Since the election and its coin-flip result, there's been a lot of angst about overhangs, and how to "fix" MMP to stop them happening. I have a rather radical suggestion: don't.
As Robbie Nicol points out in this amusing video, overhangs are a core feature of MMP. They exist because we have electorates. We have electorates because we value local representation for local people. But as long as we have them, we face the risk that a party will win "too many" for its party vote share. That happened in 2005, 2008, and 2011(with TPM); in 2014 (with Peter Dunne); in 2023 (TPM again); and this year (National and TPM).
Up until this year, they didn't really matter - the overhang didn't affect the results, didn't make coalition maths more difficult, and didn't really grant anyone any more real power. This time, they will. Whichever bloc eventually wins power will do so on the back of a substantial overhang. Those overhangs do effectively cancel each other out, and perversely have led to a near-proportional preliminary result (the right bloc's 51% of the seats is only slightly larger than its 49.9% combined vote share). That happy coincidence might not hold in the final results, and in the unlikely event that a bloc take power with more seats but a lower vote share than its rival, then its 1978 all over again.
Meanwhile, there have been calls over the years for the overhang to be "abolished". What this means in practice is "parliament doesn't get bigger". The overhang victors still get their seats, but they come at the cost of everybody else: parliament is effectively proportionalised over (120-overhang) seats; there are fewer list seats, and less proportionality. Except there's a twist, because both the 2012 and 2023 reviews defined an "overhang" as "electorates won by a party which wins less than the threshold". Its not an overhang if its a big party doing it! Which again shows the complete failure of imagination by our electoral "experts" in these reviews.
(In hard numbers: in 2023, TPM won 6 seats, but 3% of the vote. So Parliament would have been proportionalised over 114 seats, giving final numbers of National 47, Labour 34, Green 14, ACT 11, NZF 8, TPM 6 in a 120 seat house. This year's preliminary results would see parliament proportionalised over 115 seats, but National would still have 42 electorates, giving National 42, Labour 34, Green 16, ACT 13, NZF 11, TOP 7, TPM 5, a 123-seat house, and a clear majority to the regime. If the rules were somehow applied fairly, through some sort of countback, ACT, the Greens, and Labour would each lose a further seat, and it would be even worse).
In both cases, "abolition" would in fact hand disproportionate power to overhangs, and set very strong incentives encouraging them. Its not any sort of solution at all.
So why the focus in the reviews on "abolition"? Because both reviews had made a prior commitment to abolishing the "coat-tail rule" - awarding seats to any party which wins an electorate, even if it doesn't meet the threshold. In reality this helps preserve proportionality, but the reviews (and the major parties) were fixated on getting rid of it. The ostensible reason was "unfairness" - and it is unfair that some small parties get their full allocation of seats while others do not, based on a single electorate victory - but rather than resolving this by giving everyone what they were entitled to (the fair solution), the big parties wanted to give no-one what they were entitled to. As for why, there are a bunch of reasons, selfish and petty and political, which basically boil down to "wanting to eliminate National's coalition partners" and "wanting to discourage small parties from running in electorates at all". Those reasons stink, the outcome stinks, and we shouldn't accept any of that bullshit.
Anyway, if you abolish coat-tailing, every electorate won by a minor party becomes an overhang, and parliament gets larger. The reviews thought this was an unacceptable outcome of their intervention, and so "abolishing" overhangs became the solution. But as the 2023 review notes (p. 122),
there would be little point in abolishing overhangs if the one-electorate seat threshold remained.So, maybe we could just... not do that then?
Of course, we still have a problem - a 131 seat parliament, and elections hinging on contests in marginal seats, as they did under FPP. How do we fix that, within the constraints of the MMP system (proportionality, electorates, and Māori representation?) In the words of Mark Thomas, “many New Zealanders might find [it] difficult to accept” a German-style scheme of stripping "electorates" (really semi-geographically-allocated list placements) from parties who win too many, and it would be seen as a direct attack on Māori representation. So I think all we're left with is to increase the size of the House to ensure a 50:50 ratio of list to electorate seats. This would reduce the chance of large party overhangs and mitigate the impact of small party ones (because an extra seat or two matters less in a 140-seat parliament than a 120-seat one). It might not be enough to deal with major-party weakness, but it would be a start.
And of course, we should reduce or eliminate the undemocratic 5% threshold while we're at it.
