by Paul Harris
Chief Executive
Electoral Commission of New Zealand415
Introduction
All democratic voting systems create incentives and disincentives which affect the behaviour of political parties and election candidates. It is well-established that different types of voting system can affect the nature of democratic representation and the composition of government. For example, countries which use a voting system based on proportional representation are more likely to have smaller parties represented in their legislatures, to have a multi-party governments, and to have representation from a variety of social groups. On the other hand, smaller parties find it more difficult to win seats under a simple plurality voting system, there tend to be single-party majority governments elected under that system, and the make-up of the legislature tends to be less diverse than in proportional systems.
However it is not only the type of voting system that is important. Variations within types of voting systems can also affect parties' chances of winning seats, and can affect the extent to which members of various social groups are present in the legislature. For example
- What criteria do parties have to meet to be eligible to contest an election? Do these criteria allow new parties to be formed and to compete with established parties on an equal footing?
- Is there a formal or effective threshold or share of the vote which parties have to cross to win seats? Is that threshold set at a reasonable level? Is it applied on a local, state, regional or national basis?
- Do voters elect representatives directly or indirectly through voting for parties? Are seats won or allocated to parties on a local, state, regional or national basis?
- What methods and criteria are used to decide electoral district boundaries? Are they more favourable to some parties than to others?
The choices of the type and details of the voting system used to elect a legislature can thus raise significant integrity issues. But it must always be remembered that there is no perfect voting system. Although it is important that a democratic voting system should be fair, equal and transparent, those goals can be achieved in different ways. It is a matter for each country to judge which voting system will best achieve them in the context of its own history, culture and traditions.
Integrity issues are also raised concerning the process by which decisions are made about changing from one type of voting system to another. For example:
- who should decide that a change is to take place?
- who should design an alternative system?
- how should the final decision be made?
- what opportunities should the public and political parties have to influence the design of a new voting system and the decision-making process?
In 1993 New Zealand changed its voting system for parliamentary elections from the First-Past-the-Post system (FPP) to a system of proportional representation based on the German model, known as 'Mixed Member Proportional' (MMP)416. This paper outlines the way that change took place. It is not intended that New Zealand's process and experience should be taken as the model which other countries should follow. Rather the New Zealand experience is presented as a case study of why one country decided to change its voting system and the process by which the decision was made.
Why did New Zealand change to MMP?
Why did New Zealand one of the world's oldest democracies make such a major change in its constitutional arrangements, from its well-established single-member plurality voting system to the MMP system of proportional representation?
Three sets of circumstances help explain that change:
1. Election results
New Zealand's FPP electoral system was finally established in 1913. After a period of adjustment, the Labour Party and the National Party had emerged by the mid-1930s as the two major political parties.
However the well-known effects of the FPP voting system began to be felt from the mid-1950s when the Social Credit party began to gain significant electoral support across the country but without winning many seats. In the eight general elections from 1954 to 1975, Social Credit gained an average of 9% of the vote but won only 1 seat, in 1966.417
Mutterings about whether the FPP voting system was fair turned into wider grumblings after the 1978 election because the Labour Party won 10,000 more votes (0.6%) than the National Party but had 11 fewer Members of Parliament (MPs) than National in a House of 92. Social Credit won 1 seat with 16% of the vote.
Labour's policy at the 1981 election was that there should be an authoritative re-examination of New Zealand's electoral system. Its concerns about the FPP voting system were reinforced after that election since once again National won more seats (4) than Labour even though Labour won 4,000 more votes than National. Such results confirmed Labour's suspicions that the way the FPP system was working in New Zealand was biassed against it and strengthened its resolve to review the electoral system as soon as it could. The 1981 election again underlined Social Credit's difficulties; it won 21% of the votes but only 2% of the seats.
2. Royal Commission on the Electoral System
Labour's opportunity came after it won the 1984 election. In 1985 it established an independent non-partisan Royal Commission to conduct a wide-ranging inquiry into the electoral system.418 The Commission had 5 members: a High Court Judge as Chairman, a law professor, a political science professor, a retired Government statistician, and a Mäori woman who was a statistician and social researcher.
The Royal Commission invited submissions from members of the public, and travelled around New Zealand to hear oral submissions. Members of the Commission also travelled to Australia, Canada, Germany, the United Kingdom and Ireland.
The Report of the Royal Commission was published in 1986. Two of its 69 recommendations were about the voting system that should be used to elect MPs: first, that FPP should be replaced by MMP, and secondly that the final decision should not be made by Parliament but by the people in a binding referendum after a period of public debate.
Why did the Royal Commission recommend a change from FPP to MMP? It tested FPP and other voting systems against 10 criteria:419
(a) Fairness between political parties. When they vote at elections, voters are primarily choosing between alternative party Governments. In the interests of fairness and equality, therefore, the number of seats gained by a political party should be proportional to the number of voters who support that party.
(b) Effective representation of minority and special interest groups. The voting system should ensure that parties, candidates and MPs are responsive to significant groups and interests. To facilitate this, membership of the House should not only be proportional to the level of party support but should also reflect other significant characteristics of the electorate, such as gender, ethnicity, socio-economic class, locality and age.
(c) Effective Mäori representation. In view of their particular historical, Treaty420 and socio-economic status, Mäori and the Mäori point of view should be fairly and effectively represented in Parliament.
(d) Political integration. While the electoral system should ensure that the opinions of diverse groups and interests are represented it should at the same time encourage all groups to respect other points of view and to take into account the good of the community as a whole.
(e) Effective representation of constituents. An important function of individual MPs is to act on behalf of constituents who need help in their dealings with the Government or its agencies. The voting system should therefore encourage close links and accountability between individual MPs and their constituents.
(f) Effective voter participation. If individual citizens are to play a full and active part in the electoral process, the voting system should provide them with mechanisms and procedures which they can readily understand. At the same time, the power to make and unmake governments should be in the hands of the people at an election and the votes of all electors should be of equal weight in influencing election results.
(g) Effective government. The electoral system should allow Governments in New Zealand to meet their responsibilities. Governments should have the ability to act decisively when that is appropriate and there should be reasonable continuity and stability both within and between Governments.
(h) Effective Parliament. As well as providing a Government, members of the House have a number of other important parliamentary functions. These include providing a forum for the promotion of alternative Governments and policies, enacting legislation, authorising the raising of taxes and the expenditure of public money, scrutinising the actions and policies of the executive, and supplying a focus for individual and group aspirations and grievances. The voting system should provide a House which is capable of exercising these functions as effectively as possible.
(i) Effective parties. The voting system should recognise and facilitate the essential role political parties play in modern representative democracies in, for example, formulating and articulating policies and providing representatives for the people.
(j) Legitimacy. Members of the community should be able to endorse the voting system and its procedures as fair and reasonable and to accept its decisions, even when they themselves prefer other alternatives.
The Royal Commission concluded that FPP rated well in relation to effective representation of constituents, voter participation, effective government, effective Parliament, and general legitimacy. In its view, however, FPP had 'severe weaknesses' in relation to fairness to major and minor parties, minority representation and Mäori representation.421
Having concluded that FPP was inherently flawed, the Royal Commission examined a number of alternative systems, particularly the Single Transferable Vote (STV) system; the Preferential Vote system used to elect the Australian House of Representatives; the system used in Germany, which the Royal Commission called the Mixed Member Proportional (MMP) system; and the Supplementary Member (SM) system.422 It concluded that MMP was the most suitable alternative to FPP, since it would be fairer to political parties and would improve voter participation, representation of Mäori423 or other groups, and be seen as more legitimate. The Royal Commission believed that MMP had 'comparable, though sometimes different' advantages over FPP in relation to effective government, effective Parliament, representation of constituents, effective parties, and political integration.424
Some of the basic provisions of New Zealand's electoral law have been 'entrenched' or 'reserved' since 1956 in that they can only be repealed or amended by a 75% vote of all MPs or by a majority of votes cast in a referendum held for the purpose. The matters protected in this way are the term of Parliament, the minimum age for qualification as an elector, the method of voting, and the membership of the independent body responsible for deciding electorate boundaries and the criteria it must follow in doing so.
The Royal Commission concluded that fundamental matters concerning the constitutional framework of government should continue to be entrenched. It considered that some of those provisions should only be changed by referendum rather than by a vote of 75% of all MPs but concluded that the choice of the method to be used to change a particular entrenched provision should be left to political judgement.425
The Royal Commission's recommendation that a referendum should be held on whether FPP should be replaced by MMP reflected its view that a change to a country's voting system affects the most basic aspects of its democracy and is therefore properly a matter which the people should decide.426
3. The political context
In 1986 the Royal Commission's recommendations on a change to MMP seemed the least likely to be accepted. Neither the Labour or the National parties supported a proportional voting system, although there were a few individual MPs in Labour (including the Deputy Prime Minister) and in National who did support a change.
So how did the change to MMP actually come about? A number of factors were at work. A pressure group, the Electoral Reform Coalition, was formed in 1986 to promote the case for a referendum on a change to MMP. In a live television debate during the 1987 election campaign, one of the leaders of this group extracted an undertaking from the Prime Minister that Labour would hold a referendum on the issue. It later emerged that the Prime Minister had misread his briefing notes and had omitted the word 'not'.427
In 1987 Parliament referred the Royal Commission's report to a parliamentary select committee. That committee reported in 1988 and recommended that the FPP system should remain. However the Labour majority on the committee recommended that a referendum should be held on a change to the Supplementary Member (SM) system.428 The Labour Government agreed that FPP should remain, but did not agree that a referendum should be held on a change to SM.429
However despite the lack of enthusiasm among the major political parties, the issue of the voting system refused to go away. It seems that a crucial factor provoking change was the prevailing public attitudes to politics and politicians in the decade from the mid-1980s. There was a widespread public perception that the Labour governments elected in 1984 and 1987 had deceived the voters by failing to carry out their election manifestos and by imposing new models of economic and state sector management against significant public opposition. Many people and groups began to look for ways to reassert popular and parliamentary control over governments. They saw the recommendation of the Royal Commission that there should be a referendum on a change to a proportional voting system as a means of doing so.
The strength of this public opinion was such that at the 1990 election all parties, major and minor, undertook to hold a referendum on the voting system.
The 1992 referendum
The National government elected in 1990 decided that the first electoral reform referendum would be held on 19 September 1992 coincidentally exactly 99 years to the day after women's' suffrage was achieved in New Zealand. This first referendum was held in order to establish whether there was a popular desire for a change to the voting system and if so which of four alternative 'reform options' discussed by the Royal Commission would be put to the voters at a second referendum Supplementary Member, Single Transferable Vote, MMP, and the Preferential Vote system.
Although formally speaking the 1992 referendum was 'indicative' or 'non-binding', the National government pledged that if a majority of those who voted at the referendum supported a change to the voting system, a second binding referendum would be held with the 1993 election at which voters would be asked to choose between the FPP system and the reform alternative which received the most support at the 1992 referendum.
The Government introduced a Bill to Parliament to authorise the holding of the referendum, to specify the form of the voting paper and the questions to be asked, to provide for the appointment of scrutineers to monitor the conduct of the referendum and the counting of the votes, to specify who was to declare the final result and how the result could be challenged, and to invoke certain provisions of the Electoral Act 1956 relevant to the conduct of the referendum. Public submissions were invited on the Bill, and Parliament passed an amended Bill in December 1991. There were no limits to the amounts that individuals or groups could spend on promoting particular answers to the referendum questions, although advertising and literature regarding the referendum had to include the name and address of the person who authorised it.
An independent ministerial Panel chaired by the Chief Ombudsman was appointed and funded to carry out public education concerning the referendum. However its efforts to do so were hampered by the lack of an official statement about how each of the alternative systems would be implemented.
The referendum resulted in a turnout of 55.2% of registered voters. The results were as follows:
I vote to retain the present First-Past-The-Post system. 15.3%
I vote for a change to the voting system. 84.7%
I vote for the Supplementary Member system (SM) 5.6%
I vote for the Single Transferable Vote system (STV) 17.4%
I vote for the Mixed Member Proportional system (MMP) 70.5%
I vote for the Preferential Voting system (PV) 6.6%
The 1993 referendum
The second binding referendum was held in conjunction with the 1993 general election. The government introduced two Bills into Parliament relating to this referendum which were passed after a select committee had considered submissions from the public. The law authorising the holding of the referendum was similar to that for the first referendum. However Parliament also passed a new Electoral Act to implement MMP430 which provided that the new electoral law would automatically replace the old electoral law if the official result of the referendum showed that a majority of those who voted in the referendum had voted for a change in the voting system.
The select committee made a very important change to the Bill to implement MMP when it decided to insert a provision requiring a select committee to review the new system beginning in April 2000 and reporting to Parliament by mid-2002. That review would therefore come after the second scheduled general election under MMP.431 In addition to specifying a number of matters which the review committee must examine, the law required the committee to consider whether any further referendum should be held on changes to the electoral system, and if so what issues should be put to the voters and when such a referendum should be held. This provision reassured many people that a change to MMP at the 1993 referendum would not be an irrevocable step, although many also thought that a further referendum was guaranteed rather than just a matter which the review committee was obliged to consider.
Once again, an independent Panel was appointed to carry out a government-funded programme of public education about the options at the 1993 referendum.432 Unlike 1992, the Panel had a detailed law to implement MMP and could therefore be specific on the details of both FPP and MMP.
The 1993 referendum resulted in a narrow but decisive public endorsement of MMP, by 54% to 46%. Because the referendum was held in conjunction with a general election, the turnout (85.2%) was much greater than at the 1992 referendum.
The transition to MMP, 1993-1996
Once the 1993 referendum had been carried, preparations for the new system had to begin in earnest. Two factors influenced the urgency of those preparations. First, because the number of electorate seats under MMP would be reduced from 99 to 65, an election under MMP could not be held until all the electorate boundaries had been re-drawn by the independent Representation Commission. That was completed in 1995. Second, four parties were elected to Parliament at the 1993 election and the National Party government had a majority of one.
The preparations for the coming of MMP took place on several fronts. First, MMP would almost certainly bring coalition or minority government, which would in turn bring significant changes to the ways the public service carried out its responsibilities in relation to governments and to ministers. The State Services Commission and the Cabinet Office therefore began work on the nature and scope of those changes, including sending senior personnel to Europe to study the operations of the public service under different types of proportional system.
Second, the Standing Orders Committee of Parliament undertook a major revision of Standing Orders to prepare for a multi-party Parliament elected by proportional representation. It too travelled to Europe to gain first-hand knowledge of the workings of different Parliaments elected by proportional representation.
Third, MPs and parties also began to prepare for MMP. Once it was known that one of the glues holding the old parties together was to be weakened, some MPs left their former parties to join other parties or to form new ones. That affected New Zealand's governing arrangements for 2 years prior to the 1996 election, and during that period the country experienced all the forms of government that are possible within its type of constitution. These changes passed largely unnoticed by the public. These pre-MMP changes also provided the public service and Parliament with invaluable pre-MMP experience of a multi-party majority and minority government. Nevertheless the government survived in its various forms until the first MMP election was held in October 1996.
The two MMP elections
As expected, no party won a majority of seats at the 1996 election. Two parallel coalition negotiations took place for nine weeks after the election, one between National and New Zealand First, the other between Labour and New Zealand First. Contrary to many expectations, New Zealand First decided to form a coalition with the National Party.
The coalition agreement between National and New Zealand First was signed on 11 December 1996, exactly 10 years to the day since the members of the Royal Commission on the Electoral System signed the letter of transmittal of its report to the Governor-General. Public opinion polling shows a marked public preference for FPP over MMP dating from about that time in 1996 until the 1999 election.
The period from 1996 to the 1999 election marked another stage in New Zealand's transition to MMP. The National-New Zealand First coalition ended in August 1998, but National continued to lead the government to full term with the support of newly-Independent MPs who had been members of New Zealand First. These and other defections from parties provoked considerable public and political resentment. The 1999 election resulted in a change of government, with the new coalition between Labour and the Alliance being formed shortly after the election. There were signs of increased public support for MMP during the 1999 election campaign.
Effects of MMP
How has MMP fared against the Royal Commission's expectations?
- All parties that crossed the threshold in 1996 (6 parties) and 1999 (7 parties) received a share of seats in close proportion to their shares of Party Votes.
- The Parliaments elected in 1996 and 1999 have higher proportions of MPs who are women (29.2% and 30.8% respectively; 1993 21.2%), Mäori (13.3% in 1996 and 1999; 7.1% in 1993), Pacific Islanders (2.5% in 1996 and 1999; 1% in 1993) and Asian (0.8% in 1996 and 1999; none in 1993). It is notable that most of these improvements in representation have come about through party lists.
- The proportions of Party Votes which did not count towards seats in Parliament were 7.5% in 1996 and 6% in 1999 (substantially lower than in 1993 estimated at 55%).
- Voter turnout in 1996 was 3% higher than in 1993, but declined by 3.4% in 1999.
- Parliament has been more assertive over the Executive since 1996, particularly through select committees.
However there are some areas of public concern. First, some people doubt whether New Zealand's multi-party governments since the first MMP election in 1996 have been as effective as the single-party governments it had under FPP since the mid-1930s. It is generally recognised that whether MMP remains will depend on whether the minority coalition government formed after the 1999 general election is regarded as successful.
Second, some people say they are disappointed that MMP has not brought a new era of political consensus. That was never a realistic possibility since it overlooks the basic realities of democratic politics. MMP has certainly brought more political co-operation and dialogue between parties, but it has also brought more disagreements out into the open.
Third, the change to MMP has meant increase in the total number of MPs from 99 to 120. A non-binding citizens initiated referendum held with the 1999 general election on whether the number of MPs should be reduced to 99 was supported by 82% to voters. The result of this referendum has now been referred to the special parliamentary committee which has been established to review MMP.
Lessons from the change to MMP
So what have been the lessons of New Zealand's change from FPP to MMP?
The first lesson is that some of the important effects of a change to a different type of voting system take longer to emerge than many expected and longer than was the case with the only other major constitutional change in New Zealand in recent years the abolition of the appointed upper house in 1950. Adaptation to a new set of fundamental rules cannot be swift when many of the key actors have had long history and experience under the previous system and hence find it difficult to adjust quickly to the new rules and their incentives. For example, some parties' campaigning in 1996 was still focussed on winning particular electorate seats rather than on winning Party Votes across the whole country. In general, however, parties' campaigning in 1999 was much more focussed on winning Party Votes.
Second, New Zealand's basic constitutional arrangements have been shown to be extremely flexible. Of course, a good deal of planning was done prior to the first MMP election, and the changes in the composition of government beforehand allowed some of those plans to be tested. On the whole, the processes of government and of Parliament have evolved to cope with the changed environment without themselves requiring fundamental change.
The third lesson is perhaps the most important. The change to MMP has revealed that there is a general lack of public understanding among New Zealanders of their parliamentary democracy. For the first time in recent years, there were systematic attempts to promote public understanding among New Zealanders about their system of government. Those attempts revealed that New Zealand is similar to other democratic countries in having significant gaps in public knowledge about electoral and parliamentary matters (particularly among some groups in the population) and that many New Zealanders do not really understand the principles and values of democracy New Zealand-style. It seems too that there were similar levels of ignorance of the previous FPP system. Perhaps because FPP seemed so simple, New Zealand had not felt the need for many years to educate and inform people about constitutional and electoral matters in any kind of systematic and long-term way.
Conclusion
So how were integrity issues addressed in New Zealand's change to MMP?
- concerns about the fairness of the previous voting system prompted a re-examination of that system;
- that re-examination was carried out by an authoritative non-partisan body which recommended a referendum should be held on a change to the voting system;
- although the voting system is an intensely political matter affecting the vital interests of political parties, a cross-party agreement emerged that the voters should decide the issue;
- there was a two-stage referendum process, and members of the public were able to make submissions to a parliamentary committee on the Bills relating to each stage;
- an independent publicly-funded programme of public education was carried out at each referendum and at the first MMP election;
- the new electoral law approved by voters requires a parliamentary review of the new system after two general elections, including consideration of whether any further referendum should be held.
The way in which New Zealand changed to MMP is not a blueprint for the ways other countries should embark on a similar change. The history, traditions and politics of New Zealand's parliamentary democracy heavily influenced that change, just as other countries' histories, traditions and politics influence the ways they go about making significant constitutional changes. Nonetheless, it is possible that New Zealand's example might offer some useful lessons for other countries which might contemplate changing their voting system.
