Summary of the impact

Researchers in the Centre for Media, Elections and Participation (CEMaP) have improved the quality of electoral administration and enhanced the quality of understanding and debate about electoral systems in New Zealand. Following a transition from a First Past the Post to a Mixed Member Proportional System in 1996, New Zealand has become an electoral reform model of international interest and the country held a referendum in 2011 on whether to retain the new system. CEMaP research has had impacts on the NZ Electoral Commission, the general public and electoral system campaigners. The main impacts of the research have been:

• improvements to electoral administration in recording official voting data in elections;
• better understanding by the Electoral Commission of voter attitudes towards elections for more effective electoral administration;
• improvements to information campaigns to increase public understanding of the mixed member proportional (MMP) electoral system and informing media debate during a national referendum;
• informing a national review of the MMP system including to maintain thresholds on party electoral support necessary for admission to Parliament in the light of public concern about the potential of MMP to admit too many small parties.

Underpinning research

Key researchers: Jack Vowles — Professor (2007-present); Jeffrey Karp — Senior Lecturer (2006-2007), Associate Professor (2007-2010); Professor (2010-present)

Researchers in CEMaP, Vowles and Karp, developed a programme of research about the New Zealand experience of electoral reform that has created a substantial body of evidence about the transition to the mixed member proportional system and its consequences. The research is grounded in two fundamental principles of democracy: that elections maintain governmental accountability and that elections serve to translate citizen preferences into public policy. Vowles and Karp's research assessed under which electoral rules and conditions these two principles are best conducted.

New Zealand is an important case of international interest because it made a transition from a first past the post, single member plurality, electoral system to a mixed member proportional (MMP) voting system in 1996. Under the MMP system, two types of representatives are elected: one from electorates and the other from party lists, making it a moderate form of proportional representation. A person may be both a candidate for an electorate and on a party list. Evidence about experience with this system is important to debates both in countries where first past the post systems have been criticised and for new democracies looking for important examples of electoral systems to learn from internationally. In New Zealand itself, a major referendum on whether or not New Zealand should retain the MMP system was held in 2011 (with 59 per cent supporting retention of MMP in a 74 per cent turnout). In this context, research on the operation of MMP has been important for policymakers, electoral campaigners and the public.

Vowles and Karp participated in six funded studies investigating electoral systems in New Zealand between 2008 and 2012 (see research grants). Vowles was part of a major US National Science Foundation funded project, Electoral Systems and Party Personnel: the Consequences of Reform and Non-Reform (2008-12). This international study examined how electoral systems affect `party personnel strategies' through an analysis of four countries with electoral system change (New Zealand, Bolivia, Germany, Portugal), and four without, as `steady-state' cases, (United Kingdom, Ukraine, Lithuania, Japan). Karp investigated voters and coalition government through the British Academy (2008-9) and voter validation in New Zealand through Nuffield Foundation (2009-2010) funding. Research from two funded studies (New Zealand Electoral Commission and Treasury 2008-9 and McDougall Trust 2011-12) was conducted as part of the New Zealand Election Study (NZES) which has provided extensive data to help evaluate, improve and sustain the quality of electoral democracy.

Findings from these studies were broadly supportive of the retention of the MMP system but identified significant problems in the administrative arrangements for national elections. The research found reasonably informed citizen participation (Vowles 2010) and that perceptions of fairness of MMP were a key correlate of support for the new system (Vowles 2008) and (Karp 2009). However, based on their analysis of NZES data, the research (Vowles 2010) found that the administrative procedures for recording special votes, where people vote at a polling place outside of their electorate/constituency, were insufficiently robust at the 2008 general election. Comparison with the official summary data showed many `special voters' had not been recorded on the master rolls as having voted. As a consequence, the Electoral Commission has decided to move toward more centralisation and computerisation of the systems to improve the quality of the electoral administration of special voters in future elections.

The research found generally positive public attitudes towards MMP but limitations in levels of understanding of the system. The US NSF funded research found that dual candidacy in MMP is not a serious limitation on candidate accountability. Relatedly, some voters prefer the coalitions that are more common under MMP with preferences for coalition government increasing from 46 per cent in 1999 to 56 per cent in 2010 (Miller and Vowles 2009; Bowler, Karp and Donovan 2010; Vowles 2011). The findings suggested that improved voter education about the system would lend further support to MMP as a system combined with maintaining features of the system which mitigate some of the public's concerns, especially that it could potentially lead to a large number of small parties being represented in New Zealand's Parliament. These findings have led to follow-up research on their implications by Karp and Vowles through interaction with the Electoral Commission and in evidence to the Review of the MMP to the NZ Minister of Justice, with impacts on the form of MMP retained in New Zealand (see section 4, details of impact).

References to the research

Research Publications

Research Grants:

Vowles, J. Electoral Systems and Party Personnel: the Consequences of Reform and Non- Reform, United States National Science Foundation (UK and NZ partner), US$34,000 (£21,235) (2008-12).

Vowles, J. NZ Election Study 2008, NZ Electoral Commission, NZ$15,000 (£7,225) (2008-9).

Vowles, J. NZ Election Study 2011, NZ Electoral Commission, NZ$20,000 (£10,300) (2011-12).

Karp, J. Voters and Coalition Government, British Academy, £5000 (2008-09).

Karp J. Social Desirability and the Survey Response: Validating Voter Turnout, Nuffield Foundation Social Sciences Small Grants Scheme, £6,306 (2009-10).

Vowles, J. New Zealand Election Study 2011, McDougall Trust, £7,000 (2011-12).

Quality of underpinning research: All grants obtained by peer reviewed, competitive bidding processes. All research articles were published in peer-reviewed journals. Electoral Studies had an impact factor of 1.6 and the British Journal of Political Science had an impact factor of 2.3 (2012 5 year Impact Factors). All outputs are available on request.

Details of the impact

Improving electoral administration in recording official voting data:
Vowles and Karp's finding of errors in the 2008 election recording of special votes in polling places revealed that many `special voters' had not been recorded on the master rolls as having voted. Misclassification in the 2008 dataset was as much as 40 per cent. The problem was subsequently also identified in data from the 2002 and 2005 elections, and was discussed and documented (Vowles 2010). The findings were communicated to the New Zealand Electoral Commission, which is responsible for the administration of national elections, in 2009, along with the implications of the problem. Recording of special voters for public inspection is a legal requirement and could be needed if an electoral petition challenging the result were to be initiated. As a result of Vowles and Karp's findings, the Commission `reviewed the clarity of [its] instructions in producing the master rolls' to Returning Officers in the 23 Electoral Offices across the whole country and `put greater emphasis on this area when training Returning Officers' (Chief Electoral Officer, personal communication, 2009).¹ Following up his earlier research, Vowles monitored the effectiveness of the changes in the 2011 NZES. He found a reduction in the problem although not its complete elimination, with 4 per cent of those validated as non-voters claiming they had cast a special vote. The Electoral Commission has now further responded by increasing automation in the process for scrutinising the master rolls in time for the 2014 election which should further improve administration.

Better understanding of voter attitudes for more effective electoral administration:
The research findings about public understanding of NZ elections (Miller and Vowles 2009; Vowles 2011) helped the Electoral Commission decide on appropriate administrative services for elections. Vowles is part of a group assisting the Electoral Commission in improving its strategy, including leading a workshop in August 2013, to inform their policy for the 2014 NZ election. The importance of the research to this beneficiary is reflected in the part-funding of NZES by the New Zealand Electoral Commission since 2005. The 2008 and 2011 NZES contained several questions developed jointly with the Electoral Commission. Findings about the 2011 election were communicated to the Commission in March and April 2012 with research helping the Commission assess the potential to introduce voting using the internet. Vowles reported findings that most people still prefer personal voting (46%) although 36% would prefer the internet to the Commission. This supports the Electoral Commission's policy of not introducing a wholesale change immediately but to move forward with a trial of online voting scheduled for elections in 2016.

Improvements to Electoral Commission information campaigns and informing media debate during the mixed member proportional system national referendum:
The research findings that there were misconceptions about MMP in public attitudes (Vowles 2010; 2011) helped the Electoral Commission's education strategy for public understanding in the November 2011 MMP referendum. The educational materials were effective with 88% of those who received Commission information feeling confident in their ability to make a meaningful vote, and, amongst all voters, 82% of registered electors knowing that the question would be whether they wanted to keep MMP or not (compared to only 35% in May 2011 before the campaign).² Vowles communicated with campaigners and media organisations prior to the referendum. In the national press, Vowles' (2011) research about MMP's effects on the accountability of representatives was used in media stories, including the largest circulation newspaper, the New Zealand Herald. On 16 Nov 2011 the paper reported the `quality analysis being contributed to the debate' by the research and how `Prof Jack Vowles nicely demolishes the anti-MMP argument about the supposed lack of accountability under the current system.' (http://bit.ly/ve6D9p). Rudman, an influential national columnist, wrote about the referendum campaign in the New Zealand Herald (Dec 14^th 2011) (http://bit.ly/1dxKQIw) after it had been completed. He cited Vowles' finding that `the survival of electorate-defeated MPs through dual candidacy is relatively rare' as important in Rudman's own support for the retention of MMP during the campaign. The media debate contributed to high public knowledge about the referendum which was further promoted by the researchers making NZES datasets available to all campaigners, with some using it to target their campaigns (Gray and Fitzsimmons 2012).³ The high level of knowledge was confirmed by data from the NZES showing that 9 out of 10 people were aware of the referendum.

Vowles and Karp's research helped confirm the legitimacy of the 2011 referendum outcome by showing that voters made an informed choice in retaining MMP in evidence to the Justice and Electoral Law Committee select committee.⁴ Karp's analysis confirmed that the public did have a preference for MMP as expressed in the referendum result undercutting criticism and reinforcing the legitimacy of the referendum result. (http://www.parliament.nz/resource/0000190224). Vowles extended his research on voter turnout, which had previously been used by the Electoral Commission², to show that MMP had not reduced turnout in elections as some critics of it had claimed (http://www.parliament.nz/resource/0000190244). A summary of findings appeared on the parliamentary website and triggered media debate. In the New Zealand Herald (29 May 2012), Vowles research was seen to challenge the media tendency of `turning elections into a two horse presidential style contest' because `this ignores the MMP reality that an election can be very close even though the gap between the two main parties is wide' (http://bit.ly/K6HaBL). Parliament Today (28 May 2012) also ran an article reporting the research, quoting it substantially.⁵

Informing a national review of the mixed member proportional system including to maintain thresholds on party electoral support necessary for admission to Parliament:
The research findings about public attitudes to MMP were used by a major review of MMP to the Ministry of Justice in 2012 and affected the review's recommendations. The research found that nearly half of New Zealanders believed that too many parties were represented in the New Zealand Parliament, including a third of people who voted for MMP in the 2011 referendum. Vowles submitted these findings as evidence to the MMP Review in May 2012 (http://www.mmpreview.org.nz/search/apachesolr_search/NZES) MMP potentially contributes to more parties being represented in Parliament depending on the threshold of the vote that parties have to achieve to obtain representation. The information about the potential threat to the legitimacy of MMP provided grounds for the Review Panel to resist arguments that the overall threshold, currently winning at least one electorate seat or 5% of all party votes, should be lowered significantly or abolished, which would have increased the number of parties in Parliament. The influence of the research is reflected in extensive citation in both the Proposals Paper⁶ and the Final Report (with 10 references throughout the Final Report⁷).

Sources to corroborate the impact

(numbers below refer to superscript notes in Section 4)

1. Chief Electoral Officer (CEO), New Zealand Electoral Commission, letter, 6^th November 2009 and letter 14^th Oct 2013. (Corroboration of impacts on recording of voting data, understanding of voter attitudes, and role of research as evidence in inquiries).
2. Report of the Electoral Commission on the 2011 General Election and Referendum to the Minister of Justice (http://www.parliament.nz/resource/0000188866).
3. Gray, S. and M. Fitzsimmons, (2012) `Defending Democracy: Keep MMP and the 2011 Electoral Referendum', in Johansson, J., and Levine, S., ed'. Kicking the Tyres: the NZ General Election and Electoral Referendum of 2011. Wellington, Victoria University.
4. Justice and Electoral Committee (2013) Inquiry into the 2011 General Election, NZ 50^th Parliament, April 2013 (http://www.parliament.nz/resource/0001871778).
5. Paliament Today 28^th May 2012 (http://parliamenttoday.co.nz/2012/05/media-blamed-for-low-election-turnout-2/).
6. NZ Electoral Commission, Review of the MMP Voting System Proposals, 13 August 2012. Wellington, NZ Electoral Commission (http://www.mmpreview.org.nz/proposals/overview).
7. Review of the MMP Voting System to the Minister of Justice, 29 October 2012. Wellington, New Zealand. http://www.mmpreview.org.nz/have-your-say/final-report.