Double Trouble: The complex arithmetic of NZ voting

Dr Oliver Hartwich
The Australian
15 September, 2023

On 14 October, the same day Australia will vote on the Voice referendum, New Zealanders will also go to the polls, to elect a new Parliament.  

But where the choice in Australia is between a relatively straightforward ‘Yes’ or ‘No’, things are not quite as clear-cut in New Zealand.  

That is because of the electoral system New Zealand uses to elect its House of Representatives.  

Known by the acronym MMP, which stands for Mixed Member Proportional, it starts with the fact that each Kiwi does not have one vote but two.  

But making the right choice is much harder than simply making two ticks. And bizarrely, in extreme cases, you can even effectively hurt a party by voting for one of its candidates.  

Confused enough? Well, then let’s dive into the wonderfully strange world of New Zealand’s electoral system. We shall deal with the history and theory first and then explain what this might mean in the forthcoming election. 

Though a quintessential part of New Zealand’s political landscape, MMP is not a homegrown invention. In fact, it was first used in Germany after World War II.  

To build a democratic future after Nazism, the Germans aimed to combine the best of both electoral worlds: local representation and proportional representation.  

Or rather, the Germans just could not agree on one system. And so, they designed a compromise, combining the two. That was the birth of MMP. Or, as they call it there, personalisiertes Verhältniswahlrecht. It is just as straightforward as it sounds. 

As a native German trained in Hegelian dialectics, I can only apologise to the New Zealander I have become for exporting this confused electoral system to my new home.  

Of all the wonderful things New Zealand could have imported from Germany, why did it have to be our electoral system? Why not airtight windows, underfloor heating or triple glazing? Or Aldi. 

But I digress. 

As I have said, under MMP, voters cast two votes: one for their local MP and another for their preferred party. Electorate candidates get elected by winning a first-past-the-post contest in their local constituencies. Then, other candidates are admitted to parliament from each party’s ‘list’, so that the proportion of seats each party controls reflects their proportion of the party vote.   

So, crucially, each party’s total number of MPs is determined by its share of the party vote. 

Having explained that much, the ordinary voter is already confused. And indeed, even in Germany, where this system has been in operation for more than seven decades, many voters remain unclear about the meaning of their two votes. 

But wait! It gets even more complicated. 

In the game of MMP, parties must secure five percent of the party vote to enter Parliament. But if it wins even a single electorate, it gets a share of seats based on its party vote. 

So, a party could bring in list MPs even if it gets only two percent of the vote, provided it wins at least one constituency. This practice is sometimes called ‘coat-tailing’. 

Just when you thought it could not get any weirder, let me introduce the issue of ‘overhang seats.’ If a party wins more electorate seats than its share of the party vote warrants, these extra seats create an ‘overhang.’ This peculiarity expands the size of Parliament. 

For example, if Party X wins 10 constituencies, but its party vote would only allot it eight seats in Parliament, it nonetheless keeps the additional two seats. After all, we could hardly kick out an MP who has just won a local election. Parliament simply grows by those two extra members, even though that disturbs the intended proportionality. Bad luck. Also, bad luck for the taxpayer. 

Now that we have mastered the theory of MMP, let us look into the 2023 New Zealand election and its potential quirks. 

Exhibit A: the Māori Party. In New Zealand, those identifying as Māori can opt to vote either on the general roll, or for one of seven Māori electorate seats.  

There is a chance in this election that Te Pati Māori (the Māori Party) might win all of them, even though it only looks likely to gain about 3 percent of the total party vote. If they managed this feat, they would create a classic ‘overhang’ of 3 or 4 MPs. It would strengthen Parliament’s left bloc. 

However, what polls indicate may happen with  Labour’s vote is even more intriguing. 

Labour finds itself in a precarious position in the upcoming elections. One recent poll indicated only 26 percent support, a bit more than half what it won in the 2020 election.  

What would this mean in the esoteric arithmetic of MMP? Well, it raises the possibility of ‘Labour List MP Extinction.’  

Three years ago, Labour won 46 of 72 electorate seats. It then received an extra 19 seats from the Labour list – so 65 MPs altogether (in a Parliament of 120). 

Now, on 26 percent (and, for simplicity’s sake, assuming no wasted votes), Labour would have 31 seats. If Labour lost 15 of their 46 local constituencies, they would have precisely those 31 seats. Crucially, not a single Labour MP would enter Parliament via the party’s list. 

If Labour’s vote share dipped even further, say, to 22 percent, Labour could even produce an ‘overhang’, provided they hang onto enough constituencies. 

Labour’s weak polling opens up some strange options for tactical voting. That is, if voters understand what bizarre outcomes their votes could bring about. 

The Wellington seat of Ōhāriu serves as an illuminating case study.  

The incumbent, Labour’s Greg O’Connor, is conspicuously absent from the Labour party list. He is probably not too popular with the party’s leadership.  

O’Connor must, therefore, win his seat again to stay in Parliament. In 2020, he won 52 percent of the vote with a 12,000 vote lead. 

But O’Connor’s rival, Nicola Willis of the National Party, is second on her party’s list and therefore certain to enter Parliament because National is far ahead in the polls. 

Now, the constituents of Ōhāriu find themselves in a tactical quagmire. If they elect Willis, they will have a National MP representing them. If they elect O’Connor, they will get both. 

It gets weirder. O’Connor is obviously not the most influential person in the current Labour line-up. But by voting for him, the Ōhāriu electorate might deprive Labour of a more senior MP from its party list.  

So, if O’Connor wins, it could potentially lead to the ousting of a higher-ranking, senior Labour list MP. The candidates at risk include Finance Minister Grant Robertson, Defence Minister Andrew Little, and Attorney-General David Parker, depending on Labour’s share of the party vote. That is because Robertson, Little and Parker are running only on the Labour list. They probably never expected Labour’s vote to fall this low. 

For a National-leaning voter in Ōhāriu wanting to weaken Labour strategically, voting for O’Connor might be a tempting option. It would help to oust the current Labour leadership if National wins the election. 

But such gaming turns the act of voting into an exercise more resembling three-dimensional chess than straightforward democracy. 

Perhaps MMP’s saving grace is that most normal voters do not understand how it works. Because if they did, they could create a whole lot of chaos and confusion by splitting their votes strategically. 

In any case, as you vote ‘Yes’ or ‘No’ on the Voice, remember that your choice is a lot easier than having to vote in New Zealand’s general election. 

Stay in the loop: Subscribe to updates