![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
There has been a dramatic change in the effect that special votes have on the results of general elections in New Zealand. Under the former first-past-the-post system for electing Members of Parliament, special votes favoured the National Party. For example, during the 12 successive elections from 1960 through to and including 1993, Labour Party candidates received an average share of the special votes cast that was 1.58 percent less than their share of election-night votes, while National Party candidates gained, on average, 2.19 percent more special votes than their initial share of the provisional votes tallied on election night. Since the implementation of New Zealand’s mixed member proportional (MMP) representation voting system, however, special votes have favoured the Green and Labour parties. This article documents the effects that special votes have had in New Zealand for a period of more than 50 years and examines why a major change has occurred in the impact that special votes have on the fortunes of New Zealand’s political parties.
Acknowledgments
The authors would like to thank Robert Peden and Alicia Wright (respectively, a former and current Chief Executive of the New Zealand Electoral Commission) and Anthony Pengelly (a staff member at the Commission) for insights and suggestions made during discussions with them about the topic. Mr Pengelly also provided several data files for this analysis: We would also like to thank the journal’s anonymous reviewers for their helpful feedback.
The data for this article were taken on an electorate-by-electorate basis from the official published results of New Zealand general elections contained in the H.33 and E.9 publications in the Appendices to the Journals of the House of Representatives, 1961 to 2018.
The research for this article was not undertaken with the aid of any grants.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. There was one other possible case during the 1960–1993 period of a change in party fortunes as a result of the special vote count. On election night 1987, National’s candidate in the Wairarapa electorate had a 65-vote margin over Labour’s incumbent MP, Reg Boorman. The final count awarded the seat to Boorman by a margin of seven votes, but after an official recount that was reduced to one vote. Following a petition to review the results, however, an Electoral Court ruled that a total of 125 votes in the electorate were invalid and the Court consequently reduced Boorman’s tally by 78 votes and Creech’s total by 43 votes (and the Democrats’ candidate by four). The nett effect of all these manoeuvres was that on final Court-mandated figures Creech managed to hold the electorate that he had nominally won on election night by a final margin of 34 votes.
2. Party-only special votes are votes cast by electors who voted for an electorate other than the electorate for which they were qualified to vote, for example a voter enrolled in Christchurch Central who votes for Christchurch East. Their party vote will count but their electorate vote will not. Prior to 2002, the entire ballot was disallowed. Note that the party-only votes category does not include votes where the voter cast a vote for the correct electorate but did not cast a valid electorate vote (for example by leaving it blank or ticking multiple candidates). These types of votes are fully allowed and are included in the party vote results, it is just that their electorate vote would be recorded as informal.
3. The Electoral Amendment Act 1990 provided that a person who applied for registration after writ day and by 4pm on the day before polling day could vote by special vote and this applied to the 1990 election. The Electoral Amendment Act 1993 (an amendment to the 1956 Act) restored the 1987 provision so that only those who became eligible between 31 days before writ day and the day before polling day could vote. and this was applied at the 1993 general election. The Electoral Act 1993 (brought into force by the referendum on the electoral system) maintained the 1987 position. It was the Electoral Amendment Act (No 2) 1995 that introduced the current provision which applied at the 1996 and subsequent elections (noting though that the 4:00 pm deadline for the application to be received no longer applies). We are very grateful to former Electoral Commission Chief Executive Robert Peden for this clarification.
4. A marginal seat is one that requires a two-party swing of 5 percent or less to change hands; a fairly safe seat needs a two-party swing of between 5 and 10 percent to fall; and any seat requiring a swing of more than 10 percent to switch allegiance is classified as a safe seat. See Roberts (Citation1975, 112).
5. The 120 safe Labour seats illustrated in do not include the four Māori seats that were in existence for each of the seven general elections from the which the data for the Figure were extracted and analysed. This is because – as is discussed elsewhere in this article – special voting in the Māori electorates was considerably higher than special voting in the general (i.e. non-Māori) electorates, not least because of the fact that not every polling place in the Māori electorates was a Māori electorate booth. They were frequently general electorate polling places only, and in order to vote at them, Māori electors had to cast a special vote.
6. It is worth recalling that in 2005 a great deal of voter attention was focussed on which major party – Labour or National – would win the most votes (and thus also the most seats in Parliament) as a result New Zealand First’s pledge to negotiate initially with the party that won the most votes. In 2017, there were late changes in leadership for both the Labour and the Green parties – changes that were viewed positively for Labour but more negatively for the Greens.
7. An overhang occurs when a party wins ‘more seats in parliament than its proportion of the party vote would have entitled it to’ (Levine and Roberts Citation2006, 346). As Shugart and Wattenberg have pointed out, the word overhang comes ‘from the German term űberhangmandate’ and ‘overhang seats actually increase the total number of seats in the chamber’ (Citation2001, 23). Four of the first eight MMP elections in New Zealand produced overhang seats in the House of Representatives. There were no overhang seats after the 1996, 1999, 2002, and 2017 general elections. The 2005, 2011, and 2014 elections all resulted in a Parliament with one additional (i.e. overhang) seat; the 2008 election saw the Māori Party win two overhang seats, which meant there was thus a House of Representatives of 122, the largest in New Zealand’s history.
8. See Elklit and Roberts (Citation1996), regarding the significance of upper- or controlling-tier votes in two-tier proportional representation electoral systems.
9. It is worth noting that as early as the 1950s, Ralph Brookes made similar points to those by Vowles. See Brookes (Citation1953); and Brookes (Citation1959).
10. The Values Party contested five elections from 1972 to 1984 and was widely viewed as New Zealand’s first environmental-focused political party.
11. New Zealand First maintains it is centrally positioned in the party spectrum and, initially at least, negotiated actively with both Labour-led and National-led governing blocs. However, the party was only once (and briefly) part of a National-led government. New Zealand First has more commonly been part of Labour-led governments.
12. Provision for tangata whenua votes remained in the electoral regulations and if, for any reason, the Chief Electoral Officer or Electoral Commission decided against providing ordinary Māori voting facilities in any polling place, they would have been legally obliged to provide tangata whenua voting facilities at that polling place. Regulation 22, which provided for tangata whenua voting, was revoked in March 2017, which means that there is no longer provision in the legislation for tangata whenua voting. We are extremely grateful to former Electoral Commission Chief Executive Robert Peden for this clarification.
13. Official turnout in the 1984 general election was 93.71 percent. Three years later it had fallen to 89.06 percent, and in 1990 was 85.24 percent. In 1993 turnout fell still further – to 85.20 percent, which was less than 79 percent of the estimated voting age population (Electoral Commission Citation1997, 127).
14. See Section 27 of the Electoral Amendment Act (Number 2), 1995, to the Electoral Act 1993. We are extremely grateful to Bill Moore, Parliamentary Counsel Office/Te Tari Tohutohu Paremata for his help in clarifying this point.
Additional information
Notes on contributors
Therese Arseneau
Therese Arseneau specialises in the study of elections, electoral system reform, and issues of representation. She has lectured in political science at universities in Canada and New Zealand and is currently an Adjunct Senior Fellow in Political Science at the University of Canterbury, Christchurch, New Zealand.
Nigel S. Roberts
Nigel S. Roberts is an Emeritus Professor of Political Science at the Victoria University of Wellington. His main research fields are in comparative politics, focusing especially on electoral systems and voting behaviour.