MMP is better for New Zealand than every other electoral system. Here’s why.

warning: longer than usual post
The upcoming referendum on the electoral system next year will end up being a Mixed Member Proportional (MMP) walkover, or a choice between two electoral systems: MMP and either First Past the Post (FPP), or Supplementary Member (SM). Either way MMP should remain and then be modified.

Incidentally, the ‘first past the post’ label is completely misleading because there is no fixed winning post. The system was not designed for national elections, and is unlikely to enhance democratic accountability when a significant proportion of the population wishes to vote for parties other than National and Labour. Under FPP, what you need to win a local seat is just a ‘plurality’, namely more votes than anyone else. So the more parties compete in each seat, the lower the winning ‘post’ gets, and the greater the likelihood that most of the the voting public will vote for a party other than the governing party. Consequently political scientists call this system ‘plurality rule’, a much more accurate label.

MMP is proportional in that proportionality extends to all elected members – both on the list and through constituents. The SM electoral system - a bastardised FPP system which was comprehensively rejected in the 1992 referendum - has all of the claimed disadvantages of MMP, such as the status of list MPs, plus the disadvantages of FPP, including parties getting most of the seats with a minority of the votes. SM allows for a relatively small number of list MPs which “top up” the electorate MPs.

The big difference between the SM and MMP lists is that, under MMP the list is used to offset the distortions to proportionality that are inherent in the FPP electorate system, as the party vote only determines the composition of the list seats. Under MMP the party vote determines the composition of parliament. SM, however, uses the list to reinforce those distortions, and this list gives token representation to minor parties. List MPs are way outnumbered by electorate MPs, with list MPs only providing the proportionality. This means that a smaller percentage of list MPs would lead to decreased proportionality than is currently the case, and single party government will be the norm, as it is under FPP.

The advantage of a mixed-member proportional system arises because voters can indicate a preference for a candidate without supporting that candidate's party. A good candidate in an unpopular party has a stronger chance of election.

As an electoral system, SM is closer to FPP than MMP. FPP will only survive as a functioning electoral system if a high majority of people in the electorate support the two main parties, and the voters for these parties have sufficient confidence in their party to form a single party government. In most cases since 1996, SM would have produced the same government outcome as FPP, based on current voting, although minor parties would have had fewer MPs. In 1996 SM, under a 90/30 electorate/list top-up split, as proposed by the 1986 Royal Commission on the Electoral System, would have generated the same government outcome due top NZ First’s capture of the Maori seats. But in 1999, in one of the lowest election turnouts, Labour’s share of the seats was over two percent higher than its share of the party vote and had NZ First and the Greens not crossed the threshold, that share would have been higher.

Supplementary Member is likely to lead to Māori being disproportionately represented in parliament. While it won’t affect how Maori electorate seats are elected, due to split voting it may influence allocation of supplementary seats. Overall proportionality would not exist as the “top-up seats” are likely to be as low as 30 percent of total parliamentary seats. Alternative Vote is likely to rely on second choice preferences, as is STV; FPP is majoritarian in the extreme, is disproportionate, and is likely to “waste” votes. Perhaps this was among the reasons why the Royal Commission on the Electoral System, in its 1996 report Towards a Better Democracy, recommended that MMP was ‘to be preferred to all other systems’. FPP would lead to winners being over represented and losers under represented. This is particularly the case in the Maori seats, where, for example, in 1990, Labour won all the electorates with 65.4% of the vote. More than one in five Maori electorate voters voted for Mana Motuhake, which won no seats. If voters behaved the same way, the party would not be under represented as its share of the nationwide vote, nor would it have won a supplementary seat under SM either, as its overall vote was only 0.6% ( which is also under the MMP 5% threshold).

Single Transferable Vote (STV), in single member districts, as we have in New Zealand, is better than FPP, but not as proportional as MMP. Under this system candidates are ranked and the candidate with the lowest number of votes drops out and votes are reallocated among the remaining candidates and so on until a candidate gets the majority of votes. It is unclear whether this would lead to single party or coalition governments, but it would require a national top up to guarantee full proportionality between the larger parties.This system was rejected by the Royal Commission in 1986.

MMP is the better bet provided the proportion of list seats does not go below 40% - i.e. 48 seats in a house of 120. Currently under MMP it is 70/50.When discussing the choices between electoral systems, the discussion should not be which of the electoral systems - FPP, STV, MMP, SM etc – is preferable. Rather it should be whether one supports single party government or multi-party government – or whether one considers that the party with the most votes should get the most seats. Most supporters of single party government will probably prefer FPP. But when they realise that single party governments are less likely to do what the people want and are less likely to keep their promises, FPP support starts to diminish among those that value democratic ideals. Furthermore, as FPP was never designed for national elections it cannot guarantee that the party with the most votes will get the most seats. Indeed, in 1978 and 1981, the winning party, National, got more than half of the seats with less than half the vote. In 1981, National governed with just 38.8% of the vote. Had these elections been held under a Supplementary Member system (SM), with that vote, this would have still been the case.

There has been some discussion over recent years regarding the 1990 election. That year National got 67 seats for just 48% of the vote, and Labour got just 29 seats, as New Labour got the other one. That was more disproportionate than Labour’s 1972 result. That year Labour also won 48% of the vote but got 63% of the seats and Social Credit got its lowest ever election result (6.7%). Just three minor parties have since got a higher share of the party vote, all whom had leaders with parliamentary experience. Similarly, in 1951, the UK Labour Party lost the election despite outpolling the Conservatives and winning a majority of the popular vote. In 1974, Labour won the election despite the Conservatives gaining most of the votes. So FPP fails to ensure that the party with the most votes gets the most seats. FPP is not a fair system when 21% of the electorate can vote for one party and be represented by fewer than 2% of the members of parliament, as did Social Credit in 1981.

In fact, of four electoral systems - FPP, STV, SM, and Preferential Voting (PV), not one guarantees that the party with the most votes gets the most seats because some seats are allocated to parties who have not won them based on their share of the vote. The non-proportional preferential voting system maintains the basics of the Westminster system intact, but allows second-place votes to count towards determining the winner if no party wins more than fifty percent of the vote. FFP is likely to foster majoritarian extremes. Only MMP guarantees that the party with the most votes gets the most seats, and without these extremes.

While MMP makes it easier for a greater number of parties to enter parliament, it is not without problems. In the 2008 election, NZ First got more votes than Act, United Future and the Progressives combined, yet NZ First was out of Parliament because it failed to have its threshold effectively lowered by securing a constituency seat to gain representation. The others remained in Parliament purely because they did. All four parties- along with the Māori Party, polled less than the 5% threshold. Had the MMP threshold been 4%, NZ First would have relied on the threshold for their parliamentary existence, as would have the Christian Coalition when it gained 4.3% in 1996.

Open lists; or either lowering or removal of the threshold – which would prevent the situation where Act got fewer party votes than NZ First, who got no seats - is preferable to introducing a new electoral system such as SM. Also, as the population increases, this will lead to a greater number of electorate seats – and a fewer number of list seats, if the number of South Island electorate seats continues to be pegged at 16. Currently there are 52 list seats ( down from 55 in 1996) and 70 electorates (up from 65 in 1996) – including the seven Māori electorates which currently provide a two seat overhang.

But the fact remains: To the average voter, MMP, STV, FPP, PV and SM are not electoral systems, they are merely letters without much meaning. Prime Minister John Key has said that any change to the electoral system requires a major advertising campaign. "We can't ask people to make constitutional changes without understanding what the options are."

Politics is often seen as two sections: the left and the right. Act will go with National and the Greens with Labour. Many other PR systems are like this. New Zealand is different as we have centre parties that can go either way - United Future and even the Maori Party. So the left/right splits are not so relevant, as we have seen with New Zealand First when they held the balance of power in 1996, and chose National to govern when most thought Labour would have been picked.

There are a lot of matters on our electoral system options that need to be understood. The task of making sure any possible changes are understood will fall on the Electoral Commission.

Why National supports Supplementary Member and the Greens oppose it

Heres how the New Zealand House of Representatives would have looked under the Supplementary Member electoral system, with the current seat numbers in brackets for comparison.


New Zealand National Party - 65 seats (58)
New Zealand Labour Party - 40 seats (43)
Māori Party -6 seats (5)
Green Party - 4 seats (9)
Act New Zealand - 3 seats (5)
Jim Anderton's Progressive - 1 seat (1)
United Future New Zealand - 1 seat (1)

National would have governed alone, may as well have had FPP, with no bridle on executive power.

If the five percent threshold was to be removed:
Under SM, only National and Labour would be affected, both losing a seat to New Zealand First. Under MMP, National would have 3 fewer seats, Labour 2 , the Greens 1 , with Act getting an extra seat. Three additional parties will enter parliament: New Zealand First with 5 seats, with Bill and Ben and the Kiwi Parties gaining one seat each.

Had the 2008 election been under SM, Act would have still have got more seats than NZ First- even without a threshold - and with 10,000 fewer votes.

So, does anyone else think that, in the context of discussing electoral systems, a reduction of the MMP threshold would be preferable to the adoption of the SM system, with or without a threshold.

More on the MMP referenda

It will be interesting to see the next stage of Cabinets plans for the two stage referenda on the electoral system. The first stage, at the 2011 election, asks whether we should change our electoral system, the next stage provides options for change should change be desired. How many options? We don't know. And nothing will change until at least 2017.

Now I think I have this right. Take this scenario, which appears to be the way Cabinet is heading. Let's say the first referendum gets a 45/55 split with 55 percent supporting a change of electoral system.This was the split that led to the introduction of MMP in 1986. Significantly the 45 percent minority would prefer the retention of MMP than change to another electoral system. But lets say the second stage has a run off between three electoral systems - say if 35 percent vote STV, 25 percent vote SM, 40 percent vote FPP, then FPP would be the electoral system, right?

Wrong. MMP would. Despite more than 50 percent initially voting for a change from MMP. A referendum needs more than a 50 percent vote for a specified electoral system to be changed.

In the first referendum 45% wanted to keep MMP which is higher than the numbers that voted for the highest polling electoral system in the second stage. Not one voting block got over 50%, so that will be a complete waste of money. Even if MMP was one of the options in the latter referendum, as Graeme Edgeler notes, MMP will be the declared electoral system even if it got the smallest proportion of the vote, if no other system got over over 50 percent of the vote.

That would also be the wrong way to go about this*. It just won't work. MMP supporter David Farrar suggests the first referendum be done in two parts. The first part will ask if we should change our electoral system, the second part which system we would change it to should changed be desired, with the most popular option being run off in the second referendum at the following election.

So, on the above figures, if the first 2011 referendum combined the above two stages,in 2014 there could be a be a referendum between several electoral systems. Unless that referendum was conducted under STV rules*, on the above figures this will be either MMP v FPP(the current system and top alternative) or FPP v STV(the two favoured alternatives). If it was to be the latter, 45 percent of the population will be unhappy as their preferred option was MMP. If it was MMP v FPP, it will technically be a rehash of the first referendum, the only difference being which of the 55 percent that wanted change would support FPP, assuming all the 45 percent who voted against change still support MMP. Some of the 55 percent may have wanted a change to STV. If enough that did voted for MMP because they wanted at least some proportionality, the current system would be retained.

So there has to be two questions at the first referenda. In the second referenda, an electoral system other than MMP must have 50 percent of the vote for an electoral system change - and the more options there are the less likely that will happen. But if the Government follows this format, some people are not going to be able to vote for their favoured electoral system after being asked what their preference was in the second stage of the first referenda.

*unless the second referendum was conducted by ranking ala STV rules -which it looks like it can't be unless they first pass a regulation.