OPINION: Michael Swanson
Every three years, New Zealanders fill in two ticks on their ballot paper. One is a party vote, which decides how many seats each party gets in the House of Representatives. The other is an electorate vote, which decides who represents your local patch. As the Electoral Commission explains, the party vote is what largely determines how many total seats each party gets, while the electorate vote is about choosing who represents the area you live in. It is a system designed to give voters the best of both worlds: a proportional Parliament and a local face to complain to.
Thirty years after MMP was adopted, it is worth asking whether that second vote is doing what it was meant to do. In practice, electorate MPs spend most of their time in Parliament voting not as representatives of their electorate, but as instructed by their party whip.
The mechanics of the problem
Under MMP, the overwhelming majority of votes in the House are party votes rather than personal ones. As one academic analysis of conscience voting in the New Zealand Parliament puts it, the vast majority of votes in the House are whipped, with MPs voting along party lines rather than as individuals. In a party vote, an MP does not walk to a lobby and cast a ballot in person. Instead, a whip (or a member acting as one) casts the vote on behalf of the whole party, and any MP simply present in the building at the time is automatically counted in that total, whether they agree with the call or not.
The exception is the so called “conscience” or personal vote, reserved for a handful of deeply contested moral issues such as euthanasia, prostitution law reform, or marriage equality. Parliament’s own fact sheet on the practice notes that these are the rare occasions when MPs are freed from the expectation of following a collective party position; the rest of the time, the default is to vote as a bloc with the rest of the caucus. Even here, academic Dominic O’Sullivan has argued in The Conversation that conscience votes expose a deeper flaw: an electorate MP might consult their community or even run an informal survey before a conscience vote, but there is no guarantee of this, since that same conscience can just as easily be handed back to the party whip on every other question that comes before the House.
So for the vast bulk of legislation, the thing that supposedly distinguishes an electorate MP from a list MP, namely a personal mandate from a specific community, simply does not operate. A National MP in a marginal Auckland seat and a National list MP ranked fortieth on the party’s list will vote identically on almost everything that matters: budgets, confidence motions, and government bills. The electorate MP’s local mandate is, for practical purposes, decorative.
This isn’t a flaw unique to malice or poor character. It’s how any disciplined parliamentary party system with proportional seat allocation has to work. A government needs to be able to guarantee its numbers to pass a budget or survive a confidence vote, and the New Zealand Initiative has pointed out just how tightly the arithmetic of MMP already binds parties together, given the complexities of thresholds, coat-tailing, and overhang seats that can expand Parliament beyond its notional 120 seats. Layering local independence on top of that arithmetic was always going to be more symbolic than substantive.
What electorate MPs still offer
To be fair, electorate MPs are not entirely ornamental. They run electorate offices, take up individual casework such as immigration or ACC disputes, and can occasionally cross the floor or extract concessions behind closed doors, particularly if they hold a marginal seat a party cannot afford to lose.
Losing the electorate tier entirely would mean losing that single, easily identifiable point of local contact. That is a genuine cost of any reform, not a minor one.
The Scottish alternative
One model worth examining is the Additional Member System used to elect the Scottish Parliament. Scotland is divided into 73 single-member constituencies, but also into 8 larger electoral regions, each of which returns several additional MSPs elected proportionally from party lists rather than through a single-winner local contest. As the Scottish Parliament’s own guidance explains, each of those 8 regions returns 7 additional MSPs, with regional votes counted using a formula so that the total number of seats a party wins across a region ends up reflecting its overall share of the regional vote. The Constitution Society describes these as effectively “top-up” seats, since their whole purpose is to correct for parties that ended up under-represented once the single-winner constituency seats were tallied.
Applying something like this to New Zealand would mean abandoning the single-member electorate altogether in favour of, say, eight to twelve broad regions, each returning a handful of MPs allocated proportionally to party vote share within that region. Voters would still get a distinctly regional slate of MPs, arguably several who could plausibly claim to speak for that area, rather than one who is bound to vote with their party regardless. It would also reduce the electoral distortions unique to single-member seats in a proportional system, including overhangs and the “coat-tailing” loophole that lets a party bring in list MPs off a low party-vote share simply by winning one seat.
It would not be a costless change. Scotland’s own experience shows that dividing votes between two tiers creates its own gaming risks: the Constitution Society and other commentators have noted concerns about parties splitting support strategically between constituency and list votes to maximise seats. A regionalised New Zealand would need careful thought about region size, list transparency, and how casework and constituency service would be shared among several regional MPs rather than one clearly accountable local figure.
A question worth asking
Would we even miss Electorate MPs if we instead moved to a proportionally allocated group of regional MPs? You’d still have MPs for that region, but you’d only have a single vote, the party vote, which would then decide the regional allocation of seats to various parties. It’s just an idea, and it absolutely, 100%, has flaws, but every system does. The overall issue I want to highlight is that Electorate MPs don’t advocate as directly and effectively in Wellington as some may believe. And in many electorates they aren’t really held to account each election at all as many seats have barely changed party in their entire history.
None of this means electorate MPs are pointless. But if the defining justification for the seat, namely independent local judgement expressed through votes in the House, rarely survives contact with party discipline, then the case for the current single-member model rests almost entirely on constituency service and a simple, familiar mental model of representation. Whether that is worth preserving over a system that could deliver more proportional, regionally grounded representation is a genuine and open question, and reasonable people looking at the same evidence come down on different sides of it.
