Dialogue: Just make a few changes to get MMP right

By PETER AIMER and JACK VOWLES*

When 54 per cent of voters in a 1993 referendum chose MMP ahead of first-past-the-post as the future voting system, they brought to life the Electoral Act. One provision of the act is that Parliament review the working of the new system by June 2002.

A select committee is doing this. In light of public submissions and its own deliberation, it may recommend anything from no change to big changes to the voting system, including a further referendum.

Meanwhile, Opposition leader Jenny Shipley is seeking a mandatory referendum on MMP versus alternatives before the next election. This is a puzzling move since it begs the outcome of the review.

Such a referendum for or against MMP in general could be justified only if there were evidence there has been a comprehensive loss of support for it, or that it is seen to have failed. Does the evidence support the need for such a referendum?

Since 1990, data from the New Zealand election study programme show that when electoral systems in general are on offer, the contest is still mainly between two horses: MMP and the old system, FPP. Support for other alternatives is small indeed. Any referendum pitting MMP against alternatives would almost certainly lead to a rerun of the MMP-FPP race.

Our data show that overall support for MMP did decline after the 1996 election, mainly reflecting disappointment among Labour and Alliance supporters. After the surprise formation of the National-NZ First Coalition, support for MMP among Labour voters dropped from 59 per cent at the time of the 1996 election to 33 per cent by July 1998, and that of Alliance supporters from 76 to 51 per cent.

By contrast, support for MMP among National voters strengthened from 24 per cent in 1996 to 31 per cent in July 1998. It reached 38 per cent during the last week of the 1999 campaign, only to plunge back to 23 per cent immediately after the election, which saw their party lose.

Although support for MMP has softened since 1996, strong opposition to it has also declined. In the election study surveys, respondents who agreed that we should get rid of MMP have fallen from 40 per cent in July 1998 to 36 per cent in the last week of the 1999 election campaign to 29 per cent immediately after the election. An NBR-Compaq poll in May found only 23 per cent who would scrap MMP.

Wisely, considering our brief experience of the new system, by far the largest body of electors - from 49 to 51 per cent since July 1998 - say it is too soon to tell whether MMP has been a disaster or a success.

If public attitudes are a guide, a referendum on whether we should retain MMP could not be justified, bearing in mind also that opinion is often associated with short-term partisan assessments of advantage.

The foundations of support for MMP are evident in the more fundamental issues of political representation, and the composition and performance of governments.

On these matters, there is widespread acceptance of the kind of outcomes that MMP rather than FPP elections can be expected to produce. The principle of proportional representation is supported by more than those who oppose it. In 1999, 60 per cent of respondents agreed that a party's seats should be equal to its percentage of the party vote, and only 14 per cent disagreed.

When proportional representation of parties in Parliament is linked to the need to form a government, opinion is more evenly divided, but by margins ranging from 9 per cent to 17 per cent. Thus, since July 1998, respondents have consistently indicated that proportionality was more important than the typical FPP election outcome of one party gaining more than half the seats in Parliament so it can govern on its own.

Although coalition government is a novelty, and despite the tribulations of the National-NZ First Administration, public preferences for coalition or one-party governments remain nearly evenly divided, with the trend being towards greater acceptance of coalitions.

In July 1998, 58 per cent preferred single-party government and 40 per cent coalition government. On the eve of the 1999 election, when both major parties were facing the prospect of governing with the help of an acceptable coalition partner, the balance was reversed, and a majority of 52 per cent over 46 per cent preferred coalitions. Immediately after the election, there was more uncertainty again, though slightly more still preferred coalition (47 per cent) to single-party government (44 per cent).

When questioned on the properties of single-party and coalition governments, respondents have consistently and accurately judged single-party governments to be more stable. But they have ranked them more evenly on their ability to make tough decisions and have assessed coalition governments to be clearly superior on two counts - keeping promises and doing what the people want.

The data thus indicate a public which is sometimes divided or uncertain in assessing electoral issues, but one in which a preference for outcomes consistent with the proportional representation of MMP continues to have an edge.

Some of the uncertainty and division might be reduced by introducing changes to MMP without reducing the central element of proportionality.

There is still strong prejudice against list MPs. This is partly because of assumptions about the local recruitment and role of MPs embedded in the tradition of single-member territorial representation. It is also partly because of party-hopping by list MPs.

Many voters resent the lack of effective choice of who is elected from the party lists. Fifty-seven per cent of survey respondents in 1999 agreed that voters, not parties, should decide which of the candidates on the party list get the seats the party has won. Only 15 per cent disagreed.

The select committee would, therefore, be justified in investigating ways of opening the party lists to voter choice. For such a choice to be meaningful, a system of regional lists may be necessary. Such a refinement would also require consideration of means of maintaining the proportionality of MMP.

Submissions have queried whether the winning of an electorate should continue to provide a means for a party to cross the threshold for allocation of list seats if that party fails to gain 5 per cent of the party vote.

The existence of this electorate threshold increases the chances of small parties gaining seats. Polls indicating that a small party is likely to win a single electorate have encouraged more people to cast party votes for that party, for Act in 1996 and the Greens last year.

Removing the threshold would reduce the chances of small parties, and thus increase the odds of majority government and greater stability. Modification of the threshold would be a significant change to MMP - and, in fact, a valid issue for a referendum. It would not be useful to throw the whole electoral system back into the referendum melting pot, and replay the debate of the early 1990s. A more sensible strategy, based on public opinion, would be to work towards changes in the details of the new voting system.

* Peter Aimer is an honorary research fellow in the department of political studies at the University of Auckland; Jack Vowles is associate professor in the department of political science at the University of Waikato.