Summary of the impact

This research on party and voter behaviour in the 2012 Scottish Local Government elections influenced the understanding and behaviour of two audiences. First, it directly informed the Scottish Parliament's Local Government and Regeneration Committee's inquiry into the 2012 Scottish Local Government Elections, which recommended reforms to electoral law as well as other measures to enhance local democracy. Second, it made a distinct contribution to improving the understanding of the Single Transferable Vote (STV) electoral system among party agents and candidates. In particular, there is clear evidence that it informed the campaign strategy of some Scottish National Party candidates and agents.

Underpinning research

After his appointment in August 2011, Clark conducted a research project on the behaviour of political parties and voters in the 2012 Scottish local government elections with the results being published in his 2013 Representation article [1]. The data collection and analysis for this project was wholly undertaken at Newcastle while the theoretical framework and methodology used in the study builds on Clark's earlier research on the 2007 Scottish local elections (conducted, with colleagues at Aberdeen and Lancaster, while he was at Queen's University, Belfast).

This study examines how Scottish parties and voters have adapted to the use of the Single Transferable Vote (STV) electoral system in the 2012 Scottish local government elections. STV is a form of proportional representation, which uses preferential voting in multi-member constituencies. Candidates must achieve a quota (or proportion of the votes) to be elected, which is determined by dividing the number of people who have voted by the number of positions to be filled. Voters can rank as many candidates as they choose in order of preference. If their first-preference candidate receives more than the required quota or receives the lowest number of votes, their vote is transferred to their second-preference candidate. This can be repeated multiple times so that third-, fourth- and even lower preferences may play an important role in determining which candidates are elected. Under STV, a single political party might win more than one seat in a multi-member constituency so a potentially effective strategy is to put forward two or more candidates. However, parties also need to think carefully about vote management to maximise their gains from both intra-party transfers (i.e. from candidates of the same party) and inter-party transfers (i.e. from candidates of another party).

STV remains a relatively unusual electoral system, which is currently used in national or local government elections in only a handful of countries. It was introduced for Scottish local government elections in 2007. In 2007, the local and Parliamentary elections, which used different electoral systems, were held on the same day. The 2012 elections were therefore the first time that STV had been used in stand-alone local elections in Scotland (or in Britain in living memory). Therefore, the effects of STV on party and voter behaviour should be clearer in the 2012 elections than they were in the 2007 elections.

This research used aggregate ward-level data from the 2012 elections in conjunction with observation of the election campaign to examine parties' strategies and voter behaviour. In particular, it examined:

(i) to what extent parties put forward multiple candidates in constituencies;

(ii) voter turnout;

(iii) the ability of voters to use STV;

(iv) the pattern of intra-party and inter-party vote transfers;

(v) whether the position of candidates on the ballot paper had an effect on their share of the vote.

The main findings were:

(1) Only the SNP and Labour put forward more than one candidate in a large proportion of wards (SNP: 69.6% of wards; Labour 48.1%) [1, pp.58-60].

(2) Voter turnout across Scotland was 39.8%, which compares favourably with turnout in standalone local elections in many other countries, including England. There was considerable variation between wards and surprisingly the lowest turnout was in two wards in relatively affluent Aberdeen. Further investigation of the potential causes of variations in turnout should be undertaken when demographic data from the 2011 census is available. It may be that these wards include relatively high numbers of students and itinerant workers [1, p.58].

(3) Scottish voters do not seem to have difficulty using STV. This is indicated by the low number of ballots rejected (only 1.7% compared with 2% in the 2011 Northern Ireland local government election with an electorate much more used to STV) and the number of voters taking the opportunity to express more than one preference (81.3% of voters expressed a second preference and 52.6% of voters expressed a third preference) [1, pp.61-3].

(4) Vote transfer patterns indicate high levels of intra-party "transfer solidarity": where any of the four main parties ran two candidates, over 70% of surplus votes were transferred from its first to second candidate; where the SNP, Labour and Conservatives ran only one candidate, over 30% of their voters did not express a second preference. The main beneficiaries of inter-party transfers were the SNP [1, pp.62, 64-5].

(5) There was a statistically significant relationship between position on the ballot paper and the number of first preferences a candidate received [1, p.65].

References to the research

Details of the impact

Clark's research on the 2012 elections had direct effects on:

• the recommendations of the Scottish Parliament's Local Government and Regeneration Committee's inquiry into the 2012 Scottish Local Government Elections; and
• the campaign strategy of some Scottish National Party candidates and agents.

Informing the Electoral Process

First, Clark's research directly influenced the recommendations made in the Scottish Parliament's Local Government and Regeneration Committee's Report on the 2012 Scottish Local Government Elections (June 2013) [IMP1]. The aim of the Committee's inquiry was to review the 2012 elections and make recommendations to the Scottish Government for measures to enhance local democracy. Clark's research contributed directly to the Committee's inquiry in four main ways:

(i) A pre-publication version of Clark (2013) [1] was included in the briefing papers for the initial Committee session in November 2012 [IMP2];

(ii) Clark was the only academic researcher invited to give expert oral evidence to the Committee in November 2012 [IMP3];

(iii) Clark submitted additional written evidence to the inquiry in March 2013 [IMP4];

(iv) Clark was further invited to an extended meeting directly with the MSPs acting as Reporters for the Committee in March 2013 [IMP1, Annexe A, pp.28-32].

The Parliamentary Report [IMP1] explicitly cites Clark's evidence in support of two recommendations:

(1) It recommends that public information campaigns should be "better targeted at under-represented groups to assist with their inclusion in the democratic process" [IMP1, para.17].

The Report explicitly states that this recommendation is directly informed by Clark's evidence in his meeting with the Reporters: "Dr Alistair Clark from Newcastle University told us that it may be easier to identify the reasons for low voter turnout in areas once 2011 census data is available and suggested a need to engage better and more often with student, ethnic minority and itinerant populations" [IMP1, para.16].

(2) It recommends that an alternative to alphabetical ordering of candidates on ballot papers should be "looked at in time for the 2017 Scottish local government elections" and that the Electoral Commission uses "opportunities that arise in the interim period to trial potential solutions" [IMP1, para.109]. In paragraphs 105-7, the Report refers to Clark's written evidence as well as his meeting with the Reporters. In particular, it cites his finding that there was a significant statistical relationship between ballot position and the relative vote share of candidates from the same party to support the argument that alphabetical ordering disadvantages some candidates [IMP1, para.106]. The Report also emphasises that based on his further examination of the ballot paper ordering issue [IMP4], Clark "suggested a degree of caution in proceeding with any change" [IMP1, para.107]. Moreover, this is reflected in the Committee's recommendation that different potential solutions should be considered before any decision is taken.

More generally, Clark's research re-assured the Committee that STV was easy to use and turnout was relatively high for standalone local elections as well as providing more detailed analysis of how STV had worked in practice. His work was described as "very much appreciated" by John McCormick, Electoral Commissioner for Scotland [IMP3, Col.1434].

Informing Party Strategy

Second, Clark's research also made a distinct and material contribution to improving the understanding of effective party strategy under STV among party agents and candidates, especially in the Scottish National Party (SNP). Success under STV depends on strategic vote management: parties and candidates need to understand how to maximise the number of their candidates who will be elected in each constituency. In 2007, the SNP were widely regarded as having failed to maximise the number of seats that they won by offering too few candidates. To help to avoid this problem in 2012, Clark was invited to deliver a presentation on the operation of the 2007 elections and his preliminary research on the build-up to the 2012 elections to a closed party plenary session at the SNP Spring Conference (10^th March 2012). The session was attended by approximately 250 SNP candidates or their election agents. This represents around 40% of all SNP candidates (n=613), and almost all those in attendance at the Conference, with the session advertised in the programme as one "all candidates and agents should attend" [IMP5].

Audience feedback from the session indicated that 79% of respondents (n=161) agreed with the statement: "The session will inform how I campaign during the 2012 council elections" [IMP6]. In their qualitative feedback, participants gave several specific examples of how their campaign strategy would change as a result of Clark's research [IMP7]:

"Didn't really understand STV before. Will be campaigning for 2nd and 3rd preferences as well — thank you"

"Our campaign was very much based on the need to get voters for us 1 and 2. Now we may have to chase votes 3 and 4, also"

"Will affect content of campaign literature — taking account of inter-party transfers as well as intra-party"

"I now have a much clearer understanding of STV and feel more confident when campaigning and explaining STV to voters and asking for their preferences"

Other participants indicated that the session had reassured them that the party strategy made sense [IMP7]:

"Has added a why to the how of STV campaigning — I now know precisely why we are doing things the way we are!"

"It removed my concerns over fielding two candidates and maybe splitting the vote and ending up with zero"

Clark also organised a workshop open to all candidates and parties at the Convention of Scottish Local Authorities (COSLA) Conference Centre, Edinburgh (12^th March 2012) on Elections under the Single Transferable Vote: A Workshop for Candidates and Councillors. He presented his research findings, with participation from the Electoral Reform Society and the General Secretary of Irish political party Fianna Fáil. Thirty candidates and councillors attended with representatives from all the major Scottish political parties and several minor parties, including a team from the Green Party, and Independents. In feedback, 60% of respondents (n=20) agreed with the statement: "The session will inform how I campaign during the 2012 council elections" [IMP8].

In sum, Clark's research on the use of STV in Scottish local elections has reached the key Parliamentary Committee considering the process of local elections in Scotland and has directly informed their recommendations for changes to the electoral process. In addition, approximately 1/4 of SNP candidates or their agents (161/613) indicated that Clark's research would inform their campaign strategy in the 2012 local government elections.

Sources to corroborate the impact

[IMP1] Scottish Parliament Local Government and Regeneration Committee, 8^th Report, 2013 (Session 4) Report on the 2012 Scottish Local Government Elections. Edinburgh: Scottish Parliament. SP Paper 358. Available at:
http://www.scottish.parliament.uk/parliamentarybusiness/CurrentCommittees/64870.aspx

[IMP2] Local Government & Regeneration Committee meeting Papers, 27^th meeting, 28^th November 2012. Available at:
http://www.scottish.parliament.uk/parliamentarybusiness/CurrentCommittees/58583.aspx

[IMP3] Scottish Parliament Official Record, Local Government & Regeneration Committee, 28^th November 2012. Available at:
http://www.scottish.parliament.uk/parliamentarybusiness/CurrentCommittees/58583.aspx

[IMP4] Clark, A. (March 2013) Written Evidence to the Scottish Parliament Local Government and Regeneration Committee. Available at:
http://www.scottish.parliament.uk/S4_LocalGovernmentandRegenerationCommittee/Inquir ies/13.__Dr_Alistair_Clark.pdf

[IMP5] SNP Spring Conference 2012 agenda. Copy available on request.

[IMP6] A summary of the survey responses from attendees at the SNP Spring Conference session. Available on request.

[IMP7] Hard copies of questionnaires completed by attendees at the SNP Spring Conference session. Available on request.

[IMP8] A summary of the survey responses from attendees at the COSLA workshop. Available on request. Hard copies of questionnaires completed by attendees are also available on request.