if respondents are asked to choose between two propositions (Schuman and Presser, 1981). However although the first question does secure a higher level of support for proportional representation than the second does, we also find that a clear majority support it on the second question as well. Having experienced it in action, there appears to be clear support for the new electoral system in both Scotland and Wales. Table 1a How much do you agree/disagree that the Scottish Parliament/Welsh Assembly should be elected using proportional representation? Scotland % Wales % Strongly Agree/Agree Neither Agree nor Disagree Strongly Disagree/Disagree

13 Table 1b Some people prefer the new way of voting for the Scottish Parliament [Welsh Assembly] as they say it means all parties are fairly represented. Others say that the old way of voting for elections to the House of Commons is better as it produces effective government. Which comes closer to your own view? Scotland % Wales % Keep to the new way of voting Use the old way of voting But had Scottish and Welsh hearts grown fonder, and to what extent? To answer this question, we compare attitudes to proportional representation in May 1999 on the first of our two questions 7 with those recorded after the general election in May 1997 (on which we have data for Scotland only) and after the devolution referendums held in Scotland and Wales in September In Table 2 we show the resultant time series for Scotland. This shows that there was, indeed, an increase in support for proportional representation for the Scottish Parliament between 1997 and May In May 1997, just 53 per cent agreed or strongly agreed that the Scottish Parliament should be elected using proportional representation; by May 1999 that figure had risen to 66 per cent. However this rise in support does not appear to have been a response to seeing proportional representation in action. For support for its use was already as high as 75 per cent by the time of the referendum on devolution in September In short, it appears that the reality of a proportional system reduced rather than increased support for its use. All that may be argued in defence of the idea that the more that voters know about proportional representation, the more they like it, is that the case for proportional representation was apparently won during the campaigning for a Yes vote in the devolution campaign. 7 The second question was not asked in the earlier surveys. 12

14 Table 2 How much do you agree/disagree that the Scottish Parliament should be elected using proportional representation? May 1997 (%) Sept 1997 (%) May 1999 (%) Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree Experience of proportional representation does not appear to have increased support for proportional representation in Wales either (see Table 3). True those in favour of proportional representation rose slightly from 52 per cent after the devolution referendum to 57 per cent after the first devolution election; but then so equally did opposition (from 9 per cent to 14 per cent). So while the experience of proportional representation in Wales may have enabled some people to make their minds up about the subject, the effect appears to have been to polarise opinion rather than simply persuade voters of the merits of proportional representation. Table 3 How much do you agree/disagree that the Welsh Assembly should be elected using proportional representation? May 1997 (%) Sept 1997 (%) May 1999 (%) Strongly agree Agree Neither agree nor disagree Disagree Strongly disagree

15 Our results so far then give little reason to suppose, as suggested by Farrell and Gallagher, that the effect of exposing voters to new electoral systems is likely to run in one direction only, that is to encourage voters to embrace proportional systems. Of course it is possible that the effect that they anticipate is not evident in our figures because of the presence of other countervailing influences. 8 But even if this is so, it still implies that, if it is present, the impact of knowledge cannot be considered sufficient to ensure that a referendum campaign would be won by a pro-reform campaign. We are thus inclined to re-affirm the conclusion that the outcome of any referendum will be shaped by the quality of campaigning on both sides (Dunleavy and Margetts, 1997; Curtice and Jowell, 1998), and not just on the degree to which the reform camp can raise voters understanding of the issues. Still, perhaps the knowledge gained by the experience of devolution meant that voters became more consistent in the answers they gave to questions about proportional representation. If so, this might suggest that while experience of proportional representation does not necessarily produce a beneficial reaction, at least voters do end up with firmer views on the subject. That at least might mean that, as the campaign proceeds, we could gradually expect voters to become less likely to change their views in response to the blandishments that they hear in any referendum campaign. In Table 4 and Table 5 we show the answers our respondents gave to the two questions about electoral reform for the House of Commons which were the model for the questions detailed in Table 1, and which have previously resulted in highly inconsistent results (Curtice and Jowell, 1997). We can see that they still produced very different results. In Scotland, nearly three in five backed the introduction of proportional representation in response to the first question, but only two in five did so when presented with the second 8 One possible explanation for the drop in support for proportional representation in Scotland between September 1997 and May 1999 could be that it reflects the disappointment of those who lost out under the new system. For example, Labour supporters might have been disappointed at failing to win an overall majority while Nationalists might have been disappointed at failing to win power at all. However, the fall in support for proportional representation happened amongst all party identifiers. 14

16 question. The difference in Wales is almost of identical proportions. It appears that in fact experience did little to generate consistency on the subject. 15

17 Table 4 How much do you agree/disagree that the UK should introduce PR so that the number of MPs each party gets in the House of Commons matches more closely the number of seats each party gets? Scotland (%) Wales (%) Strongly agree/agree Neither agree nor disagree Strongly disagree/disagree Table 5 Some people think we should change the voting system for general elections to the House of Commons to allow smaller parties to get a fair share of MPs. Others say that we should keep the voting system as it is, to produce effective government. Which comes closer to your own view? Scotland (%) Wales (%) Change the voting system Keep it as it is This is confirmed if we compare the results we obtained for Scotland in our 1999 survey with those obtained two years previously when both questions were also asked in the 1997 Scottish Election Study. As we have seen in 1999 there was a 20-point difference between the two questions in the level of expressed support for proportional representation. In 1997 the equivalent gap was just 15 points. Of course, it may be the case that it is unreasonable for us to expect the experience of using proportional representation to elect one institution should have increased the consistency of answers given to questions about its use for another institution. But our results do give reason to doubt that voters' apparent ambivalence towards proportional representation will be easily resolved in any referendum campaign. Doubts about reform 16

18 We identified earlier four arguments that have been used against proportional representation, arguments that might indeed help account for the apparent ambivalence of voters towards reform. These arguments were that proportional representation is too complex, and that it means coalitions, multi-member constituencies and party lists, all of which British voters do not like. We now examine how far these arguments appear to have been an issue for voters in Scotland and Wales. The complexity of an electoral systems is of course relative concept: how many voters must find a system perplexing before it is deemed to be complex? And how much do voters need to know in order to be able to participate effectively? Neither question can be answered straightforwardly. But one useful distinction that can be drawn is between whether voters understand enough to use an electoral system and whether they know so much that they can be said to understand how it works (Farrell and Gallagher, 1999). Our measures of voters' understanding of the new electoral system enable us to identify where they lie on this spectrum. Our first measure is based on subjective evaluations, that is, how voters themselves experienced the new arrangements. This suggests that at the very basic mechanical level, voters experienced few problems. Only ten per cent of voters in Scotland, and 13 per cent in Wales, found it difficult to fill in the ballot paper. We might anticipate that similar proportions would say the same about a first past the post ballot paper. On the other hand when it came to understanding how the Additional Member System worked, voters acknowledged they had difficulties (see Table 6). In both Scotland and Wales, around two in five people said they found it difficult to understand how votes translated into seats. Indeed, in Wales though not in Scotland, those who said it was not difficult were outnumbered by those who said it was. 17

19 Table 6 Some people say that it is difficult to understand how the number of seats a party wins is worked out under the new system. Others say that it is not at all difficult. How difficult do you think it is understanding how seats are worked out? Scotland (%) Wales (%) Very difficult 7 9 Fairly difficult Not very difficult Not at all difficult 12 7 Can t choose Similar results were found by our second measure, that is the answers to an objective test of knowledge that we also administered (for further details see Curtice et al, 2000). In Scotland, for example, as many as 78 per cent correctly identified as 'True', the statement that 'You are allowed to vote for the same party on the first and second vote'. Yet only 30 per cent correctly identified as 'False' the claim that, 'The number of seats won by each party is decided by the number of votes that they get'. In Wales the equivalent figures were 63 per cent and 21 per cent. Knowing how to fill in the ballot paper is one thing; understanding how it all gets translated into seats is evidently quite another. What then appears crucial in the argument about complexity is how much voters need to know. Do voters need to understand how votes are converted into seats in order to be able to participate effectively? We have one clue that suggests that they might not. The level of abstention amongst those who said they thought it was difficult to understand how votes were translated into seats was no higher than it was amongst those who did not think it was difficult. In contrast, the small minority that felt the ballot paper was difficult to fill in were twice as likely to abstain as the majority who did not. Voters do need to feel comfortable with the ballot paper if they are not to be discouraged from voting, but on the other hand 18

20 understanding how the seats are calculated may not be particularly important. And by that test at least, voters in Scotland and Wales did not find their first experience of proportional representation too complex. What then of the second claim, that voters do not like the coalition governments that are the likely product of proportional representation, especially given that in Scotland at least the election was actually followed by the formation of a coalition. 9 Mindful that survey responses in this area might also be sensitive to question wording, we again administered more than one question on the subject. The first of these addressed the principle of singleparty versus multi-party rule in general without referring to any particular institution. In Scotland, well over half (55 per cent) believed it "better to have two or more parties in government so that more people s views are represented. Only one in three said that it was 'better to have just one party in government so that it is very clear who should be blamed if things go wrong (33 per cent). In Wales the gap was narrower, but even here more (47 per cent) supported multi-party rule than backed single party government (41 per cent). However, views were not so clear cut when we asked our respondents which form of government they preferred in their own devolved legislature. As Table 7 shows, in Scotland no less than two in five would prefer single party government in Edinburgh while in Wales the majority for coalition disappears. 9 The bulk of the fieldwork in Scotland was carried out in May and June, after the Partnership for Scotland was formed (14 th May 1999). 19

21 Table 7 Which do you think would be generally better for Scotland [Wales] nowadays? To have a government in Edinburgh [Cardiff] formed by one political party on its own? To have a government in Edinburgh [Cardiff] formed by two political parties together, in coalition? Scotland (%) Wales (%) Voters may then favour the idea of coalition more in the abstract than in reality. Yet we should note that on both of our measures support for coalition is higher in Scotland, where a coalition was formed, than in Wales where it was not. If coalitions are an anathema to the public, as indeed much of the media commentary surrounding the formation of the Scottish coalition appeared to assume, then we would expect to find that the idea of a coalition was less popular in Scotland than in Wales. But this is evidently not the case. Indeed, if anything, the idea of coalition government was more popular in Scotland after its first devolved election than it was before. In 1997, only 35 per cent of Scots thought that coalition government would be better at Westminster while as many as 56 per cent backed single party government. In 1999 in contrast, those favouring coalition government, 48 per cent, actually outnumbered those who backed single party rule (47 per cent). Of course, whether people in Scotland will continue to feel that way about coalition government when they have experienced its fruits for some period of time remains to be seen. But so far at least neither coalition nor complexity appear to be criticisms of proportional representation that resonate with the public in the light of experience. 10 But 10 We should also note that two of the arguments often used by those opposed to coalition government were decisively rejected by Scottish and Welsh voters. Just 16 per cent of both electorates believed that the new voting system gives too much power to small parties, with 44 per cent (Scotland) and 35 per cent (Wales) disagreeing. Similarly, neither group of voters thought that the new system would lead to 20

22 what about the merits of single-member rather than multi-member districts? Are the frequently made claims of politicians themselves that single member districts ensure the electorate is better represented supported by the public even when they have experience of both? It appears that they are. 11 As Table 8 shows in both Scotland and Wales more than six in ten voters prefer single-member constituencies to multi-member areas. We did not examine views on what is claimed as a particular virtue of the Additional Member System, namely that it includes both types of constituency (Linton and Southcott, 1998). But it may well be that one of the reasons why their new electoral system apparently is acceptable to Scottish and Welsh voters (Table 1) is because it retains single member constituencies. Certainly, our results suggest that proportional systems that dispense with this arrangement will face some scepticism from British voters. Table 8 People have different views about how their area should best be represented in a parliament. Which of these statements comes closest to your view? Scotland (%) Wales (%) I would rather have one member of parliament fo the area I live in? I would rather have several members of parliamen possibly from different parties, but covering a larger area Can t choose unstable government. In Scotland, more than one in four voters rejected this view, against just 14 per cent who agreed. In Wales, just 16 per cent agreed, while over one third rejected the proposition. 11 Our survey was of course conducted before many voters would have had much experience of how constituency and list members operated in practice. However, in Scotland at least, their respective roles became the subject of public argument soon after the election following attempts by the Labour, party to limit the allowances paid to list MSPs (McCabe and McCormick, 2000: 41). Voters might then have been expected to have developed some views on this subject, though doubtless it will be particularly important to revisit voters' views on this subject by the time of the second devolved election in

23 The final criticism often made by proportional systems, and one that can certainly be levelled against the Additional Member System as implemented in Scotland and Wales, is that closed party lists put power in the hands of parties rather than voters. Indeed this issue provoked a major constitutional rumpus in the parliamentary session, with the House of Lords passing amendments on five separate occasions to provide for open lists for European Parliament Elections in defiance of the government which wanted closed lists. 12 Our results confirm that, at least so far as people in Scotland and Wales are concerned, their Lordships were closer to public opinion than the government. As many as 43 per cent of people in Scotland and 44 per cent in Wales agreed that they would have preferred to have voted for individual candidates on the regional vote rather than for a party list. Just 17 and 14 per cent respectively disagreed. Only around one in four voters in both Scotland and Wales agreed that parties rather than voters "should decide which of the candidates on their regional list get the seats their party has won", while around half disagreed. And more voters agreed than disagreed that those elected through the regional party list "will be too much under the control of their party". However, the strength of public opposition to closed lists should not be exaggerated. On none of these questions does the proportion who did not express a clear view, that is either say they did not agree or disagree or said that they did not know, fall below a quarter; in one case it was over a half. The public may be suspicious of party control, but for many of them it is not evidently a particularly salient objection. Conclusion The first devolved elections gave voters in Scotland and Wales their first experience of proportional representation in action. Although that experience undoubtedly still has to mature, by looking at their reactions to proportional representation in the aftermath of the 12 See HL Deb, 4 th November 1998, cols Also see the Second Reading debate in the House of Commons: HC Deb, 25 th November 1997, cols

24 first devolved elections, we have been able to cast significant light on a number of controversies about attitudes towards electoral reform in Britain. People in Scotland and Wales largely reacted to their new electoral system with approval. The use of proportional representation certainly did not detract from and may well have contributed to the legitimacy of the new institutions. Nevertheless, we have also found that becoming better acquainted with proportional representation does not necessarily mean more public support for reform. Indeed, it looks perfectly possible as apparently happened in Scotland that, on occasion, the opposite could happen. Meanwhile, there is no evidence from our study that acquaintance makes voters more consistent in their views. As a result it seems to us that the outcome of any referendum on electoral reform for the House of Commons must be considered uncertain. Much will depend on the relative success of the two camps in the way that the manage to frame the issue in voters' minds. At the same time we have also been able to examine which criticisms of proportional representation appear to resonate with a relatively well acquainted public, and which appear to have less force. The claims that proportional representation is too complex and that it results in unstable coalitions appear to fall into the latter group. In contrast the public does appear to be suspicious of party lists, but at the same time may well not be much exercised by them. Rather, the Achilles heel of the pro-reform argument appears to be the public's regard for having their own local MP. Even after being given an electoral system that gives them both single member and multi-member constituencies, voters in Scotland and Wales still exhibited a clear preference for the former. Fortunately for the pro-reform camp, the proposals put forward by the Jenkins Commission attempt to meet these two criticisms of proportional representation head on. The commission proposes that those MPs who are elected through the 'top-up' part of the system should be elected on 'open' rather than 'closed' party lists. Meanwhile it is suggesting that around five-sixths of MPs should still be elected through the familiar single member plurality system. Whether this latter feature will be enough to weaken the force of 23

25 the claims of the champions of single member constituencies remains to be seen - so long, that is, as the referendum on the Jenkins' proposals is eventually held. 24

26 References Brown, A., McCrone, D., Paterson, L. and Surridge, P. (1999), The Scottish Electorate; the 1997 election and beyond. London, Macmillan. Budge, I. (1998) Stability and Choice: Review of Single Party and Coalition Government, University of Essex: Democratic Audit Crewe, I. (1985) MPs and their constituents in Britain: How Strong are the Links?, in Bogdanor, V. (ed) Representatives of the People? Parliamentarians and Constituents in Western Democracies, Aldershot: Gower Curtice, J. (1993), 'Popular Support for Electoral Reform: The Lessons of the 1992 Election', Scottish Affairs, no. 4, Curtice, J. and Jowell, R., 'Trust in the Political System', in Jowell, R., Curtice, J., Park, A., Brook, L., Thomson, K. and Bryson, C. (eds.), British Social Attitudes: the 14 th. report: The end of Conservative Values? Aldershot: Ashgate. Curtice, J. and Jowell, R. (1998) Is There Really a Demand for Constitutional Change?, special edition of Scottish Affairs, Curtice, J., Seyd, B., Park, A. and Thomson, K. (2000) Wise After the Event? Attitudes to Voting Reform Following the 1999 Scottish and Welsh Elections, University College London: The Constitution Unit Curtice, J. and Shively, P. (2000), 'Who Represents Us Best? One member of Many?', paper presented at the 18 th World Congress of the International Political Science Association, Quebec. Delli Carpini, M. and Keeter, S. (1996), What Americans Know about Politics and Why it Matters, New Haven: Yale. Dunleavy, P., Margetts, H. and Weir, S. (1993), 'The 1992 election and the legitimacy of British democracy', in Denver, D., Norris, P., Broughton, D., and Rallings, C. (eds.), British Elections and Parties Yearbook 1993, Hemel Hempstead, Harvester Wheatsheaf. Dunleavy, P. and Weir, S. (1995) Sleaze in Britain: Medial Influences, Public Response and Constitutional Significance, Parliamentary Affairs, 48(4), Dunleavy, P. and Margetts, H. (1997) The Electoral System, Parliamentary Affairs, 50(4),

27 Dunleavy, P., Margetts, H., O Duffy, B. and Weir, S. (1997) Making Votes Count: Replaying the 1990s General Elections under Alternative Electoral Systems Dunleavy, P., Margetts, H. and Hix, S. (1998) Open or Closed Lists Voting for the European Parliament Elections 1999: The State of the Nation Report, paper to Constitution Unit/ Nuffield College seminar on Regional Lists, 20 th February Dunleavy, P. and Margetts, H. (1999) Mixed Electoral Systems in Britain and the Jenkins Commission on Electoral Reform, British Journal of Politics and International Relations, 1(4), Farrell, D. (1997) Comparing Electoral Systems, Hemel Hempstead: Prentice Hall/Harvester Wheatsheaf Farrell, D. and Gallagher, M. (1999) British Voters and their Criteria for Evaluating Electoral Systems, British Journal of Politics and International Relations, 1(3), Home Office (1998) Attitudes to Euro-elections and Electoral Reform, qualitative research carried out by NOP Jenkins Commission (1998) Report of the Independent Commission on the Voting System, chaired by Lord Jenkins, Cm 4090-I, London: The Stationery Office Kellner, P. (1992), 'The devil-you-know factor', Representation, 31, Kerley Committee (2000) The Report of the Renewing Democracy Working Group, chaired by Richard Kerley. Edinburgh: The Stationery Office Lijphart, A. (1999) Patterns of Democracy: Government Forms and Performance in Thirty Six Countries, New Haven: Yale University Press Linton, M. and Southcott, M. (1998) Making Votes Count: The Case for Electoral Reform, London: Profile Books Luppia, A. (1994), 'Short Cuts versus Encyclopaedias: Information and Voting Behaviour bin California Insurance Eleections', American Political Science Review, 88, McCabe, A. and McCormick, J. (2000) Rethinking Representation: Some Evidence from the First Year, in Hassan, G. and Warhurst, C. (eds) The New Scottish Politics: The First Year of the New Scottish Parliament and Beyond, Edinburgh: The Stationery Office, Sartori, G. (1997) Comparative Constitutional Engineering: An Inquiry into Structures, Incentives and Outcomes, Basingstoke: Macmillan 26

28 Schuman, H. and Presser, S. (1981) Questions and Answers in Attitude Surveys: Experiments on Question Form, Wording, and Context, New York: Academic Press Smith, T. (1991) Citizenship and the British Constitution, Parliamentary Affairs, 44(4), Vowles, J. (1999), 'Electoral Systems and Proportional Shares of Government: Renewing the Debate', paper presented at the Annual Meeting of the American Political Science Association, Atlanta. Vowles, J. (2000) Evaluation Electoral System Change: The Case of New Zealand, paper presented at the International Political Science Association conference, Quebec City, August 1-5. Weir, S. (1992) Waiting for Change: Public Opinion and Electoral Reform, Political Quarterly, 63, Wyn-Jones, R. and Trystan, D. (1999) The 1997 Welsh Referendum Vote, in Taylor, B. and Thomson, K. (eds) Scotland and Wales: Nations Again? Young, K. (1985) Shades of Opinion, in Jowell, R. and Witherspoon, S. (eds) British Social Attitudes: The 1985 Report. 27