Explaining New Zealand

As expected, Saturday’s New Zealand election delivered a close result (see my preview here). The shape of the new parliament will not be clear until after postal and absentee votes (called “special votes”) have been counted (official results here), and figures incorporating them will not be released until 27 November. Unfortunately, it’s impossible to explain how they might change things without getting quite technical.

For those who don’t want the technical detail, here’s a one-paragraph summary of where we are so far. The incumbent right-of-centre coalition government has 67 seats (unchanged) from 49.1% of the vote (down 2.9%); its left-of-centre opposition has 57 seats (up two) from 43.2% (up 1.6%). The centrist Opportunity Party, which wasn’t in the last parliament, has 6.0% (up 3.7%) and seven seats. But it’s possible that the later counting will change the totals for the two main blocs enough to give Opportunity the balance of power.

Now for the detail. Although New Zealand’s electoral system is called Mixed Member Proportional (MMP), it’s generally a pure proportional system. In the normal situation – or at least what those who devised the system expected to be the norm – only the party list votes count in determining how many seats the parties hold in parliament. The electorate votes only determine the identity of the local MP, not the overall numbers.

The exceptions to that situation, when a party elects more electorate MPs than its total entitlement, creating an “overhang”, were expected to be rare. Of the nine elections held under MMP prior to this one, only two produced an overhang of more than one seat: two in 2008 and three in 2023 (one of which was due to a supplementary election caused by the death of a candidate). So the projected 11 overhang seats this time are off the chart.

The problem is that two parties, National and the Māori Party (TPM), have learnt to game the system. Large numbers of people voted for their candidates at electoral level, but gave their party vote to an allied party (ACT or NZ First for National, Labor or Greens for TPM). National got 29.2% of the party list vote but 41.3% of the electorate vote; TPM had just 0.6% of the party vote but almost six times as much, 3.4%, in the electorate vote. That produced an overhang of seven National seats and four TPM seats, for a parliament of 131 instead of the normal 120.

Since those two parties are on opposite sides, their efforts partly cancel out. But that doesn’t undo the harm done to what is supposed to be a proportional system. National and Labor have almost exactly the same party vote (Labor is just 3,650 votes ahead) but National is winning six more seats, 42 to 36. This is just the sort of thing that supporters of proportional representation criticise non-proportional systems for allowing.

So what about the special votes? In a normal election they are a zero-sum game; if they skew left, as they generally do, they might give Labor an extra seat or two at National’s expense, by increasing its party vote relative to National’s. They might also change the result in one or two electorate seats, but that doesn’t affect the overall numbers, because in a normal election the electorate seats don’t matter for that.

With a large overhang, that all changes. If Labor outperforms National in the special votes, that might increase its seat entitlement by one or two. But it won’t change National’s entitlement, because that’s being determined by the number of electorate seats it’s winning: all it will do is increase the overhang. Starting, as we saw above, at 67-57-7, a gain of two seats for Labor would therefore take it not, as you might imagine, to 65-59-7 (a majority for the opposition plus Opportunity), but only to 67-59-7 (a one-seat majority for the government).

So does that mean the government is safe? Not quite, because National could still lose seats, but only by losing electorate seats. It’s currently leading in two of its electorate seats by margins of less than 500 votes; if the special votes are enough for Labor to overtake them there, then the overhang will be reduced accordingly and National’s total will go down. But Labor’s total won’t go up, because its entitlement is well in excess of its number of electorate seats.

That means that for the opposition to be in a position to bid for Opportunity’s support to form government, the things described in the last two paragraphs both have to happen: it has to pick up two party list seats, and National has to lose two electorate seats. It also has to have TPM hold all its electorate seats – in one of them, Te Tai Tokerau, it leads the Greens by just 436 votes. If it should lose, that seat would disappear from the opposition’s total.

We’ve seen in many places around the world the ways in which electoral systems have been subverted by parties playing tricks (this 2020 example in South Korea is particularly relevant). Now it’s come to our neighborhood. Whoever ends up in government in New Zealand needs to do a repair job on its electoral system to ensure that proportionality is allowed to work as intended.