This piece started life as a general survey of the state of the race as the 2026 campaign officially began. Over the weekend, events decided to rudely transpire. Instead we’re looking into coups present and past.

The tl;dr

• National were still favoured to win the election prior to this latest round of leadership speculation.

• A scarcity of winnable list spots and nervous MPs in swing seats could be a factor in said speculation. As could the rise and ruling-out of the Opportunity Party.

• Challenges to leaders with their head above water are rare, as are public challenges to Prime Ministers.

• Most new leaders (after both jumps and putsches) see a short-term bump. Not all maintain it, but National don’t have the luxury of the long-term.

Like the All Blacks and a coach with a win rate under 80%, sometimes political parties decide the leader just isn’t working out.

It’s not you, it’s us.... While this is not a reflection on your performance, we received many high-quality applicants for the position and, regrettably we have decided…

This is, in some ways, the position National and Christopher Luxon find themselves in ahead of the leadership vote later this morning.

Is it really so bad?

Based on recent public polling, National had pulled ever so slightly ahead of Labour (what private polling says is anyone guess - maybe yours - leak it to me 😉).

The coalition in some form was favoured (albeit far from guaranteed) to return to power in November.

This does force us to ask…

But why?

A scarcity of winnable list places and nervous MPs in swing seats may be a factor. Coalition prospects could be another.

Current polling gives National 38 total seats on the party vote. Opportunity entering Parliament would lower the wasted vote share and take seats off other parties (likely two from National). Add to that the continued strength of NZ First and ACT, and you have National needing to lose five electorates just to avoid an overhang, and six or seven seats (depending on whether Hutt South is one of them) for Nicola Willis and Chris Bishop to return. Turkeys may not pray for Christmas, but Gerry Brownlee and Paul Goldsmith are probably advised to.

The fact that they likely would drop six is little comfort to the backbenchers taking dives to save #2 and #3. The fact that they very well might not is little reassurance to said #2 and #3. What felt far-off and theoretical during April’s round of murmurings is now imminent and real. The list is set, the campaign has started, and a rebound of brand National (or a few good list spots fortuitously opening) might be the only chance for some.

The captain’s “no deal” call on Opportunity is also notable. Current polling trends have them on 5.4%, and our modelling has them at 2:1 odds to make it over 5%. Unilaterally ruling Wong’s party out when they would be necessary for forming a government ~30% of the time (and useful to forming one in many other situations – if only as a counterweight to ACT and NZ First’s bargaining power) has narrowed National’s options, whatever else it did.

Further speculation as to motivations (policy, factionalism, ego, boredom) I will leave to better-connected observers and to the eventual tell-all memoirs. Depending on how the vote today goes, by lunchtime this analysis (the original subject of this newsletter) will be so much fish-and-chip wrapping.

Is it really so bad?

While the above numbers aren’t stunning (they have a distinct second-term vibe), compared to past precedent they’re not exactly coup-inducing. Only one leadership challenge since 1980 has involved a leader with positive figures: Robert Muldoon all the way back in 1980. Even then, the plot fell apart before a vote could be called. Brash’s resignation (under a cloud, but not underwater) and Key’s surprise mic drop both happened above zero, but both were voluntary exits (at least nominally in Brash’s case).

In this context, a narrow lead with a decent chance of a return doesn’t seem so bad. It certainly isn’t enough to warrant a Mike Moore hospital pass or Jacinda Ardern Hail Mary.

Well did it work for those people?

I mean, kind of? Sometimes? Maybe? If the winds are favourable and the auguries propitious?

In the immediate aftermath of a leadership change, almost everyone gets at least a short honeymoon. The main exceptions are Goff and English (on their second goes) both as somewhat anti-climactic follow-up acts, and Cunliffe after his successful ousting of Shearer.

Against these, Luxon’s bluffing off the April rumblings stacks up well at a net 6% gain vs Labour.

Over the medium term, the picture is more mixed.

• Two changes saw an initial bump that faded, ultimately ending in electoral defeat (Shipley and Brash).

• Three essentially failed on their own terms: falling off quickly and never recovering (English #1, Goff #2, Cunliffe).

• Two held for a time before falling away, usually due to subsequent events (English #2 with Jacindamania, and Hipkins with successive Ministerial scandals).

• A few didn’t really do anything (Little, Bridges, Muller/Collins).

• One saw a lift in the short term, but not a big enough one, never making it to an election (Shearer).

• The most successful – and the ones I suspect are foremost in any plotters’ minds – actually worked: a bump that stuck and that led to a win (Key, Ardern, Luxon).

The one successfully-defended plot in the MMP era – Clark seeing off Goff’s 1996 challenge – worked in the long term: a miraculous, face-saving run into the election, followed by eventual victory in 1999.

Overall, these results in the medium term are a wash: seven ahead, seven behind, with the one survivor (Clark) eventually do spectacularly.

Worth noting that for the successful cases, a lot of the time the difference was as much, if not more, down to consolidating the vote among aligned parties rather than just pulling over voters from the other side. Clark, Brash, Ardern, and Luxon had to gather wayward Alliance, NZ First, ACT, or Green support back into the big tent alongside winning converts to it.

With 87 days to go to the election, the long term probably isn’t on anyone’s mind. There isn’t enough runway for a slow take-off. That short-term high is what people are chasing and, for now, it’s all they need.

Whether it hits (and what reaction it provokes from a Labour with their own Chris to bear), we’ll have to see.

Hi.

A much-delayed effortpost, but a comprehensive and hopefully useful one.

It covers:

• the overall outcomes of the 2025 boundary review; and

• a deep dive into the changes in each seat.

This is the sixth and final part of a series on the review. For backgrounders on the review process see issues one, two, three, four and five of this series. The Representation Commission’s own final decision provides a pretty good account of the reasons for its decisions and how it responded to submissions.

A note I’ll repeat a few times: all this is presented based on results from the 2023 general election. It’s a rerun, not a forecast.

This work will be used as an input to forecasts for 2026 which should be coming out over the summer break.

tl;dr version

• On net Labour lose one seat based on 2023 numbers

• Labour loses the abolished seat in Wellington (Ōhāriu/Mana)

• National gains one seat from Labour (Wigram)

• Labour gains one seat from National (Waitākere)

Winners

• Voters in Auckland – Final boundaries leave the Auckland region with 2/3rds of a seat more than its population entitles it to.

• West Auckland Labour – More efficient distribution of Labour voters will make winning all three seats (Glendene, Waitākere, and Henderson) easier.

• Carlos Cheung and Cameron Brewer – Changes to Mt Roskill’s boundaries of make the seat more defendable for the surprise 2023 winner, while Upper Harbour (a shock 2020 Labour win) is now out of Labour’s reach for the foreseeable.

• … and Chrises Luxon and Penk – Changes to Botany see the largest swing in the country and Kaipara ki Mahurangi sees decent National gains (not that either candidate needs it).

• … and Jenny Salesa – On the other side, the incredibly safe Labour incumbent in Ōtāhuhu gets even safer.

• Chlöe Swarbrick – The Greens co-leader sees a fairly green seat get greener.

• Wellington Greens (maybe) – The adjustment between Wellington North and Wellington Bays leaves the seats more evenly balanced in Green-voter terms.

• The 2029 Representation Commission – The obvious eye that the 2025 Commission had on the future should make the 2029 review easier: so long as population forecasts are correct.

Losers

• Wellington – Simply put: Wellington has one less MP. Population changes made this largely inevitable, but the Commission’s decisions exacerbate the impact.

• Voters elsewhere in the North Island – The cost of under-packing Auckland is leaving the rest of island over-packed and underrepresented.

• Greg O’Connor – The seat lost in Wellington (Mana or Ōhāriu, depending how you slice it) is a straight -1 loss for Labour squeezing O’Connor out.

• 2023 National newcomers (Paulo Garcia, Greg Flemming, Rima Nakhle, Tim Costley) – Modestly unfavourable boundary shifts in Waitākere, Maungakiekie, Takanini, and Kapiti respectively will make defending 2023 flips harder.

• Technically also Judith Collins and Suze Redmayne – Both Papakura and Rangitīkei remain deeply blue, but unfavourable shifts still reduce their margins.

• Tangi Utikere and whoever replaces Megan Woods – Palmerston North’s outward expansion into bluer exurbs makes re-election more challenging for Utikere, and the same phenomenon flipping Wigram could have contributed to Woods’ decision to go list-only.

• The Greens’ chances of another Auckland seat – While the headline Labour/National contest in Mt Albert is unchanged from the 2023 nail-biter, the price the Greens pay for the boost in Auckland Central is Ricardo Menéndez March’s chances fading.

• Wellington Greens (maybe) – the flipside of spreading their vote more evenly is it makes losing both easier too.

• Ashurst and Newlands – because of the high-level decisions the Commission made and the constraints it faced, both Ashurst and Newlands find themselves attached to electorates across geographic boundaries.

Overall changes

Recapping

As previewed in Parts 1-4 of this series, the Commission’s basic challenges were:

• Needing to lose a seat on the North Island

• Overpopulation in Auckland’s outer suburbs and Selwyn

• Underpopulation on Auckland’s isthmus and in Wellington

Their ability to make changes was constrained by:

• Respecting existing electorate boundaries

• Physical barriers (mountains, rivers) and connections (roads)

• Communities of interest

What that resulted in is…

The big picture

Overall changes

The changes are among the most significant we’ve seen since MMP was introduced:

• 54 seats have seen changes.

• Only the Waikato, Upper South Island, and Otago are unaltered, along with the four Māori seats north of Taupō.

• Around 270,000 voters (10%) are now in different electorates.

• Population in Auckland has been shifted “inwards” (or boundaries shifted “outwards”) from the north-west and south-east to make up for the seats on the isthmus all being well below quota.

• In the lower North Island, the merger of Mana and Ōhāriu kicks off a cascade of changes as far north as Taranaki.

• In Christchurch, moving Selwyn’s excess population "‘inwards” also causes a chain reaction of changes through the city.

As you can see on the diagram, these changes leave seats in Auckland predominantly under quota (purple), while North Island general seats elsewhere are largely over-quota (orange). This has been driven by two factors:

• The decision to remove a seat at the southern end of the island.

• A clear prioritisation of forecast population growth through 2029.

It’s minor in the scheme of things (and doesn’t affect proportionality given the party vote) but leaves Aucklanders with 0.6 more of an MP than they ‘should’ have and the rest of the island with 0.6 less of one. In theory this reverses out by 2029 when the next review is scheduled, but that’s putting a lot of faith in medium-term population forecasts.

Partisan impact

The chart below shows the left/right swing in candidate margins for each seat. Dashed lines indicate tipping points in four scenarios:

• The 2023 general election

• An “average” election with Labour and National evenly matched

• A “National landslide” (slightly better than 2011)

• A “Labour landslide” (based on 2020)

Two seats cross the 2023 threshold (Wigram and Waitākere) meaning they would have flipped. None cross the average line, but several get closer to it. Seats beyond the landslide lines are likely out of reach in even the best (or worst) scenarios.

A change might be big, in a safe seat it’s not going to matter. The chart below compares the size of the swing to the existing margin to highlight the changes that do matter. The closer to the top-left a seat is, the more significant the change.

Seat-by-seat

Now to go through things in excruciating and unnecessary detail. Seats are sorted roughly north-south of you want to skip to one you care about.

Explanatory notes about the maps below before we get into it:

1. Areas outlined in green have been gained from another electorate, areas outlined in orange have been lost.

2. The red-blue spectrum shows left (Labour, Green, TPM) vs right (National, ACT, NZ First) party vote PVI for each area: how much more left or right leaning the area is compared to the country as a whole.

3. The info boxes show estimates of 2023 party votes for each area that got moved.

What shifts the partisan dial is how the moved population votes relative to the electorate it’s being moved into or out of. As we’ll see, this can mean a boundary change helps one party in both the donor and the recipient seat.

Methodology for estimating the party and candidate vote

Brief note on how the partisan estimates were calculated:

1. Calculate the total number of voters moved (and not moved) based on turnout at an SA2 (neighborhood) level. Given the granularity of the official statistics, this step is an accurate account of how many voters were reallocated (as opposed to total population the Commission uses - including children, the ineligible, and the non-voting).

2. Estimate party vote shares (separately for each of advance, on-day, and specials) for each SA2 based on nearby polling places (interpolated using an inverse distance weighted heat-map). This is an imperfect approach as it assumes neighbourhoods reflect the polling places in/near them, but is as close as we can get.

3. Apply an adjustment so the sum of SA2 votes matches the 2023 electorate vote.

4. Estimate candidate vote shares based on 2023 vote-split data to account for candidate strength and contest type (Labour vs National or three-way contests) and on the assumption rates of vote-splitting doesn’t vary much within electorates.

There are some inevitable inaccuracies in this approach, hence why everything is framed as an estimate. As a cross check I compared the SA2 level totals aggregated up to 2023 electorate boundaries and they’re accurate to within ~1% on average which gives me confidence the estimates are meaningful.

With that out of the way, now to take it from the top…

Northland-North Auckland

Changes North of the Waitematā are all fairly minor, with only one seat (Upper Harbour) seeing a material swing. For the rest, it’s largely blocks of conservative-leaning voters being traded back and forth.

Northland

Over-quota Northland sees an area around Ōakura and Whangaruru with 1,160 votes move into Whangārei to hit weight.

The National/Labour race in 2023 seemed closer than it really was given active campaigns from NZ First’s Shane Jones and former National MP-turned DemocracyNZ leader Matt King. Given the small number of voters moved and their similarity to the overall electorate mix it doesn't do much.

Whangārei

Whangārei was over-target but still within the 5% tolerance, so absorbs 1,160 voters from Northland (allowing Northland’s southern border to remain aligned to the Northland/Auckland regional one). The partisan impact is similarly marginal.

Whangaparāoa

Further south, fast-growing north Auckland seat Whangaparāoa was well over quota. This sees it lose:

• Albany Heights and Pāremoremo with 3,730 voters to East Coast Bays

• Eastern Coatesville and 370 voters to Kaipara ki Mahurangi

Negligible partisan impact given the relatively homogenous right-lean of both the seat as a whole (currently the largest majority in country) and the areas moved.

East Coast Bays

ECB picks up 3,730 voters in Albany Heights and Pāremoremo from Whangaparāoa while shedding:

• Murrays Bay and Windsor Park (1,300) to North Shore

• Rosedale and Schnapper Rock (2,780) to Upper Harbour

The lost areas are relatively more conservative than the electorate as a whole so technically should benefit Labour (+1.1% on net) but with the second largest National majority in the country it’s comically moot.

North Shore

The three inner North Auckland seats were all under-quota so absorb population from the over-populated seats to the north.

And oh wow, another consistently right-leaning seat with a small (1,300) and inconsequential (National +0.5%) boundary tweak that will have no meaningful impact on the results. Ground-breaking.

Northcote

Northcote is what passes for a “swing” seat on the Shore so a change here hypothetically could matter. However, the 2,040 voters around Glenfield that it picks up from Upper Harbour more or less match the seat’s existing mix so - once again - no meaningful impact (things get more interesting soon folks I swear).

Kaipara ki Mahurangi

KKM loses its suburban fringe around Massey (to Henderson - 2,550) and Whenuapai (to Upper Harbour - 4,940) while picking up 360 voters from Whangaparāoa. As with any urban/rural fringe shuffle this benefits National (by +2% on net) but given a 44% margin…

Upper Harbour

The first change that could make a difference (albeit only in a 2020-style red wave). The gain of deep-blue Schnapper Rock and Rosedale from ECB (2,780) and more mixed Whenuapai and Westgate from KKM (4,940) net the right 2.3% on the party vote. However, the bigger impact comes from losing light-blue Glenfield to Northcote (2,040) and especially left-leaning Massey to Henderson (9,160) for a net gain of 7.2%.

Combined these give the parties of the coalition a gain of 9.4% on the party vote and National MP Cameron Brewer an estimated net gain of 7.6%. This moves a seat Labour were able to win in 2020 completely out of reach.

West Auckland

The three seats in West Auckland see significant changes (all three have been renamed) that benefit Labour. More than the gains from the north-west and losses to the south-east, the boost Labour sees comes from a more efficient distribution of its voters among the three seats.

Henderson (was Te Atatū)

Henderson gains 2,550 voters from Kaipara ki Mahurangi and 9,160 from Upper Harbour. While these areas were left-leaning by KKM and UH’s very blue standards, they’re slightly more National-friendly than Te Atatū’s base was, so they push the seat 0.8% to the right.

This is more than cancelled out by a population exchange with Glendene. The loss of the moderate Te Atatū peninsula (17,140 voters) in exchange for a solidly Labour area around McLaren Park (2,080) combine to shift the seat 2.1% to the left on the party vote.

In candidate vote terms this should add 1.4% to a much-needed buffer in a traditional Labour seat National all-but took in 2023 (incumbent Phil Twyford hung on by just 131 votes).

Glendene (was Kelston)

Glendene was and remains (or rather Kelston was and Glendene will be) the most solid Labour seat out west. In addition to the exchange with Henderson, Glendene swaps Glen Eden and the northern edge of Titirangi (13,410) for parts of Avondale and New Lynn (3,010) with Waitākere. Finally, it sheds Waterview and 2,010 votes to Mt Albert.

On the party vote this moves the seat 6.8% to the right, and narrows Labour’s candidate vote by 5%. However, this still would have left Labour deputy leader Carmel Sepuloni with a healthy (in a bad year like 2023) 8.4% lead.

Waitākere (was New Lynn)

Waitākere sees the first of two flips on 2023 numbers. The swap with Glendene nets the left a 2.2% boost, while losing Blockhouse Bay (10,540) to Mt Roskill adds a further 2.3% on the party vote.

In candidate vote terms, this would have been enough for Labour to have held off surprise National winner Paolo Garcia by a still-narrow 1.2% margin.

Auckland Isthmus

Given relatively low (and in some cases negative) population growth, the seats on the Auckland Isthmus have had to stretch north-west and south-east to scrape back within quota. In 2023, 4/6 seats hosted two-tick campaigns from minor parties (three of them successful) meaning the partisan impacts from boundary changes are slightly more nuanced than for simple two-party contests.

Mt Roskill

Mt Roskill extends westwards to the edge of the isthmus picking up Blockhouse Bay from Waitākere (10,540 votes) and passing Wesley to Mt Albert (8,430). Both these changes benefit National and its partners on the party vote (2.9% and 0.4% respectively) and a net 2.4% swing on the candidate vote turn this from a soft-red seat National can win in a good year into a genuine toss-up that surprise 2023 winner Carlos Cheung now has a good show of defending against Labour’s Michael Wood.

Maungakiekie

Maungakiekie reabsorbs Ōtāhuhu’s awkward Panmure panhandle, adding 5,200 votes. It also sees some nips and tucks with Epsom (160) and Tāmaki (10). The southward move nets Labour +2.2% on the candidate vote, slightly boosting their chances of recovering what has been a reasonably solid bellwether seat from National’s Greg Flemming.

Mt Albert

Mt Albert saw a third-party challenge from the Greens and a nail-biting 18-vote result between Labour and National in 2023. It picks up Waterview from Glendene (2,010), Wesley from Mt Roskill (8,430), and Kingsland from Auckland Central (780) while handing back greater Grey Lynn to Auckland Central (9,400).

These changes are equally good for Labour and National candidates (+0.9% each) at the expense of the Greens who drop by -1.7% to a more distant third, muting the chances of another three-way contest.

To give a sense of how the changes are affecting the Greens’ chances, the map below shows the 2023 party vote on a Labour vs Greens basis. The net impact of losing Grey Lynn and gaining Wesley is pretty clear:

Auckland Central

In addition to the swap with Mt Albert, Auckland Central loses Grafton and Eden Terrace (3,610) to Epsom. The gains from Mt Albert bolster Green co-leader Chlöe Swarbrick’s lead, slightly offset by the losses to Epsom.

Taken together with the changes in Mt Albert, these changes consolidate rather than extend the Green’s position (in contrast as we’ll see later to the changes in their other seats in Wellington) leaving them with just the one solid seat.

Epsom

Epsom gains 3,610 voters from Auckland Central and a tidy-up 160 from Maungakiekie. On net these changes are good for the left but not in a way that matters given (*checks calendar*)… Deputy Prime Minister David Seymour’s 20% lead over National’s Paul Goldsmith.

Unlike the Labour/Green contests, ACT’s candidate are largely reliant on split-ticket voting from National party voters, so the blue/yellow dynamic matter a lot less. On the party vote though, the results are slightly less-worse for ACT (-0.3%) than they are for National (-1.7%).

Tāmaki

Tāmaki - ACT’s second seat as of 2023 - picks up 2,120 votes from Ōtāhuhu around Point England. Similar to Epsom, this change is notionally good for Labour (+1.5%) but with a party vote share of just 18% and third in the candidate vote, it scarcely matters.

Similar to Epsom on a National vs ACT basis (given the seat is a blue-on-yellow one) the move very slightly favours ACT. More than anything this reflects both party’s overall weakness in Point England rather than any relative strength.

On the whole these changes do little to alter the dynamics of the National-ACT contests.

South-East Auckland

Excess population in Takanini and the need to push population both onto the isthmus and up towards Pakuranga see a cascade of changes around Manukau. Given the sharp partisan cleavage (three safe Labour seats, three safe National seats, one blue-marginal) despite all the population shifts it largely adds up to not a lot.

Ōtāhuhu (was Panmure-Ōtāhuhu)

Ōtāhuhu loses Panmure (hence the name change) to Mangakiekie (5,200) and Point England (2,120) to Tāmaki, almost completely retreating back off the isthmus. At its southern end it picks up Rongomai from Botany (4,210) and Clover Park from Takanini (780).

All the changes except for Takanini boost Labour. While the lost areas around Panmure are solidly Labour-leaning, they’re less solid than the rest of the seat. Conversely, the area gained from Botany is even redder. In many ways Labour would rather have had these voters allocated somewhere (like Takanini) where they could actually use them, given incumbent Jenny Salesa already had a 27% margin.

Māngere

Māngere was very slightly below quota, so picks up 1,250 votes from Manurewa (and gives two (2) back via a tidy-up in a largely unpopulated industrial area). Net gain of 0.8% for whoever vainly flies the National flag next time around.

Manurewa

Offsetting the losses to Māngere, Manurewa pickups up 2,320 voters from an over-quota Takanini. Also a +0.8% gain for a doomed National challenger.

Takanini

The only seat in South Auckland that’s anything like marginal, Takanini was the fastest growing seat in the country between the 2018 and 2023 Censuses meaning it has to lose population in all directions.

• 3,830 to Botany around Mission Heights (partially offset by a 2,170 gain around Donegal Park)

• 780 to Ōtāhuhu on the northern edge of Clover Park

• 2,320 to Manurewa around Manurewa South

• 2,380 to Papakura to the south east of the seats namesake suburb

The losses to Manurewa (R+2.1%) and Ōtāhuhu (R+1.1%) help the right, but are more than offset by the loss to Papakura (L+1.3%) and net loss to Botany (L+4.4%) helping the left and leaving the seat 2.5% more left-wing on the party vote.

On the candidate vote Labour gain an estimated 2.3%. This still leaves National with a solid 20% lead on 2023 numbers, but could be consequential in more balanced nationwide conditions.

Botany

Botany sees the biggest changes of any seat in the country.

It loses:

• 3,490 voters to Pakuranga around Cockle Bay

• 4,210 voters to Ōtāhuhu around Rongomai

• 2,170 voters to Takanini in Donegal Park

It gains:

• 3,830 voters from Takanini in Mission Heights and Flat Bush

• 8,130 voters from Papakura along the coast from Whitford to Maraetai

This leaves incumbent (*check newsfeed*)… Prime Minister Christopher Luxon with the largest candidate vote gain of anyone in the country at +11.5%. Given he already held it by a margin of 44% in 2023 it’s won’t be good for much more than a morale boost.

Pakuranga

Pakuranga picks up Cockle Bay, Shelly Park, and 3,490 voters from Botany. Given the areas gained look more or less like the rest of the seat, this has effectively no impact.

Papakura

Finally on Auckland’s southern fringe, Papakura loses the eastern beaches (8,130 votes) to Botany while gaining 2,380 from Takanini.

The losses to Botany cause a large 4.3% leftwards swing on the party vote, and would narrow National’s margin by 3.1% on the candidate vote, but given this still leaves them with a 29% lead it is (once again) largely inconsequential.

Bay of Plenty

In what I think is some of the Representation Commission’s best work this review, they’ve finally tidied up the boundaries around Tauranga, rationalising a messy inner/outer split into a sensible east/west one.

Mt Maunganui (was Bay of Plenty)

What was Bay of Plenty loses a semi-rural area to the southwest of the city (6,700 votes) to Tauranga in exchange for the now-eponymous Maunganui (5,500). At the same time it sheds two areas (Waiorohi and a rural strip south of Papamoa) to Rotorua (1,350 votes between them).

Essentially no partisan impact.

It’s worth diverting here for a moment to look at how the Bay of Plenty electorate has retreated over time. What’s now Tauranga’s eastern suburbs once stretched all the way to Cape Runaway and the border with Tairāwhiti. Amazing what 30 years of population growth can do.

Tauranga

Reverse of the changes described above and similarly negligible partisan impact.

Rotorua

Rotorua picks up 1,350 voters from Mt Maunganui (necessitated by quota considerations). As with the other seats in the Bay, this doesn’t change much.

Eastern North Island

Napier was very slightly (~1%) under quota, so had to pick up voters from either East Cape or Tukituki. In a (sensible) change from the draft decision, this now comes at the southern end from Tukituki rather than a messy change at the northern edge.

While it didn’t see any border changes, East Coast is now (more accurately) renamed East Cape.

Napier

Napier gains a rural wedge of Tukituki between the Ngaruroro and Tutaekuri rivers (1,610 votes). Very slightly favours National candidates (+0.4%) but would only matter in a close year.

Tukituki

Reverse of the change to Napier. Slightly favours Labour candidates (+0.4%).

Western North Island

On the western side of the lower North Island every seat sees changes. It’s tempting to start this chain of seats from the south rather than continuing from the north, given the whole cascade is kicked off by the loss of a seat in Wellington.

Taranaki-King Country

TKC absorbs Taumaranui (1,850) from Rangitīkei given the area was isolated from the rest of Rangitīkei by the Whanganui-Rangitīkei boundary shift. It also loses a few blocks of Waitara (110) to New Plymouth in a CoI tidy-up. Narrows National’s margin by -0.7% but not in a way that matters.

New Plymouth

This is only included because I’m a completionist.

Whanganui

Whanganui picks up the western Central Plateau (2,900) from Rangitīkei. While this nearly doubles the area of the seat, the partisan impact is small (widening by +0.4%) but given the seat is a blue-leaning marginal in a close year that could matter.

(It also picks up an uninhabited area north of Ōhakune from Taupō).

Rangitīkei

In the middle of the LNI chain of changes, Rangitīkei is seeing nearly 40% of its voter base turn over. As a result of the three-into-two merger in Wellington, Rangitīkei needs to absorb 17,390 voters around Foxton and Levin from Kapiti. To partially off-set this it then loses:

• Taumaranui (1,850) to Taranaki-King Country

• The western Central Plateau (2,900) to Whanganui

• The outer suburbs of Palmerston North (6,530) to… Palmerston North

• Ashurst (1,970) over the hill to Wairarapa

Of these, the gains from Kapiti have the biggest impact, moving the party vote 6.2% to the left and narrowing the candidate vote margin from 24% down to 19% on 2023 numbers. This is one of the larger changes we’ve seen and moves the seat to more of a blue-marginal than safe National seat.

Palmerston North

Palmerston North was under quota, and as an enclave of Rangitīkei could only absorb voters from there. It extends outwards to pull in all of the city’s urban area and some of the rural fringe, adding 6,530 voters. As you’d expect, this pulls the seat 3.9% to the right, and narrowing the Labour candidate’s margin by 2.8%.

In most cases this should see the seat remain with Labour, but it’s now gettable in a National landslide year.

Wairarapa

A slightly odd decision (but one necessitated by overpacking everything south of Taupō) sees Ashurst and its surrounds (1,970 voters) moved into Wairarapa despite the significant natural barrier of the Tararua ranges.

While you can critique it in community of interest terms, in partisan terms it doesn’t have any impact.

Wellington

Forced by apportionment maths to abolish one seat in the North Island, the Commission chose to effectively merge the Wellington seats of Ōhāriu and Mana, reducing the number of Wellington seats by one and kicking off a cascade of changes as far north as Taumaranui and spilling over the Tararuas and into the Hutt Valley.

On one level, the overall changes are bad for Labour given the Commission abolished a seat they hold. As compensation though, they now have better chances in the remaining seats versus National to the north and versus the Greens to the south.

Kapiti (was Ōtaki)

Kapiti sees over 40% of it’s population turn over, losing the Horowhenua (17,390) north to Rangitīkei while picking up the southern Kapti Coast and Porirua City as far as Whitby (20,580) from what was Mana.

Both these changes help move the seat to the left on the party vote (1.8% and 2.2% respectively) and narrow the estimated candidate margin by 1.5%. Not quite enough to move it into Labour’s column, but certainly a target in a close race.

Kenepuru (was Mana)

I’ve notionally chosen to analyse the new seat of Kenepuru as if it replaces Mana (given Labour have confirmed Barbara Edmonds as their candidate) but given the shifts it could just as easily be seen as a replacement for Ōhāriu.

On this basis it picks up 22,080 votes from Ōhāriu while losing 20,580 northwards to Kapiti. The gains from Ōhāriu pull the seat significantly rightwards (5.1% on the party vote) but are more than offset by the leftward push from loses to Kapiti (6.3% leftwards). This leaves a seat that on the candidate vote looks very similar to Mana, with a slight 0.2% increase in Labour’s estimated margin.

Hutt South

Both seats in the Hutt were within quota and did not need changes in their own right. However, to make the rest of the regional adjustments work, Hutt South needed to pick up Woodridge and Newlands (4,890 votes) from Ōhāriu. At the same time it passes 1,780 voters to Remutaka.

On net these changes move the seat very slightly to the left (0.6%) and narrow National incumbent Chris Bishops’s already slender margin by 0.4%. In most other seats that wouldn’t matter but in this case it’s equivalent to 1/8th of his existing lead.

Remutaka

Picks up 1,780 voters from Hutt South to no partisan effect.

Wellington North (was Wellington Central)

Wellington North picks up 16,140 voters from Ōhāriu around the northern hills while losing 8,000 to Rongotai Wellington Bays around Brooklyn and Mt Cook. In partisan terms this is mostly bad for the Greens (-4.4% on the party vote) and good for National (+4.1%).

The candidate changes are similar, although given 2023 was a first-time two-tick campaign for the Greens they don’t read easily across to future races. On these numbers the Tamatha Paul would hold on. The boost in National party voters could end up working to Labour’s advantage if National voters break a bit better for them in a two-way red/green contest, or it could keep things as a three-way contest.

Given the seat is likely a Green vs Labour contest in future elections, it’s worth also looking at how the areas lost and gained look between those two, with the added northern suburbs benefiting Labour and the lost southern ones costing the Greens.

Wellington Bays (was Rongotai)

Under quota and only bordering Wellington North, Wellington Bays has had to pick up 8,000 voters around Brooklyn and Mt Cook.

On a left-right basis these changes move the seat 1.8% to the left.

Even more than Wellington North (where National perform better), Wellington Bays was and remains a green/red contest. On that front, the changes marginally improve the Green’s chances, lifting Julie Anne Genter’s lead over Labour by 1.4%.

In contrast to the Auckland Central/Mt Albert changes, the Wellington changes more evenly redistribute the Greens’ vote. This is a somewhat risky change, as it makes it more likely that they’ll hold both seats, but also more likely they’ll lose both (while reducing the chances they win just one of them).

Christchurch

While the Christchurch metro seats were all within tolerance, excess population in neighbouring Selwyn kicks off a cascade through the city that may see Labour lose a once indomitable stronghold of the left.

Selwyn

Selwyn - host to greater Christchurch’s rapidly growing southern exurbs - was 14% over quota, so in a continuing trend from past reviews loses Templeton and Prebbleton (6,940) inwards to Wigram.

While still right-leaning on net, these areas are closer to the centre electorally as well as geographically. As a result of losing them, Selwyn moves 0.7% to the right on the party vote and sees Nicola Grigg’s already gigantic margin increase by 0.5%.

Wigram

In the most consequential change from the 2025 review, Wigram picks up 6,940 voters in exurbs Prebbleton and Templeton while losing inner urban Addington and Spreydon (5,420).

In a familiar pattern for cities with growth on the urban fringe, both these changes move the seat to the right. Adding Prebbleton/Templeton shifts the dial 5.2% towards the governing parties, with subtracting Addington/Spreydon moving it -3.5% away from the opposition, for a combined 8.6% swing on the party vote.

This change transforms what has been a reliable working-class left-of-centre seat (National have never won it or its predecessors) into a hotly-contested suburban marginal.

On the candidate side, in 2023 the new boundaries would have flipped Megan Woods’ majority on its head - going from a 3% win to a -3.2% loss, and may have contributed to Woods’ decision to go list only from 2026.

Christchurch Central

As the inward cascade continues, Christchurch Central adds two lobes to its cephalopod-like borders (Addington and Spreydon, 5,420 votes) while also lopping off two - Northcote to Ilam (950) and Linwood to Christchurch East (3,110)

These changes more or less offset, with the gains from Wigram pulling things 1.5% to the left and the losses pulling them back 1.4% to the right.

Ilam

As noted Ilam picks up 950 voters from Christchurch Central around Northcote. Limited party vote effect (also note this was a three-way race with TOP in 2023, so candidate figures probably aren’t a good guide in future).

Christchurch East

Christchurch East picks up Linwood and 3,110 voters from Christchurch Central, moving the seat 1.8% to the left on the party vote and bumping up incumbent Reuben Davidson’s margin by 0.9%.

Southland

Finally among the general seats, slow-growing Invercargill continues its slow march northwards into faster-growing Southland.

Invercargill

Invercargill was just barely under quota, so takes another nibble out of rural Southland around (*zooms in on map*) Drummond and Taringatua? This notches the seat 0.5% to the right and adds 0.4% to National’s Penny Simmonds’ margin.

Southland

Rather than the more thoroughgoing rearrangement proposed by some submitters both now and in past reviews, Southland instead just loses the now-customary minimum population necessary to Invercargill.

Māori seats

Note than rather than left-right PVI, the maps below show Labour vs Te Pāti Māori PVI compared to the Māori seats overall (so how much an area leans towards Labour or TPM). Lost areas are shown in purple rather than orange for clarity. As this is a 2023 rerun, it does not account for candidates who have since been removed from Te Pāti (a modelling headache for another day).

The Māori seats have been left largely as-is, with enrolments and population growth on the Māori roll not adding enough for an eighth seat. Only Ikaroa-Rāwhiti and Te Tai Tonga see changes.

Ikaroa-Rāwhiti

Ikaroa-Rāwhiti (the only Māori seat held by Labour after 2023) was slightly under quota, so absorbs 1,530 voters in Te Awakairangi from Te Tai Tonga.

Given the urban areas around Te Upoko o Te Ika lean more towards TPM (and the Greens) than the more strongly Labour-aligned Tairāwhiti, this sees a 1% net TPM swing on the party vote, and would have narrowed Labour’s candidate majority by 0.4% (smaller effect due to higher levels of vote-splitting).

Te Tai Tonga

Rounding out the adjustments on a fairly quiet note, Te Tai Tonga loses 1,530 voters to Ikaroa-Rāwhiti with no especially meaningful partisan effect (weirdly good for National, given they do better on Te Waiponamu).

Wrapping up

If you’ve made it this far: well done and I’m sorry. This probably should have been split into a series but it was easier to just do it all in one go. A slightly hyperfixated piece but one that’s thrown up some data and topics for future pieces.

This is the final part on boundary adjustments, and with it out of the way (and using some of the results of this analysis) we can turn to previewing the 2026 races.

Ka kite anō.

Well I said I didn’t have time to look over council races outside Wellington and I don’t. However, a reader provided thoughts in the DMs and from afar (my knowledge of Christchurch is patchy at best (it’s the one north of Dunedin right?)) they seemed like a reasonable take, so they’re reproduced below with permission.

Basics

Christchurch elects the mayor citywide by FPP and has 16 single member FPP wards for 17 people around the council table (each are also elects 2 local board members by FPP - but I’m not going that in depth).

A majority of candidates tend to run as independents, so for simplicity’s sake I have split candidates into 3 groups (1 group has 2 sub-factions).

Note that sometimes the distinctions can be rough and can sometimes basically a judgement call based on who they’re running against for non-incumbents

• 🔵Centre right (CR): right leaning independents and independent citizens candidates

• 🟣Centre: councillors Scandrett and Moore, who have been the swing votes this council, Manji also somewhat goes here although he was probably CR when on council previously.

• Left: made up of two sub factions, 🔴People’s Choice (PC, basically Labour) and 🟢Greens (councillors Donovan and Templeton, both are green aligned independents)

Mayor

I’d be surprised if 🔵Mauger lost (CR hold) and faces his main challenge from sitting Green-aligned councillor 🟢Sarah Templeton.

Unopposed races

Given single-member FPP wards where others would face little chance against established incumbents, three candidates have no opposition.

• Waimari: 🔵McDonald

• Harewood: 🔵Keown

• Hornby: 🔵Peters

Safe holds

• Papanui: 🔵Henstock ran a good campaign to oust Davidson last time (now the new Green MP), only campaign I’ve ever seen say “actually we don’t need flyers delivered we have enough volunteers”, up against 🔴Britten for PC, I’d say she wins easily (CR hold)

• Fendalton: most right-leaning seat. Gough retiring, 🔵Cartwright will win easily (CR hold)

• Halswell: 🟣Moore is a popular centrist and wins easily, current median councillor (Centre hold)

• Spreydon: 🔴Coker (PC) wins easily, Craze is entering perennial candidate territory and Tubby Hanson is one (PC hold)

• Cashmere: Popular centrist 🟣Scandrett wins easily (Centre hold)

• Linwoood: 🔴Johnson wins easily, probably the most mavericky left bloc councillor (PC hold)

Likely holds

• Coastal: likely 🟢Donovan but some potential for an upset as she’s running against local board member Jo Zervos (green hold)

• Burwood: incumbent 🔵 Barber probably holds here over 🔴Chapman (CR hold)

• Central: Lowest turnout (and most left-leaning in a general election) ward in the city and 🔴McLellan (people’s choice, Tracey McLellan’s son) usually runs fairly close, former TOP leader 🟣Raf Manji carpetbagging to run here. Probably leans PC (PC hold).

• Banks Peninsula: 🔴Fields probably easily holds here after barely scraping in 2022 (PC hold)

• Riccarton: 🔴Harrison-Hunt (PC) barely won here in a car crash multi candidate race (BP was similar), likely favoured to hold in here up against 4 other candidates although the CR do like their chances so probably the best CR pick up opportunity after Heathcote/Innes (PC hold)

Key match-ups

• Innes: Marque matchup, 🔴Cotter beat 🔵Jones last time only once the specials came in. Jones won back her seat she vacated on the community board a few months later, pure toss up and a key race for council control (toss up)

• Heathcote: 🟢Templeton is retiring to run for mayor [editor’s note: must be nice] so this becomes a the other marque matchup for council control, 🔴Herz vs 🔵Kearney so a loss for the greens either way (Toss up)

In summary

• 🔵 Centre: mayor and 6 councillors (they’re favoured in all their seats)

• 🟣Centre: 2 councillors (both incumbents favoured)

• 🔴People’s Choice: 5 councillors (5/6 incumbents favoured)

• 🟢Greens: 1 councillor (1/2 incumbents favoured)

• Toss up: 2 wards (1PC and 1 green)

Barring an upset council likely won’t shift left (I’d have the right picking up a seat outside the toss up column rather than the left) so council ends up anywhere from:

• 7-2-8 to 9-2-6 in the usual scenario

• 6-2-9 in a great night for the left

• Something crazy like 10-3-4 in a landslide for the right

Likeliest seats to flip are probably (in descending order):

• Heathcote

• Innes

• Central

• Mayor

• Riccarton

• Coastal

Given The Overhang is Wellington-based we’re doing a quick preview of how the Wellington City Council races might shake out, with a few things to look for along the way. Apologies to the rest of the country, feel free to skip this one if you like – it’s a mildly self-indulgent one.

This is a structured reckon rather than a proper forecast. There’s no especially strong quantitative basis for any of it beyond previous results and a general awareness of how STV dynamics work.

tl;dr version

• Mayor 🔴Andrew Little is going to be mayor.

• Takapū/Northern will see at least two incumbents (🔴McNulty and one of 🔵Randle or 🟣Apanowicz ) and probably all three back.

• Wharangi/Onslow-Western should see one progressive (🟢Matthews or 🔴Gribben ), C🔵alvert, and probably 🟡Chung elected.

• Pukehīnau/Lambton 🟢Rogers and 🔵Young both feel safe, but the contest for the third spot in the middle could be close and consequential (likely 🔴Al-Rubayee but potentially any one of seven moderate-to-conservative independents).

• Motukairangi/Eastern with no incumbents running is the most interesting, uncertain, and important race. Best guess is one Labour/Green, one moderate/Labour, and one conservative.

• Paekawakawa/Southern will see both Foon🟢 and Abdurahman🔴 back in almost any plausible scenario.

• Te Whanganui-a-Tara is a coinflip between 🟢 Whanau and 🔴Reweti.

Basics

A few common themes to remember as we rip through:

1. The city is very, very left-wing: even in a relatively bad general election (2023) the left won 66%-33%, and the Greens were the largest party. It feels trite to say that, but I think with some coverage over the campaign the point gets lost.

2. The local government electorate is significantly more conservative: given lower turnout and fuzzier partisanshipk, across the wards the broad left took 51% in 2022, with the right taking 45%.

3. Incumbents almost always win. In low-information, low-participation elections public profile and even simple name recognition count for a lot. Since 2007, 80% of incumbents who run are re-elected (83% if you exclude mayors).

4. People are not happy. While incumbency is usually an asset, this might not be the case in a situation where the council’s overall approval rating is just 27%.

5. Because of how STV works, races are less a free-for-all between every candidate and more a contest between ideologically adjacent candidates within “lanes”.

Mayor

🟢>🔴

Andrew Little is going to win. There’s really not much more to it than that. As noted, the city has a red/green lean, the 2022 mayoral race saw progressive-adjacent candidates win 60% of the first round vote, and he’s leading every poll.

A race with either Whanau remaining in the contest or a formal Green alternative could have been a more interesting test between Labour and the Greens, but that was not to be. The lack of a Green contender at the top may end up costing the Greens down ballot if there’s an enthusiasm gap.

The interesting questions we’re left with are with how the count unfolds and how runners up perform:

• Does Little romp home on the primary vote, or does he need to wait on preference flows from Baker and to a lesser extent Calvert?

• Which of the three main conservatives (Calvert, Teifenbacher, Chung) performs the best? How they finish will tell us something about the overall crankiness (in that order) of the electorate.

• How well does lower-case green candidate Alex Baker do? Second on the primary vote and third overall would be an encouraging benchmark for future upper-case Green mayoral attempts.

Te Whanganui-a-Tara

🟢>🔴

The Māori ward left vacant by surprise 2022 Green winner Nīkau Wi Neera (who won by just 67 votes) should see another photo-finish in 2025 between departing mayor Tory Whanau for the Greens and returning Labour candidate Matthew Reweti.

On the one hand, Whanau is the literal mayor. On the other hand, that might not be a good thing given a… turbulent… term. On the other other hand, Whanau may find a more sympathetic audience in the overwhelmingly progressive Māori ward than she did in the city at large, and so the criticisms may not land. *Growing a worrying number of hands* Labour’s Reweti ran an incredibly close second in 2022 and now has an across-the-board iwi endorsement.

Best guess is the iwi endorsement, the controversy, and Little at the top of the ticket swing things to Reweti but 🤷‍♂️.

Takapū/Northern

🔴🔵🟣> 🔴🔵🟣

Wellington’s northern-most and right-most ward sees all three incumbents (Labour’s Ben McNulty, conservative independent Tony Randle, and moderate-ish independent John Apanowicz) recontest their seats.

Given the above-mentioned advantage of incumbency and the other advantages the candidates have, on balance all three will be re-elected. Even if they’re not, the result is unlikely to affect the overall partisan complexion of the council.

Centre-left lane

Cr McNulty has the progressive lane to himself given the absence of a Green or independent left candidate. While the ward is baseline (in a general election) the city’s most conservative, progressive parties still out-polled conservative ones 53/47 in the 2023 general. In 2022 (even with the more right-leaning local turnout) progressive candidates got 43% of the primary (first round) vote and gave McNulty an effective majority of 10% in the 7^th round. Add in incumbency and profile and he’s a safe bet.

Centre-right lane

The closest contest is likely to be for the slightly more centre right lane. Currently occupied by Cr Apanowicz, in 2022 he scraped in ahead of progressive centrist incumbent Jenny Condie by just 104 votes, or an effective majority of 1% in the 9^th and final round (the smallest margin in the city). That said, he started with a respectable 15% of the primary vote and also now benefits from incumbency. The most likely competitors for his spot are independents Michael Hill and Joan Shi, or Cr Randle if he get pushed out of the right-most lane.

Right lane

For the spot on the furthest right, Cr Randal sees challenges from ACT Local’s Mark Flynn and former Independent Together (IT) candidate Andrea Compton, as well as Cr Apanowicz and others to his centre. Facing a similarly sized field in 2022 Randle managed to garner 17% of the primary vote and had an effective majority of 8%, also in the 7^th round. Again, this was without incumbency. On the other hand, there is the potential for the brand-recognition the ACT and (former) IT candidates have in a conservative suburban seat with a grumpy voter-base to work against him.

Wharangi/Onslow-Western

🔵🟡🟢>🔵🟡🟢

Closer to the centre geographically and politically, Wharangi also see two conservative incumbents (conservative independent Diane Calvert and IT leader Ray Chung) and one progressive incumbent (former-Labour-now-Green Rebecca Matthews) defending their seats.

While the seat has a Khandallah-inflected reputation as a conservative bastion, at a national level it’s significantly further left than Takapū with a 20% progressive party lead. While Khandallah lives up to its reputation, this is out-weighed by centre-suburban Karori and the redder/greener inner hill suburbs.

This isn’t neatly reflected in local elections, with a combined Labour-Green primary vote of just 30% in 2022: still good enough for one lane but not enough to make a second competitive and leaving two for the centre-right/right.

As with Takapū, the results probably won’t change the left/right balance of the council table, but could affect the relative position of the Greens vs Labour or the right vs the further-right.

Left lane

The progressive lane is a two-person race between Green incumbent Cr Matthews and Labour challenger Joy Gribben. The two parties are approximately even (at 27% a piece) in parliamentary elections, while in 2022 Labour performed better on the primary (18% vs 12%) and final round (27% vs 20%) ward vote. Of course, Matthews was the incumbent Labour candidate then, so the main question is whether she can bring approximately 1/8 of her voters (~4% of total) with her to the Greens. Given name-recognition and softer partisan loyalties among local voters that seems achievable. If it turns out the vote was more Labour and less Matthews, then Gribben could sneak in.

Centre-right lane

The centre-most of the right lanes seems to be a lock for incumbent conservative independent Cr Calvert: a primary take of 21% and an effective majority of 5% (albeit in a later round than Cr Chung) combined with her more pragmatic reputation make her stand out in field crowded with candidates to her right.

Her most likely challenger is IT-defector Lily Brown who has been very active and visible if nothing else. The other, less likely option could be a Chung blow-out spilling over to IT understudy Guy Nunns but given how the IT campaign has gone that seems unlikely.

Far right lane

Purely on the numbers Chung seems very safe, and his place as right-wing standard bearer in the mayoral race gives him added prominence. The question is – given the controversy he’s courted – is whether that attention is a good thing. Difficult to see him being pushed out by one of his own or by ACT Local’s Ray Bowden unless the soft pendulum has really swung against him.

Pukehīnau/Lambton

🟢🔵🟣>🟢🔵🔴

The second most progressive and the Greenest general ward - not only in the city but in the whole country - the city centre sees 2024 by-election winner Geordie Rogers and long-term conservative incumbent Nicola Young defending. Retiring green independent Iona Pannett leaves a lane in the centre open, and the first spot that could tip the overall balance.

Left lane

Despite a scare in the 2024 by-election, in a three-seat race Rogers is essentially a lock for the left lane. His predecessor Tamatha Paul won with a show-stopping 67% effective majority (enough to be elected nearly be elected twice) and while the 2024 result was close, in a three seat race rather than a one seat one his primary vote was similar.

Centre lane

The spot vacated by Pannett leaves one of the widest races in the city: with anyone from Labour’s Afnan Al-Rubayee, through a gaggle of moderate independents (Rodney Barber, David Lee, Teal Mau, Tim Ward) to erstwhile IT independent Dan Milward possible contenders.

Given her close fourth place behind Pannett last time, the top-of-the-ticket boost from Little, and likely preference flows from Rogers, Al-Rubayee seems the best bet. That said, Ward’s prominence in the Te Aro hospitality community and moderate-business friendly vibe could pull votes from both sides. Finally, Lee is a former Green and probably best fits the Pannett-shaped gap.

Centre-right lane

Young is all but assured of holding this spot with a 17% primary vote and an effective majority of 12% once she absorbed fellow conservative Karl Tiefenbacher’s preferences. Given the crowd on the right and an anti-incumbent mood you could see her do worse on the primary vote, but as less prominent right candidates drop she’ll harvest a lot of preferences.

Motukairangi/Eastern

🔴🟣🔵>🔴🟣🔵

With no incumbents running, it’s not much of an exaggeration to say Motukairangi will decide the overall balance of power on council. Retiring incumbents Labour’s Terri O’Neill, independent green Sarah Free, and maverick-conservative Tim Brown leave a left, centre, and right lane wide open for a broad field. In general elections the seat is a 66/33 left/right split and has tended to produce similar splits locally.

Centre-left lane

Notionally held by Labour’s Cr O’Neill, the left-most spot could fall to any one of: Labour newcomer Sam O’Brien, capital-G Green Johnny Osborne, or lower-case green and mayoral aspirant Alex Baker.

O’Neill ran third on the primary vote at 15% in 2022 which is a good sign for O’Brien (tho he’s not running as an incumbent). On the other hand, the 2022 Green candidate was close at 14% – which was high given Free (a former Green nominee) will have held on to some green voters. Very hard to pick which of O’Brien, Osborne, or Baker will take this spot but if you really pushed me I’d say O’Brien does with a turnout lift from Little’s race.

Centre lane

The centre lane previously occupied by Green-turned-independent Cr Free (who is attempting to kick herself upstairs to the regional council) offers O’Brien and Baker a second chance, joined by progressive-ish independent Trish Given and moderate independent (and former councillor) Chris Calvi-Freeman.

Based on vibes and positions, Given seems nearest replacement for Free and to have campaigned strongly. Baker could also stray into the centre lane and take it if the Labour or Green candidate bump him into it, with a boost from his mayoral run. Finally Calvi-Freeman has won before and could do so again.

I think this spot is probably the hardest call in the city.

Centre-right lane

Finally the right-most lane vacated by Cr Brown. While notionally on the right, Brown occasionally tacked to the centre in office, siding with the progressive majority on several key votes – similar to Takapū’s Apanowicz but with more uh… flair.

The most prominent conservative in the race and probable favourite is Karl Tiefenbacher – unless he fails to dodge the carpetbagger problem having twice run and lost in Pukehīnau. There’s also returning candidate Ken Ah Kuoi – now on the IT ticket and with a respectable if unexceptional 8% primary vote in 2022, probably a better bet than fellow IT candidate Michelle McGuire or ACT Local’s Luke Kuggeleijn. The less said about former councillor Rob Goulden the better. And then there’s finally the frankly unclassifiable independent Thomas Morgan who’s proposed… privatising playgrounds??

With conservative-ish candidates taking ~50% in 2022, one of them is basically guaranteed to win. Most likely Ice Cream Karl but with an outside punt on Ah Kuoi.

Paekawakawa/Southern

🟢🔴>🟢🔴

The most left-wing ward in the whole country (75% in the 2023 general) Paekawakawa will be re-electing Green deputy mayor Laurie Foon and Labour’s Nureddin Abdurahman.

Left lane

In 2022, Foon took the first left spot with 37% in the first round and an effective majority of 10% (second only to Pukehīnau’s Tamatha Paul). Guilt-by-association with mayor Whanau and general dissatisfaction with the Greens might see her bumped into second by Abdurahman, but it’s a big majority to lose.

Also the left lane

While it took 10 rounds of preference flows for Abdurahman to take the second spot – building from a 20% primary vote to a final round vote of 37% and an effective majority of 10% over returning challenger Paula Muollo. In the rematch I suspect he’ll do better Muollo’s IT endorsement notwithstanding.

And with that, in the wise words of perennial fan-favourite Don ‘Newt’ McDonald, I am McDone with this.

Summing up

Best guess predictions to hold myself accountable:

• Mayor: 🔴Little

• Te Whanganui-a-Tara: 🔴Reweti (maybe, really could be Whanau)

• Takapū/Northern: 🔴McNulty, 🔵Randle, 🟣Apanowicz

• Wharangi/Onslow-Western: 🟢Matthews, 🔵Calvert, 🟡Chung

• Pukehīnau/Lambton: 🟢Rogers, 🔵Young, 🔴Al-Rubayee

• Motukairangi/Eastern: 🔴O’Brien, 🟣Given, 🔵 Tiefenbacher

• Paekawakawa/Southern: 🟢Foon, 🔴Abdurahman

Bringing all that together the parties’, definitely-not-parties’, and factions’ chances look something like:

• 🟢Greens: min 2 (Rogers, Foon), med 3 (+Matthews), max 5 (+Whanau, +Osbourne/Baker).

• 🔴Labour: min 3 (Little, McNulty, Abdurahman), med 6 (+O’Brien, +Al-Rubayee, +Reweti), max 7 (+Gribben).

• 🟣Moderates: min zero, med 2 (+Apanowicz, +Given), max 4 (+Calvi-Freeman, +Lee/Ward/Mau).

• 🔵Conservatives: min 3 (Randle, Calvert, Young), med 4 (+Tiefenbacher), max 7 (+Compton/Shi, +Brown, +Milward).

• 🟡Independents Together/ACT: min/med 1 (Chung), max 4 (+Flynn, +Wong, +Ah Kuoi/Kuggeleijn/McGuire).

Local elections are weird, turnout looks to be slightly up, and polling is thin so I’m probably going to eat crow on several of these.

Vote if you haven’t.

PS: as the Pukehīnau by-election should have taught us all, special votes can make a difference - with anything up to a ~5% swing to the left possible. So don’t panic/celebrate prematurely on Saturday night.

See, I said we would be back soon. (Welcome to neurodivergent hyperfixation week).

Today we continue an unnecessarily deep-dive into the process for redrawing electorate boundaries. Parts one and two dealt with the number of electorates and the broad population changes since 2018. This part looks at the rules of the game.

Parts 4 and 5 will look at how this applies to the adjustments we might see on the ground (a review preview, if you will).

tl;dr version

• Electorates should in general be equal to the quota

• Electorates must be within plus or minus 5% of the quota

• When drawing boundaries, the Representation Commission must consider:

  • existing boundaries;

  • communities of interest (including for the Māori seats, communities of interest to Māori (ie: hapū and iwi));

  • facilities of communications;

  • topographical features; and

  • forecast population changes.

• These considerations are not deterministic and judgment matters. The Commission faces some high-level choices of ‘philosophy’.

• Despite this wide discretion, the Commission (or anyone objecting to or predicting their choices) face unavoidable geographic and demographic constraints.

Statutory criteria

The Electoral Act 1993 sets out what the Commission must or may do, and what they have to consider. The primary considerations are the electorate quota and tolerances (seats of roughly equal size). Secondary considerations relate to community and physical links or barriers across the country. Finally, there are things the Commission is (implicitly) not allowed to factor in.

Quotas and tolerances

The quota is the average size of an electorate for each of the North Island, South Island, and Māori seats. Either side of this is a 5% tolerance.

The quota itself is a target rather than a requirement: it’s good if you can hit it, but if you have good reasons to depart from it, you can. The tolerances are mandatory: seats must fall within them, no exceptions.

To recap from the previous issue:

• North Island quota: 69,875 (tolerance 66,381 to 73,369)

• South Island quota: 70,037 (tolerance 66,535 to 73,539)

• Māori seats quota: 74,367 (tolerance 70,649 to 78,085)

Existing boundaries

First on the list of what the Commission has to consider is the boundaries of the electorates as they stand. In practice this has meant using them as a starting point. Giving priority to existing boundaries is important because:

· It minimises disruption (and confusion) for communities and their representatives.

· It constrains the near infinite choices presented by a blank map.

· It helps foreclose opportunities for gerrymandering by limiting partisan innovations. This is only true to the extent the existing boundaries are fair, but given the fairly robust processes over multiple years, this is largely the case.

Communities of interest

Every kind of collective, geographically-rooted identity. The Independent Electoral Review used the paraphrase “a group united by shared interests or values”. These are as many and varied as the varieties of human life, including but not limited to:

• Urban areas like towns, cities, or suburbs

• Employment and commercial centres

• Community facilities like schools, clubs, or marae

• Ethnic or other identity communities

• Local government boundaries at a regional, local, and ward level

• For Māori seats in particular (though arguably not exclusively) hapu, iwi, and their rohe

To make this a little more tractable, we will tend to follow pre-defined geographies in our analysis: local government boundaries, the StatsNZ statistical area hierarchy, urban/rural area classifications, or Te Kāhui Mānhai’s Iwi Map.

In a lot of ways, this criterion could stand on its own. Communication and topography will where relevant give rise to community links. On the other hand, it gives wide discretion to the Commission. A lot of the arguments in the objections (consultation) process tend to focus on what gets to count as a community, what forms of collective identity get prioritised, and who gets to define them.

Facilities of communications

A somewhat old-fashioned term (the wording in the Act goes back to at least 1927) that basically boils down to infrastructure that links communities. Historically, this might have included postal facilities and print or broadcast media markets. With ubiquitous, geography-less digital communication, it now essentially means transport infrastructure:

• Most importantly road links, as physically-adjacent parcels of land may be not be connected by road (topographically close, but not topologically close).

• For island communities, ferry links like those between the Hauraki Gulf islands and Auckland Central or Rakiura/Stewart Island and Invercargill.

• For Rēkohu/the Chatham Islands, airports and specifically the link to Wellington airport (in the Rongotai electorate).

• Arguably it could also include public transport corridors, but to my knowledge this has never been used as a basis for a boundary since at least 1996 (as far back as I checked).

Topographical features

This includes natural features like mountain ranges, rivers, coastlines, or forests, but also human structures like motorways or railways. Usually these present barriers (reasons to separate areas) but in some cases (valleys, islands) can give reasons for keeping areas together.

The most prominent example is the Main Divide along the spine of each island. You also see it on a smaller scale like dividing Hamilton largely by the Waikato River, or keeping the two Hutt electorates contained to the valley.

Forecast population changes

Finally, as well as current population the Commission can consider how the population is forecast to change over the life of the electorates (essentially until the next Census).

In most cases this will mean erring on the high side of the quota in shrinking areas or the low side in growing ones.

Irrelevant considerations

Beyond the defined list of factors, the Commission is permitted to consider anything else.

This means excluding partisan political matters. Obvious things like voting patterns or the residences of incumbent MPs are clearly off the table. But so are things like ‘proportionality’ (the results in electorates approximately mirroring the nationwide vote) and ‘competitiveness’ (making seats that give more than one party a decent chance). These can be important and legitimate for FPP systems, but are not an issue under MMP where the party votes takes care of things.

Not that any of this will stop parties and candidates submitting on the process from worrying about the effect on them. But they still need to frame any arguments in statutory terms.

It also excludes ‘aesthetic’ considerations like size, compactness (the ratio of area to perimeter: a measure proffered as an algorithmic cure to the USA’s cursed partisan redistricting processes), or shape (seats can look ‘weird’ provided they follow the rules).

Philosophical choices

There’s the rules as they’re written, and then there’s the rules as they’re applied.

Boundary-drawing is an art as much as a science, and the Commission (or anyone else involved in the process) has decisions to make about how to resolve tensions between the factors they’re required to consider.

To simplify how this judgement might be applied, I’ve broken things down into three broad trade-offs:

• Existing boundaries vs optimal boundaries

• Quota-targeting vs tolerance-targeting

• Present focus vs future focus

When predicting, analysing, and critiquing the work of the Commission, we will try to do it through the lens of these choices.

Existing boundaries vs optimal boundaries

As mentioned above, past Commissions have generally started with existing boundaries and then made adjustments at the margins.

This is a legitimate choice, but it is still a choice.

The alternative is to start with defining communities of interest, looking at the geography that divides or unites them, then assemble them into sensible units. This is more difficult and more disruptive, but can avoid path-dependence and a calcified status quo.

As for all three axes, rather than two mutually exclusive approaches these are poles on a continuum.

If you will permit me to lapse from analysis into opinion for a moment – I think past Commissions have been unduly conservative in their approach and too deferential to existing boundaries. This can lead to perverse (albeit sometimes unavoidable) outcomes. We will get to some examples later in the week, but the split of Queenstown and Wānaka or the stretched mess of Taranaki-King Country come to mind.

Given the scale of disruption already necessitated by the number of electorates outside tolerance and the loss of a seat, this go around could present an opportunity for some much-needed rationalisation. But that’s hard and controversial, so I don’t see it happening. For the purposes of previewing changes, we’ll assume the Commission strongly favours existing boundaries.

Quota-targeting vs tolerance-targeting

This choice is about whether you’re aiming for the quota target, or whether you’re just trying to sneak inside the tolerance limits. The 2020 review staked out a clear minimalist position given the issues with the 2018 Census and the disruption of COVID.

For the purposes of analysis, we’ll assume the Commission takes a moderate quota-targeting approach (still not aiming for perfection, but less willing to accept exceedances than in 2020).

The minimalist approach is easier: do as little as you have to and leave less for people to object to. On the other hand, especially at the draft decision phase more thorough quota-targeting can both flush out what communities people care about and leave room to make changes in response to objections.

Too conservative an approach also risks stocking up problems for future reviews which brings us to…

Present focus vs future focus

How much you solve the problem in front of you versus how much you solve tomorrow’s problems. Too much focus on the present and your boundaries are stale by the time you use them and you leave a mess for the next guy. Conversely, predictions about the future are necessarily speculative and growth patterns over the medium term are fickle.

You can square this circle by erring on the high side for areas in long-term decline and on the low side in areas of higher growth.

Given the distance between the Census and the first election where the seats will be used (three years, about the longest gap you could have) there’s a good case for a slightly more future-focused approach than in the past.

You could bundled this distinction in with the quota/tolerance choice above, but as we’ll get into tomorrow, the degree of future focus matters a lot when choosing which seat to remove.

A long, bumpy string

Despite all these requirements, considerations, discretions, and philosophies in the end we are all slaves to context.

The over-riding reality is that Aotearoa is a long, skinny country divided by mountains.

This means that in a lot of cases you simply don’t have a lot of choice about where the boundaries go. The seats are laid out in chains between the mountains and the sea (or the sea and the sea) and the boundaries are (within certain margins) just a matter of where the people are.

Which is as good a way as any to introduce this monstrosity: the Electoral Sefirot. I know it’s a lot, but believe me when I say it’s a surprise tool that will help us later.

The graph shows:

• Each of the current electorates and their connections to their neighbours.

• Orange electorates are over quota, purple ones are under quota.

• Electorates with white text are outside tolerance.

• Solid lines show boundaries where an adjustment can be made without real difficulty

• Dashed lines show boundaries where you could at a stretch make an adjustment (given transport links) but that where physical barriers mean you should try to avoid it.

Basically, the boundary review is a puzzle-box game where you have to:

• Choose one box in the North Island to get rid of; and

• Shovel people along the lines from the areas of surplus to the areas of deficit until you’ve complied with the rules and submitters stop yelling at you.

Right. With all the groundwork out of the way, we’ll be back tomorrow at start actually looking at how things might go.

Ka kite āpōpō (or maybe Thursday)

Last month, the Representation Commission met to agree draft changes to Aotearoa’s electorate boundaries for the 2026 and 2029 elections. The results will be released for consultation on Monday 24 March. Given significant (relative) population changes and the need to lose a seat in the North Island, this round of adjustments will be more disruptive and likely controversial than previous rounds.

In the week ahead, we are doing a short series on what is driving those changes, the problems the Commission has to solve, and what the solutions might look like. Once the draft and final results are released, we will do some follow-ups on what it could mean for candidates’ fortunes at the 2026 election and beyond.

Today’s piece focuses on:

• Part 1 - The electorate math behind the change to the number of seats

• Part 2 - The broad regional and geographic patters in population changes

(These were originally two separate newsletters but I’m getting closed to being timed-out, so sorry for another longread).

The tl;dr version

• After every Census, the boundaries of the general and Māori electorates are redrawn to account for population changes.

• The total number of electorate seats is decreasing from 72 to 71. The number of list seats will increase from 48 to 49 in response.

• The number of general electorates in the South Island general electorates is fixed (at 16), with the number of North Island general and Māori electorates changing if necessary to maintain parity.

• Because the South Island grew faster (+7.0%) than the North Island (5.5%) between 2018 and 2023, the North Island will lose one electorate (down from 49 to 48). This is the first time this has happened.

• A large number (24) of North Island general seats are over- or under-sized and will need to be redrawn. Only two South Island general seats are out of tolerance so changes will (likely) be modest.

• The number of Māori seats is unchanged (7) and any boundary changes are likely to be marginal, with only one seat under tolerance.

Part 1 - How many seats?

Starting with what should be a simple question: how many electorates should we have?

You would think it would be a straightforward matter of looking at where population has grown and adding one or two new seats. Given the historical pattern since 1996 of adding a seat or two to Auckland after each Census, it would not surprise me if a lot of people thought this was how it works.

It’s not.

It’s more convoluted.

We will start with a primer on how the mathematics of electorate allocation. The official version from StatsNZ is here but I have more pictures and unlike Stats I am allowed to swear and attempt jokes. Skip to Part 2 if you already know all this and just want the upshot.

Electoral populations

First basic concept, everything is done of the basis of total population: not voters, enrolees, adults, or citizens. All ordinary residents are counted. Non-residents (including overseas citizens entitled to vote) are not.

Second basic concept: this is calculated as-of the 2023 Census. This has caused some weirdness because of the COVID-era disruption to growth patterns and will cause some more weirdness given the gap from the Census to the 2026 election.

Anyway, there are (or rather were) 4,993,923 of us.

Māori and general electoral populations

Because Māori voters can opt to vote in either a Māori seat or a general seat, we need to work out who counts towards general seats (the general electoral population) and who counts towards the Māori seats (the Māori electoral population) to keep everything proportional.

A simple answer might be just using the number of enrolees on each roll. Again, it’s more complicated.

Because everything is done on a total population basis, we need allocate the whole population including those on neither roll: children, non-permanent residents, people who have not gotten around to filling in the paperwork. To solve for this, we pro-rate everyone of Māori descent based on the ratio of Māori roll voters to general roll Māori voters.

10.4% of the total population is counted in the Māori electoral population. This is a product of:

• The 20% of the population identifying as of Māori descent in the Census (this is a separate question on the Census to identifying as of Māori ethnicity, slightly lower at 18%); multiplied by

• The 52% of Māori enrolees who chose to be on the Māori roll (as of 1 April 2024).

The remaining 89.6% of the population is in the general electoral population. This is the sum of:

• The 80% who do not identify as of Māori descent; plus

• 9.6% from people of Māori descent allocated to the general electoral population, based on Māori enrolled on the general roll.

This pro-rating is repeated at smaller geographic scales (electorates, area units) when working out things like whether an electorate is within tolerance or when redrawing boundaries.

It’s worth noting that there is an issue where the Census consistently under-counts Māori due to non-response bias. This and the timeliness and overall accuracy of the Census are behind proposals to allow a wider set of input data for the population counts.

North and South Island general populations

Because the South Island is treated separately in working out the number of general seats (general seats cannot straddle the main islands) we next need to work out the South Island and North Island general electoral populations. This is simple* counting.

Random digression: the population of Rēkohu/the Chatham Islands is counted on the South Island for general purposes but the North Island for Māori purposes. This seems backwards given they are in a North Island general electorate and a (mostly) Te Waiponamu Māori one.

South Island seats and quota

Representation for the South Island is fixed by law at 16 seats. This is due to concerns about the North Island growing faster than the South, and ensuring electorates on the Mainland do not grow too large. Were it not for this requirement, the number would have fallen to 14 since 1996, given relative growth.

This creates its own complications/problems, but that will have to wait for a future post. For now, the process for working out South Island general seats is simple: divide the population by 16. This gives 70,037 people on average in each seat.

North Island and Māori seats and quota

To keep things proportional, the number of general seats in the North Island and the number of Māori seats is determined by the size of the South Island seats.

This gives:

• 47.89 North Island seats, rounded up to 48; and

• 7.42 Māori seats, rounded down to 7.

This rounding does lead to the North Island being very slightly over-represented and the Māori seats being under-represented, but there is really no way around this.

With 0.42 of a seat “missing”, the disparity for Māori seats is about as bad as it can be. It would only have taken a further 3,500 enrolees (or 0.6% of total Māori on both rolls) switching from the general roll to the Māori one before 1 April 2024 to increase the fractional number of seats to 7.5. This would then be rounded up to 8 (and would then flip to over-representation, but again there is not much to be done about this).

This compounds the abovementioned issue of under-counting the Māori population.

(Before you ask: no, it’s too late. While voters can switch between the Māori and general rolls any time, doing so won’t affect the boundaries until after 2029).

Total seats

All that maths leaves us with:

• 16 South Island general seats

• 48 North Island general seats

• 7 Māori seats

Add to this 49 list seats for a 120 seat Parliament (assuming no overhangs that increase the number of list seats when they occur).

Easy.

Part 2 – Patterns of change and quota deviation

In the second half of this piece, we will zoom in on patterns of population change from a nationwide view, through a broad regional view, and finish up with a quick summary at the electorate level.

The macro view – Islands and Māori seats

Population growth

The seat numbers calculated above represent a loss of one general seat in the North Island. You can see below the population growth – and more importantly the relative population growth – that has lead to this outcome. While the North Island did grow by +5.5%, the South Island and the Māori electoral populations grew by more, so in relative terms the North shrunk by -0.8%. Given the way the seat maths works, it’s this relative growth that matters.

The decision of which seat to ditch is one of the big decisions the Representation Commission will need to make (and something we will get to later in the week).

Quota and tolerances

The seat math also gives us another important number for the boundary adjustment process: the quota. The quota is the average size of a seat in each of the three groups (North, South, and Māori). Every seat is required to be within 5% of this number. If an existing seat is over or under this ‘tolerance’, then the Commission is required to shift the seat’s boundaries.

In addition to the Census population, StatsNZ also produce forecasts for the two elections covered by these boundaries (2026 and 2029). These forecasts highlight some of the weirdness mentioned above: by 2026, the average North Island electorate is already forecast to be larger than the permissible maximum. By 2029 the same will also be true of the Māori seats.

Put another way by the time we actually get around to having election using the redrawn electorates, relative population growth in the North Island would (on paper) necessitate adding back the electorate we are getting rid of to maintain parity with the South. In the end it does not matter all that much (the party vote preserves overall proportionality in an MMP system) but it’s still an odd disruption.

These differential population growth dynamics also apply at a regional and electorate level and will (combined with the need to lose a seat) drive the boundary adjustments.

The meso view – regional dynamics

To understand the growth and boundary change dynamics, we have grouped up the general and Māori seats into 16 “regions”. The groupings are somewhat arbitrary but mainly reflect the physical barriers that divide the country: mountain ranges like the Main Divide, pinch-points like the Auckland Isthmus, or depopulated areas of the country like the southern King Country.

The Commission has a statutory obligation to consider these barriers (among other things) when redrawing boundaries and tends to respect them where they can.

Regional growth

The chart below shows growth from 2018 to 2023 on a regional basis – again absolute and relative.

Focusing on relative population growth, the main takeaways are:

• The decline of inner Auckland, Wellington, and the Central North Island (absolute decline, in the case of the Auckland Isthmus)

• The growth of Northland and North Auckland, the inside of the Golden Triangle, Rural Canterbury, and the Māori seats

• The stability of the North Island’s East Coast and the South Island’s West Coast

The consistent pattern across most of these regions is growth on the rural/urban fringe compared to stagnation or even decline in established inner-urban cores and deeper rural areas. Relative rural decline is not new, but the inner-urban decline is. (There is a bigger story there about urban planning regulations, but I am not the person to write it).

Quotas

The population figures allow us to compare the number of seats a region currently has to the number it “should” have based on population.

The pattern of this deviation more or less repeats the pattern in relative growth (with differences due to seats already being under or over quota in 2020).

• The six seats on the Auckland Isthmus between them are missing 0.7 of a seat

• Greater Wellington is down 0.4 of one

• The Central North Island is 0.2 short

• Conversely, the Waikato and Rural Canterbury are 0.2 over each

Despite growth in the Māori electoral population overall, because this growth was fairly even across the motu, on a regional basis they’re close to target.

The micro view - electorates

Finally we will look at an electorate level. In some cases, the growth pattern is consistent across the region. In others, the changes at a regional level are being driven by one or two big movements.

Northland (+0.09 of a seat) Both the Northland electorate (+0.07) and Whangārei (+0.02) are over quota, but Northland more so.

North Shore (+0.06) Overall growth, but highly bifurcated. The two urban fringe seats (Kaipara ki Mahurangi +0.15, Whangaparāoa +0.06) are significantly over, while the more established urban areas are all under (get used to this pattern)

West Auckland (-0.03) Similar pattern to the North Shore, with the urban fringe in Upper Harbour (+0.08) over and the rest under or flat.

Auckland Isthmus (-0.66) Probably the most significant change in the whole country: the Isthmus is two-thirds of a seat short, and every electorate is not only below target, but outside the -5% tolerance, with 5/6 being more than 10% under. Basically the same pattern as the rest of Auckland, but with no fringe suburbs providing growth.

South-East Auckland (+0.01) While the area as a whole is basically on target, that growth is concentrated in two seats: Takanini (+0.18, the seat introduced in 2020 when we gained one, and the most over-quota in the whole country) and Papakura (+0.06).

Waikato (+0.16) Broad-based growth, with all five seats over quota. Again, growth is concentrated on the fringes: Auckland’s southern edge in the case of Port Waikato (+0.05), Greater Hamilton in the case of Waikato (+0.03) and Taupō (+0.05, despite its name this seat extends to cover Cambridge and the growth corridor along the Waikato Expressway).

Bay of Plenty (+0.01) Rounding out the Golden Triangle exurban growth pattern, the Bay of Plenty region is flat overall, with growth in the Bay of Plenty (+0.08) electorate (somewhat misleadingly named as at this point the seat is just the outer suburbs of Tauranga) and Coromandel (+0.01, noting that Coromandel extends as far south as Tauranga’s northern suburbs) offset by shrinkage in Rotorua (-0.08) and Tauranga proper (-0.01, the first time since 1996 that it’s under).

East North Island (-0.05) On the eastern side of the North Island’s main divide* things are mostly stable, except for Napier (-0.05).

*Mostly. To deal with a long-term relative decline in Tairāwhiti’s population East Coast has been extended around the Raukumaras to cover the whole Eastern Bay of Plenty.

Central North Island (-0.19) The second of three areas with a significant deficit. All the seats are under, except for New Plymouth (+0.02) and in Palmerston North’s case significantly so (-0.10). Part of an overall complex of regional decline in the lower North along with…

Wellington (-0.39) Greater Wellington replicates the overall pattern of relative urban decline and exurban growth, but unlike the Golden Triangle in an (relative) shrinking context. The three seats in Wellington City proper – Ōhāriu (-0.08), Wellington Central (-0.10) and Rongotai (-0.12) – have seen the sharpest contractions, while the outer areas on the Kāpiti Coast and up the Hutt Valley have seen more modest declines.

West South Island (-0.06) Crossing the Cook Strait (while we still can given the state of the ferry fleet amirightfolks) both Nelson (-0.03) and West Coast-Tasman (-0.03) have seen similar levels of relative decline. Given the firmness of the Southern Alps as a boundary, it’s probably for the best this decline was not greater.

Rural Canterbury (+0.15) Really a story of exurban Christchurch, with Selwyn (+0.14, the only seat in the South Island over the +0.05 tolerance) and Waimakariri (+0.04) making up most of the growth.

Christchurch (-0.01) Really not much to say (and for the Commission not all that much to do). Bucking the overall trend we see in the other main centres, all Christchurch’s seats are within tolerance.

Otago-Southland (-0.08) A two-fold pattern: decline in the established coastal areas (Dunedin -0.03, Invercargill -0.05, and Taieri -0.04) and growth in the inland areas around Queenstown-Lakes (Southland +0.01 and Waitaki +0.04). Both Southland and Waitaki combine a mix of high-growth Central Otago areas (Queenstown and the Upper Clutha respectively) with declining rural areas (Southland and Gore districts, or Waitaki and Waimate districts).

Northern Māori (-0.01) Like it’s tauiwi cousins in the Auckland general seats, Tāmaki-Makaurau (-0.046, just squeaking within tolerance) sees its relative losses offset by growth to its north (Te Tai Tokerau, +0.02) and south (Hauraki-Waikato, +0.01).

Southern Māori (0.0) Relative losses here are concentrated in Ikaroa-Rāwhiti (-0.05, the only Māori seat out of tolerance) with relative growth in Te Tai Tonga (+0.04). Again roughly mirroring regional patterns in the general seats.

In conclusion

*phew*

One day I will write something shorter than 2000 words I swear. A bit of content to get through, but necessary to set the scene for the changes we are likely to see.

Having laid out the problem, we’ll be back later this week to preview possible solutions.

Ka kite (actually, this time) anō

Kia ora, happy new year, and welcome back.

We’re using the start of the new (political) year as an opportunity to look back on the polling performance of the Sixth National Government so far, how it compares to past administrations, and where things might go.

The tl;dr version

• A little under halfway through the term, the government parties have a combined +0.8% lead over the opposition.

• This is below average compared to all government terms this far in (2.0%) and well below the average for a first term government at this point (6.2%).

• The two first-term governments who were performing similarly (Kirk and Bolger) also faced significant economic challenges.

• One went on to lose, the other to (very narrowly) win.

• Recoveries are possible, but second terms are not a given – even for National.

Halfway(ish) through

Nearly halfway through their term, National and its support parties hold a narrow polling lead over Labour and the other opposition parties. National sit on 31.7%, ever-so-slightly ahead of Labour on 31.2%. Coalition partners ACT (9.5%) and NZ First (7.3%) combine with National to put the government on 48.5%. Combining Labour’s numbers with their likely support partners the Greens (11.2%) and Te Pāti Māori (5.3%) gives the opposition 47.7%. On net, this is good for a 0.8% government lead.

How we got here

After winning the 2023 election with an above-average 11% margin, Luxon and co initially maintained a similar lead in the polls. From about the middle of 2024, this dropped back to around 5%. However, through the back end of 2024 and into the new year this has decayed to just the abovementioned 0.8%, with the opposition even taking a lead in some polls.

On the face of it this is… not great. To be secure in re-election, you would want to be doing better than a statistical dead heat. At the same time, the government would still cling to power on these numbers and there’s a while to go before the next election, so we shouldn’t over-read these trends.

But this doesn’t exactly feel like a freshly-pressed government basking in the long summer days of a waning honeymoon. As we’ll see below, if anything it resembles either a second term beginning to stall out and go stale, or more to the point one of a couple of seemingly-doomed first term government struggling to get traction amid economic chaos.

Some history

But how bad is it really? To put the sixth National government’s performance in perspective, we can compare it to past governments at approximately the same point in the electoral cycle.

(This analysis is deeply indebted to the diligent work of Patrick Leyland of The Progress Report for compiling polling data prior to 2002 – a massively invaluable resource for anyone doing political science in Aotearoa).

This chart shows a rolling average of the government’s polling lead over the opposition for each of the electoral cycles for which we have data. A positive figure shows the government leading, a negative figure shows them trailing. The dots at either end show the result of the elections.

From 1990 onwards (as the FPP system began to break down and was replaced by MMP) it also shows the difference between the left and right blocs (in paler red/blue).

The dashed lines highlight where past governments were sixteen months into their terms: the same point the Luxon administration is at now. Zooming in on this point in time we can get a sense of how today compares to various yesterdays.

While below the all cycles average of 2%, Luxon’s position is by no means disastrous: especially compared to the pits of popularity that governments of both hues experienced in the 1980s and 1990s. He’s no John Key (we knew that already) but he’s also not second-term Lange.

However, a government’s popularity tends to decline over its lifespan. Isolate just first terms, and things look worse for the current government.

Most first-term governments experience an extended honeymoon (+6.2% on average). While not as commanding as Clark or Key’s double-digit leads, Muldoon, Lange, and Ardern all retained reasonable goodwill and went on to second terms.

And it’s worth noting that this isn’t a case of a narrow initial win putting the current government off to a bad start: the 11% margin at the 2023 election is above the historic average of 8%. Both the current level and the declining trend to get there are out of step with the historical norm.

Of the first terms for which we have data, the two who were performing like the current government were Norman Kirk’s third Labour government in the mid-1970s, and Jim Bolger’s fourth National government in the turbulent early 1990s.

A departure from the historical (Big) Norm

Kirk’s government had taken power with a respectable if not earth-shattering +7% win over Jack Marshall’s National (who were at that point four terms deep into a Natural-Party-of-Government imperium).

After a year of governing, the shocks of the 1973 oil crisis and Britain’s accession to the European Economic Community helped reduced this to little better than an effective tie with the as-yet still Marshall-led National. Subsequent events – Robert Muldoon’s takeover and Kirk’s near co-incident death – would see Labour go on to defeat under replacement leader Bill Rowling.

The Mother of All Comebacks

By contrast, Bolger had won in a +13% landslide over world-historic hospital pass recipient Mike Moore. But then the aftermath of the Mother of All Budgets, the supertax reversal, and sundry other economic crises saw a complete collapse to -11% behind Labour (an almost difficult to comprehend -24pp reversal.) Factor in the rising popularity of the nascent Alliance, and National were a catastrophic -28% behind the combined opposition. Moore – whose position Hipkins in some ways resembles – looked certain to return to power.

But then he didn’t. Unlike Rowling and thanks to the split on the left in the final FPP election, Bolger was able to claw back by the barest of margins (literally a zero-seat majority before a Labour dissident took the speaker’s chair) and held on in 1993.

… but it rhymes

The common thread between these two cases – one that is shared with the current government – is economic strife. The current recession (so far) is yet to reach the retching depths of the 1970s and 1990s, but doubtless Luxon’s fortunes will be influenced by the recovery (or lack thereof) of the economy.

Luxon’s situation looks to be significantly better than Kirk/Rowling’s: Kirk’s death was an unaccountable tragedy, and the opposition seem unlikely to replace its previously-defeated leader with a pugnacious bruiser in the Muldoon mould. Disquieted as the global trade situation is given the Trump administration, it’s not quite dealing us the bodyblows that the loss of the British market to the EEC did.

As spelled out in greater depth by Henry Cooke, there are closer similarities to Bolger’s position. Like Moore, Hipkins has stayed on. The left is again displaying fissile tendencies (the Greens’ support remains stable and Te Pāti Māori’s has risen on the back of opposition to the Treaty Principles bill) but this won’t help in an MMP contest. Somehow – defying the logic of actuarial tables and the space-time continuum – like Bolger Luxon faces the risk of Winston Peter’s acting out. Most importantly though and despite the historical rhymes, Luxon has a significantly smaller hill to climb than Bolger did.

Early on-set second term-itis?

The government’s performance thus far might be middling compared to most first terms, but it is better than the truly dire examples Kirk and Bolger provide. If you squint a bit, the current slide to a narrow lead does not look out of place beside governments in their second terms.

Muldoon was 5 points ahead, although this may be a case of the polls flattering him given he undershot them by about that much in both 1978 and 1981.

At this point in his second term Bolger sat -7% behind the combined strength of Labour and the Alliance – which by his standards was a massive improvement: an improvement that would continue in the run up to 1996.

Lange and Ardern were both half-way through precipitous flop eras that would end in eventual replacement and defeat. Clark too was beginning her Ōrewa-induced slide – albeit one her government would recover from.

Even the seemingly invincible John Key was struggling at this point in his second term. While National held a respectable lead over Labour, its lack of dance partners briefly put it on level terms with Labour and the Greens combined.

All these National governments went on to a third term while only one of the Labour ones did.

Always remember that you are mortal

It’s tempting to read this as reinforcing National’s “natural party of government” invincibility. I think that’s a mistake. While true over the Holland and Holyoake post-War boom years (largely outside the scope of this piece given the lack of polling data), between 1972 and 2005 National and its allies only won the popular vote 3/12 times (1975, 1990, and 1996).

Despite an initial burst of popularity Muldoon would go on to lose the popular vote twice: propped up by FPP and – if you want to describe it charitably – a favourable electorate redistribution in 1977 (gerrymandering, if you want to be less charitable).

Bolger’s second majority was jerry-rigged together with duct-tape (speakership deals) and bits of string he found behind the couch (the United party rebels, Ross Meurant).

The Key years look like a restoration, but that can equally be viewed as part of the stable three-term turnabout ushered in by MMP and the benign macroeconomic environment at the end of history. This context also benefited Labour in government. The nine-year pattern – as previously noted and subsequently validated by Labour’s 2023 loss – has begun to fall apart.

Where to from here?

If you’re a government supporter looking for cope:

• Things aren’t as bad as they could be (your guy isn’t 30 points down).

• Governments have come back from worse (the guy who was won).

Conversely, if you’re an opposition supporter looking for cope:

• This is pretty bad for a first term government.

• The previous two times a defeated Labour PM stuck around, the opposition won the popular vote.

Finally if you’re a pure and objective centrist (or just generally checked out and indifferent):

• The polls as they stand are finely balanced.

• Assumptions about natural parties of government or stable multi-term cycles are outdated.

I’m conscious that this was a very long-winded way to say “National could win in 2026 but on the other hand they could lose”. Hopefully this helps guard against conventional wisdoms or confident predictions.

The uncertainty is the point.

Ka kite (hopefully) anō.

As set out in yesterday’s final polling update, big-picture the election doesn’t look especially in doubt. However:

• The analysis based on past polling error could be wrong: 2023 could be a 3σ event, maybe there has been a structural break in the reliability of polls.

• There is still a plausible-but-narrow path to victory for Labour even within this model.

• Even if the model’s central estimate is right, the path way from final polling to final results will be bumpy.

With that in mind, let’s take a look at how results could evolve tomorrow based on how they’ve evolved in the past. We’ll take it step-by-step from polling to the final results.

Polling error

Polls can be wrong, they often are. As discussed in the last post this uncertainty is what gives Labour any path to power.

There are some patterns in polling error. In short:

• Labour and NZ First lift a bit (~1.4% and ~0.3% on average)

• National and the Greens drop a bit (~-1.3% and ~-1.1%)

If those patterns hold, it’s possible (although I can’t stress this enough – not guaranteed) that Labour will do a bit better than expected, especially early in the evening.

Advance votes

Trends in advance voting

Advance voting (voting before election day) has been growing in popularity since the rules around it were liberalised in 2011. The numbers so far have been down versus 2020, but modestly up versus 2017.

A note here that this excludes special votes taken ahead of election day, which will be included with special votes in the count.

This could be due to a combination of two factors:

• Advance votes as a proportion of total votes are returning to trend, and on-day vote will be proportionally higher; or

• Advance vote as a proportion of total votes will be in line with 2020, with a significant decrease in turnout overall.

Before we know the numbers on election day, it’s impossible to say for sure.

As a share of all votes, advance votes reached a high of 58% in 2020. But even under the somewhat more normal conditions of 2017, they had reached 38%. We’ll find out tomorrow whether this trend persists, but I’ll take a completely baseless punt and say it’ll be closer to 48%.

Partisan lean of advance voting

Over time, the partisan lean of the advance vote has reversed: ACT, National, and other conservative parties used to do better, Labour used to do worse.

As noted above, before 2011 you had to have a reason for advance voting. Possible reasons associate with age (like surgery) or economic status (overseas travel). These factors tend to correlate with conservative voting, so help explain the pre-2011 trend.

Since liberalisation, the early vote initially shifted to mirroring the on-day vote. But in 2017 and even more so in 2020, the early vote has favoured the left. Again, a combination of two things could be going on here:

• The various demographics of voters who vote early could differ from voters who vote on-day in a meaningful way that correlates with partisan (eg: young vs old).

• A late swing over the course of the campaign could shift voter preferences – as in someone who voted early would have changed their minds had they voted on election day.

It could be possible to tease these factors apart with further research, but I’m not aware of any.

What it could mean

At any rate, this effect could make for a dramatic few hours between 7pm and midnight Saturday. The advance vote is counted over the course of election day, and released fairly promptly after polls close. On-day votes come in over the course of the evening.

It’s difficult to say with certainty what this means in partisan terms, but if 2017 and 2020 replicate, then an initial bout of excitement for the left and disappointment for the right could slowly ebb as the night goes on.

There’s a theory that because small polling places get counted quicker, and these tend to be in more rural areas, that National do better in the early returns for on day voting, but I can’t find any real evidence for that from polling-place level data (slightly shittier chart done in a hurry for 2020 shown below).

Special votes

Trends in special votes

As shown above, special votes were a relatively static ~11-13% from 1996 to 2014. In 2017 and 2020, they have increased to around 17%. In 2020, this was in part due to on-day enrolment for special voting being legalised. You might have thought this could lead to an increase even over and above 2017 levels, but it didn’t seem to.

Partisan lean of special votes

The partisanship of special votes shows a much clearer pattern: since 1999, progressive parties generally (the Greens, Labour, Te Pāti Māori, Mana, and TOP) do significantly better overall (by ~10% combined).

In 2017 and 2020, this has favoured Labour more than it has the Greens.

In terms of the difference the specials can make, for a while the consistent pattern was the Greens gaining a seat usually at the expense of National. With the rise in special voting in 2017 and 2020, there has been a two-seat switch.

(This was done based on re-rerunning a Sainte-Laguë calculation for the ordinary and post-special vote, and may be subject to errors. I couldn’t find reliable records of the change given time constraints).

What it could mean

If results on the night are close with the left narrowly behind, this consistent trend could lead to a nervous few weeks while special votes are counted and validated.

How things could evolve

So, bringing all that noise together, a very uncertain and not guaranteed view of what could happen over the next two weeks.

When the advance vote drops

The advance vote released in a big bag after 7pm could look good for Labour vs their polling, a combination of their tendency to over-perform at least a little and the lean of the advance vote. Call it conservatively +2% up.

• National conversely could look a bit worse given the reverse tendence. Again very ball park -2% down.

• Same for ACT but given they’re a much smaller party, call it -1%.

• The Greens and NZ First could be about even, as their two effects (polling and early vote) offset.

• Te Pāti Māori maybe even, given limited swings.

As the night proceeds

• Labour will drop a bit if the on-day vote is more about demographics than swing in the final two weeks. Conversely, it’s more about a swing at the end, then maybe they’ll go up. Same effect for the Greens.

• Conversely, if the shift is about demographics National and ACT will go up. If it’s about swing they might go down. If it’s about both it could be flat.

• NZ First will probably start improving regardless.

If that thing about small polling places is true (somewhat doubtful), then it’ll go better for National early on, then edge back a little bit as big urban booths drop.

In any event, the magnitude of this swing could be larger than in 2020, as on-day votes should be a larger % of the total vote.

The longest two weeks of the year

Individual candidates in close races (more on that later this evening) will have an especially nervous wait, but the shape of Parliament overall should be relatively settled. Odds on National (maybe ACT or NZF) lose 1-2 seats combined, maybe 3 in a screaming fit but don’t count on it. Who among the left picks them up will depend.

To sum up

This is all largely a structured reckon to temper expectations when everything goes live.

Just because Labour might look like they have the horses at 7pm does not mean it’ll end that way. Similarly even if National have them by Sunday morning, if they’re only 2 seats up (from either not needing NZ First, or in an extreme case of not having the numbers even with them) then fuck me, what a fortnight.

Anyway, remember to breath and drink plenty of water.

Subscribe now

The big picture

In terms of the latest poll tracking, National (35%) hold a substantial lead over Labour (27%). Figures for the Greens (14%) and TPM (2.5%) give the left a combined 44%. ACT (9%) and NZ First (7%) give the right a combined 51%, for a 7% lead.

TOP are on 2.4%, other right parties combined 1.4%.

Trends

Labour reverse some recent losses in the final stretch (+1.4). Green gains continue (+1.5). NZ First recover (+2.1). National give up some recent gains (-2.1) while ACT's slide continues (-1.8). TOP up (+1.4). TPM and the miscellaneous right stable.

The overall polling phase over October has been 🟢<🔴=⚫<🔵<🟡 or "leftwards" compared to September’s 🟢<🔴⚫>🔵<🟡 "towards National and the Greens" phase.

Breaking down polling by firm, there’s a reasonable consensus among firms. Roy Morgan’s September numbers are still not available, and given the delay may not be useful anyway.

Seats in the house

NZ First’s recovery to over the threshold combined with National and ACT’s relative decline means NZF will be necessary to form a government. Combined the right have 66 seats, the left 55.

The death of the ACT candidate for Port Waikato also means National will almost certainly have an overhang seat once the byelection is completed, leading to a 121 seats house at minimum.

There is also a possibility Te Pāti Māori create an overhang, but we have not modelled this.

Were TOP to win Ilam, they would have three seats, one each from National, the Greens, and ACT.

Performance relative to polling

Polling error is impossible to predict with any accuracy. An error in the previous election does not mean there will be one in the same direction this time.

However, there are fairly consistent patterns.

• National and the Greens underperform

• Labour and NZ First overperform

On an individual party level that averages and 95th percentile ranges are:

🟡ACT: -0.3% (-2.2% to +1.6%)

🟢Gre: -1.0% (-2.9% to +0.9%)

🔴Lab: +1.4% (-2.3% to +5.0%)

🟠TPM: +0.2% (-0.2% to +0.6%)

🔵Nat: -1.3% (-4.8% to +2.2%)

⚫NZF: +0.27% (-1.7% to +2.2%)

We can use these tendencies and the uncertainty associated with them to estimate the range of plausible outcomes once votes are counted.

Election day outcomes

Top line numbers - National highly favoured

The move to a likely 121 seat house means hung outcomes (ignoring other overhangs and Ilam) are impossible. Otherwise the overall position points overwhelmingly to a National victory, even accounting for a Labour’s tendency to over-perform their polls.

Shape of coalitions - NZ First required

Much more significant shift in the likely government formation. With NZ First recovering enough to make their odds of entering house ~97%, they will almost certainly be required to govern.

Labour’s paths

Labour's chances come from two situations.

1. Where NZ First miss out (3%), they have the numbers 61% of the time, or 2% of all cases.

2. Where NZ First make it in (97%) they have the numbers only 3% of the time, or 3% of all cases.

The sum of these situations gives their 5% odds. A narrow path, but about the odds the bookies were giving of Fiji beating Australia at the men’s rugby world cup so…

If Ilam…

An addendum on the effect TOP winning Ilam could have. (The electorate and nationwide models are not integrated so can't be run together). Were Manji to win, TOP would have the balance of power 17% of the time. This lowers Labour's chances to 1%, and National's to 82%.

Latest estimate of Manji's chances were approximately 18%, so combined they have a 3% chance of being able to choose between National's bloc and Labour's.

That’s all for today. Will be sending around a final update on seats and a “what to expect” primer about the swings from polling, through early votes, votes on the night, and specials.

Pakuranga

🔵Safe National hold

Candidate vote

🔵Nat – Brown (list 9th) 70% – odds 100%

🔴Lab – Henry (61st) 22% – 0%

🟡ACT – Parmar (9th) 3% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 52

🔴Lab 16

🟡ACT 13

🟢Gre 7

⚫NZF 6

🟣Con 3

Commentary

National’s Simeon Brown did the best of any National candidate with Labour voters in 2020, with a full 25% of them. Even with a slight expected decline in this figure as many of those voters return to National, he will easily retain the seat.

On the list, Labour’s Nerissa Henry will not be elected at 61^st, but ACT candidate (and former National MP) Parmjeet Parmar should be.

Palmerston North

🔴Safe Labour hold

Candidate vote

🔴Lab – Utikere (list 25th) 52% – odds 100%

🔵Nat – Bansal (52nd) 36% – 0%

🟢Gre – Tuiono (5th) 6% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 34

🔵Nat 30

🟢Gre 11

🟡ACT 9

⚫NZF 8

🔝TOP 4

Commentary

Labour candidate Tangi Utikere was the best-performing newcomer among Labour candidates in 2020, attracting 8% of votes from National voters, although some of this may be due to controversies surrounding then National candidate William Wood.

In 2023, now as an incumbent and with former Labour voters moving to National on the party vote, it’s highly likely he will improve on that figure.

National’s Ankit Bansal is not in an electable position at 52^nd on the list. Green candidate Tenau Tuiono is at 5^th.

Panmure-Ōtāhuhu

🔴Safe Labour hold

Candidate vote

🔴Lab – Salesa (list 24th) 56% – odds 100%

🟢Gre – Collins (11^th) 22% – 0%

🔵Nat – Randhawa (46th) 15% – 0%

🟡ACT – Modkova (12^th) 3% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 64

🔵Nat 14

⚫NZF 6

🟢Gre 5

🟡ACT 4

🔣Oth 3

Commentary

A member of the M-block in spirit if no longer in name, Panmure-Ōtāhuhu (FKA: Manukau East) remains a safe Labour seat for incumbent Jenny Salesa, despite a two-tick challenge from the Greens’ Efeso Collins (former local councillor and unsuccessful 2022 mayoral candidate).

Applying the same assumptions as other Green two-tick campaigns, there are simply not enough Green voters here for Collins to have a realistic chance, even if he over-performs with Labour voters. At any rate, at 11^th on the list he is safe on current polling.

A more interesting question for the Greens here is whether Collins’ campaign and profile in the community will be enough to lift the party’s party vote performance in an electorate they have previously struggled in.

National candidate Navtej Randhawa is at an unelectable 46^th on the list, but ACT candidate Antonia Modkova is in a probably but not guaranteed 12^th.

Papakura

🔵Safe National hold

Candidate vote

🔵Nat – Collins (list 10th) 54% – odds 100%

🔴Lab – Kanongata'a (34th) 30% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 41

🔴Lab 27

🟡ACT 12

⚫NZF 8

🟢Gre 5

🟣Con 2

Commentary

Then National leader Judith Collins won this exurban Auckland seat with a safe 14% margin in 2020. Despite the loss of a “favourite daughter” effect now that she’s left the leadership, she will still be safely returned in 2023.

At 34^th on the party list, Labour candidate Anahila Kanongata'a will not see a return to Parliament.

Port Waikato

🔵Safe National hold

Candidates

🔵Nat – Bayly (list 15th)

🔴Lab – Keel (65th)

⚫NZF – Costello (3rd)

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 52

🟡ACT 15

🔴Lab 12

⚫NZF 8

🟢Gre 6

🟣Con 3

Commentary

With the sad news of ACT candidate Neil Christensen passing away, the candidate vote in Port Waikato will not be counted, and instead a byelection will be held following the general election.

The seat is safely National’s in any election (by or otherwise). The absence of 2020 Heartland candidate Mark Ball (who drew an impressive 21%) makes Bayly’s job significantly easier.

On the list (the only way candidates here can now qualify) Bayly is safe at 15^th, as is NZ First candidate Casey Costello at 3^rd. At 65^th, Labour’s Gwendoline Keel will not make it to Parliament.

Rangitata

🔵Safe National GAIN

Candidate vote

🔵Nat – Meager (list n/a) 59% – odds 100%

🔴Lab – Luxton (19th) 29% – 0%

History

🔴🔴🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 46

🔴Lab 21

🟡ACT 12

🟢Gre 8

⚫NZF 6

🔝TOP 2

Commentary

One of the signal victories for Labour in 2020, with Jo Luxton pulling of a sound 11% win over then-National candidate Megan Hands (who took over late in the piece from disgraced incumbent Andrew Faloon) Rangitata will be safely returning to National in 2023.

National’s James Meager will benefit from the expected net-41% swing on the party vote, and while Luxton will benefit somewhat from incumbency, it won’t be enough to overcome the disadvantage.

On current polling Luxton will not make it on the list at 19^th, but could if other electorate results go against her colleagues in lower list spots.

Rangitīkei

🔵Safe National hold

Candidate vote

🔵Nat – Redmayne (list 21st) 55% – odds 100%

🔴Lab – Butt (n/a) 24% – 0%

🟡ACT – Hoggard (5th) 5% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 41

🔴Lab 18

🟡ACT 14

⚫NZF 9

🟢Gre 9

🔝TOP 3

Commentary

A dramatic moment early on election night showed then Labour candidate Soraya Peke-Mason ahead of National incumbent Ian McKelvie on the advance vote. The swing back to National once the on-day vote was counted eventually left McKelvie with a modest 7% lead.

This won’t be repeated in 2023. With McKelvie retiring and Peke-Mason moving to Te Tai Hauāuru, National’s Suze Redmayne will safely win the open contest.

ACT candidate and former President of Federated Farmers Andrew Hoggard will probably put a dent in Redmaye’s figures in this deeply rural seat, but without a formal two-tick run he won’t come close enough to affect the outcome. Ranked 5^th, he’ll be elected from the list.

Remutaka

🔴Safe Labour hold

Candidate vote

🔴Lab – Hipkins (list 1st) 56% – odds 100%

🔵Nat – Chatterton (27th) 29% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 35

🔵Nat 29

🟢Gre 11

⚫NZF 9

🟡ACT 9

🔝TOP 4

Commentary

Then-COVID Response Minister now-Prime Minister Chris Hipkins was secondly only to then-Prime Minister now-Dame Jacinda Ardern in terms of cross-partisan support in 2020, pulling in 14% of National voters (on top of all the swing voters Labour pulled over).

As party leader, he can be expected to benefit from the “favourite son” premium, and running in a deep red seat, will be re-elected despite the overall swing back to National.

National candidate Emma Chatterton will currently be the first one to miss out on the list, but a swing of a few basis points or some unsuccessful candidates further down the list could see her in at 27^th.

Rongotai

🟢Toss-up Green GAIN

Candidate vote

🟢Gre – Genter (list 4th) 38% – odds 54%

🔴Lab – Fitzsimons (56th) 37% – 46%

🔵Nat – Muthu (51st) 21% – 0.07%

🟡ACT – McKee (3rd) 1% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 33

🟢Gre 32

🔵Nat 17

🔝TOP 6

🟡ACT 5

⚫NZF 4

Commentary

When Julie Anne Genter and the Greens decided to launch an attempt on Rongotai in late 2022 following the party’s success the Wellington local elections, it looked like a long-shot but a respectable one. With events since then, Genter is *just* favoured to win.

First, incumbent Paul Eagle decided to retire following his unsuccessful mayoral bid (although whether this helped or harmed Labour’s chances is unknowable). Then Labour fell 10% in the polls and the Greens climbed 5%.

However, with no real precedent for a Labour vs Green race Labour candidate the contest with former councillor Fleur Fitzsimons will be incredibly close. National’s Muthu should be a distant third in any case, but in 1/1500 scenarios managed to perfectly split the difference.

As above for the other Green two-tick races, we’ve assumed Green voters are slightly more loyal to their candidates than Labour voters are to theirs (and artificially lowered the Lab>Green numbers out of an abundance of caution).

The specific split assumptions in this case is are (L/G/N):

• 🔴Lab voters 78/17/1

• 🟢Gre voters 18/81/1

• 🔵Nat voters 10/5/81

Nationwide polling and the seats party-vote lean of Gre+20 and Lab+7 leaves Labour and the Greens expected to be essentially level on the party vote (at 38 and 37 respectively). So while this is a hack point generally: turnout and ground game really will make the difference.

Rotorua

🔵Safe National hold

Candidate vote

🔵Nat – McClay (list 12th) 58% – odds 100%

🔴Lab – Sandford (69th) 32% – 0%

🟠TPM – Raukawa-Tait (8th) 2% – 0%

History

🔵🔴🔴🔴🔵🔵🔵🔵🔵

Party vote

🔵Nat 40

🔴Lab 21

🟡ACT 12

⚫NZF 10

🟢Gre 8

🔝TOP 3

Commentary

A very close contest in 2020, National incumbent Todd McLay was only able to hang on by 2% off the back of a sizeable support from Labour voters.

With the swing back, the result here is not in doubt. And at 69th on the list, Labour's Sandford has no chance there either.

Of note is TPM candidate Merepeka Raukawa-Tait. She is not in serious contention for the seat or on the list at 8th, but the seat is one of TPM's best general seats, so a run makes sense in terms of lifting the party vote.

Selwyn

🔵Safe National hold

Candidate vote

🔵Nat – Grigg (list 19th) 65% – odds 100%

🔴Lab – Jones (74th) 27% – 0%

🟡ACT – Harvey (15th) 6% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 47

🟡ACT 16

🔴Lab 14

🟢Gre 12

⚫NZF 5

🔝TOP 3

Commentary

As blue as it gets, the race in Selwyn isn't that exciting, with National's Grigg coasting home.

At 74th on Labour's list, Labour's Luke Jones will not be elected.

ACT candidate Ben Harvey also misses out at 15^th on current numbers, but could be in if the party's vote turns around.

Southland

🔵Safe National hold

Candidate vote

🔵Nat – Mooney (list 53rd) 62% – odds 100%

🔴Lab – McCallum (70th) 21% – 0%

🟡ACT – Stephenson (4th) 4% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 49

🟡ACT 15

🔴Lab 13

🟢Gre 10

⚫NZF 6

🔝TOP 3

Commentary

After a run of troubled candidates, Southland finally sees some stability with returning champion Joeseph Mooney.

At 4^th on ACT’s list, newcomer Todd Stephenson will be joining him. At 70^th, Labour candidate Simon McCallum will not be.

Taieri

🔴Solid Labour hold

Candidate vote

🔴Lab – Leary (list 52nd) 45% – odds 95%

🔵Nat – French (49th) 37% – 5%

🟢Gre – Willis (12th) 8% – 0%

⚫NZF – Patterson (4th) 4% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔵Nat 33

🔴Lab 30

🟢Gre 14

🟡ACT 10

⚫NZF 7

🔝TOP 3

Commentary

It didn’t matter in 2020 given the overall state of things, but the boundary changes Taieri saw when it ceased being Dunedin South could now have a significant impact on this race. Labour incumbent Ingrid Leary should be safe: the seat still leans Labour by +4 on the party vote and incumbents do better than new candidates.

However, if she’s failed to win over voters in her term in office, or the polls are off in National’s favour, then National’s Matthew French could get there.

At 52^nd and 49^th on their lists respectively, both Leary and French depend on the seat. At 12^th, Green candidate Scott Willis is most likely safe unless the Greens undershoot their polls significantly. At 4^th for NZ First, Patterson is as safe provided they make the threshold.

Takanini

🔴Toss-up Labour hold

Candidate vote

🔴Lab – Leavasa (list 38th) 48% – odds 63%

🔵Nat – Nakhle (41st) 46% – 37%

History

➖➖➖➖➖➖➖➖🔴

Party vote

🔵Nat 43

🔴Lab 29

⚫NZF 9

🟡ACT 7

🟢Gre 6

🟣Con 2

Commentary

With the Labour wave in 2020, new seat Takanini did not attract much attention, with Labour’s Neru Leavasa getting home by 20%. However, the rematch with National’s Rima Nakhle will be close.

The seat –­ which mixes traditional Labour territory similar to the M-Block with more affluent parts of Manukau – is finely balanced in normal times, leaning towards both National (PVI+8) and Labour (PVI+3).

Given the limited available data here, there’s a lot of uncertainty about what will happen. Leavasa should see an incumbency bump, and unlike other close contests Hutt South or Northcote, Nakhle doesn’t have the benefit of being a former winner.

It’ll matter too, as neither candidate is in an electable list position.

Tāmaki

🔵Toss-up National hold

Candidate vote

🔵Nat – O'Connor (list 54th) 43% – odds 59%

🟡ACT – Van Velden (2nd) 39% – 41%

🔴Lab – Solomone (59th) 17% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 52

🟡ACT 16

🔴Lab 12

🟢Gre 11

⚫NZF 4

🔝TOP 3

Commentary

Some analysis we did earlier in the year suggested Tāmaki would be a good candidate for ACT seeking to expand their reach beyond neighbouring Epsom: National and ACT combined have a super majority of voters, and it is reasonably ACT-leaning (PVI+6).

So Van Velden’s decision to contest it properly makes strategic sense in terms of providing a second anchor seat. The modelling (using a similar approach to the one we have for Labour-Green races) suggests she and National’s Simon O’Connor will run close.

A larger National bloc on the party vote with ACT falling away recently, and residual FPP voting patterns may hold O’Connor up. Difficult to predict is what Labour and Green voters will do given Labour candidate Solomone has no up-the-middle chance.

Previous results aren’t an especially useful guide given the circumstances, but O’Connor did score well with 2020 Labour voters at (16%, 8/30 incumbents). Van Velden’s below-average split performances in radically different Wellington Central is no guide.

Maybe the best comparison is Rodney Hide’s original 2005 run for Epsom, where he picked up 60% of National voters and even 10% of Labour ones without a nod from the mothership. This time around the party’s survival is not at stake though.

Polling (Curia, 28 Sep: 🔵Nat 40, 🟡ACT 38, 🔴Lab 14 and Sentio, 20 Jul: 🔵Nat 36, 🟡ACT 34, 🔴Lab 8) has been consistent with these assumptions, but it’s still a coinflip.

Tāmaki Makaurau

🔴Toss-up Labour hold

Candidate vote

🔴Lab – Henare (list 14th) 43% – odds 63%

🟠TPM – Kemp (6th) 40% – 37%

🟢Gre – Tana (13th) 11% – 0%

🔵Nat – Te Hau (38th) 5% – 0%
🟣FNZ – Tamaki (3rd) 1% – 0%

History

➖➖🔴🟠🟠🟠🔴🔴🔴

Party vote

🟠TPM 36

🔴Lab 34

🟢Gre 11

⚫NZF 9

🔵Nat 4

🔣Oth 3

Commentary

Another close Māori seat contest sees Labour incumbent Peeni Henare take on newcomer Te Pāti Māori candidate Takutai Kemp. Factors discussed below make it difficult to pick, but even with a substantial swing towards TPM on current polling Henare is forecast to hang on.

The first difficulty is the high level of split-tick voting in both directions between Labour and TPM. It’s difficult to predict, but based on past results we’ve assumed Labour voters split 69/22 and TPM 60/30.

On top of that is how voting patterns will shift given that swing. On current numbers ~23% of voters will switch their party vote from Labour to TPM. As with Labour-National races, incumbents in Labour-TPM races tend to hold on to some (but not all) swing voters.

The a less stark partisan divide between Labour and TPM may make that more likely.

Then there are the other candidates. Green coleader Marama Davidson makes way for Darleen Tana. Without the coleader running a serious two-tick campaign (19% in 2020) more Green voters might split for either Labour or TPM.

National have also entered the candidate race with (former TPM Upper Harbour) candidate Hinuera Te Tau. It’s been 21 years since National stood a candidate, and they’ve never run against TPM in the Māori seats, so what the ~4% of National voters do could make a difference in a close race.

Finally, there is VisionNZ leader Hanah Tamaki. Vision (part of FreedomsNZ) are not forecast to do well in the Māori seats on past performance by Christian conservative parties (although this could be wrong given their different focus in 2023).

In terms of the list, Henare is in a relatively (but not completely) safe 14^th spot. Kemp will in all likelihood miss out at 6^th (and if TPM over-perform that much, she will probably win the seat regardless).

Tana is fairly (although again not completely) safe at 13^th for the Greens. Te Tau will miss out at 38^th for National.

Taranaki-King Country

🔵Safe National hold

Candidate vote

🔵Nat – Kuriger (list 36th) 62% – odds 100%

🔴Lab – Roberts (35th) 26% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 46

🔴Lab 15

🟡ACT 14

🟢Gre 8

⚫NZF 8

🟣Con 3

Commentary

A safe National seat for incumbent Barbara Kuriger. The lack of an ACT candidate with the late withdrawal of will boost her performance further, but with the seat leaning PVI+11 towards National she won’t need it.

At 35^th on the list, current Labour list MP Angela Roberts would have been safe in an OK year for Labour, but with their nationwide vote under 30%, she is all but guaranteed an exit.

Taupō

🔵Safe National hold

Candidate vote

🔵Nat – Upston (list 6th) 62% – odds 100%

🔴Lab – Al-Bustanji (64th) 20% – 0%
🟡ACT – Cozens (19th) 5% – 0%

History

🔴🔴🔴🔴🔵🔵🔵🔵🔵

Party vote

🔵Nat 47

🔴Lab 16

🟡ACT 13

⚫NZF 9

🟢Gre 7

🔝TOP 3

Commentary

Once upon a time Taupō was a contestable seat for Labour, but with boundary, demographic, and economic changes over time that has ceased to be the case (on average losing 1% per cycle on the party vote, net of nationwide swings). National incumbent Louise Upston wins easily.

At 64^th on the list, Labour’s Al-Bustanji is out of reach. There was probably a point earlier in the year where ACT’s Zane Cozens had a look in at 19^th, but given the shift in polling he’ll miss out.

Tauranga

🔵Safe National hold

Candidate vote

🔵Nat – Uffindell (list 57th) 56% – odds 100%

🔴Lab – Tinetti (6th) 21% – 0%

⚫NZF – Harvey (10th) 5% – 0%

🟡ACT – Young (18th) 5% – 0%

History

⚫⚫⚫🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 47

🔴Lab 13

🟡ACT 12

⚫NZF 10

🟢Gre 9

🔝TOP 4

Commentary

Starting again with another fairly mundane blue seat. 2022 byelection opponents Sam Uffindell of National and Janette Tinetti of Labour square off again. Despite a scandals. Uffindell has retained his party’s backing, and with it will retain the seat.

Byelection ACT candidate Cameron Luxton over-performed expectations, so there’s a possibility replacement Christine Young also makes a good showing, but in 2023 that won’t affect the outcome.

Tinetti is one of the few Labour MPs who’s genuinely safe on the list, so will return to Parliament regardless at 6^th. NZ First’s Erika Harvey is probably out of contention at 10^th, but if the party can replicate some of their previous over-performances she might sneak in.

Te Atatū

🔴Solid Labour hold

Candidate vote

🔴Lab – Twyford (list 49th) 47% – odds 99.6%

🔵Nat – Nicholas (39th) 32% – 0.3%

🟠TPM – Tamihere (26th) 10% – 0%

🟡ACT – Court (8th) 4% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 37

🔵Nat 32

🟢Gre 10

⚫NZF 7

🟡ACT 7

🔝TOP 2

Commentary

Labour veteran Phil Twyford ought to be returned for a fifth time around. Despite the swing, the seat is still Labour PVI+11 on the party vote, and Labour and the Greens combined still comfortably leade National and ACT by a combined 8%.

The one complicating factor in some extreme edge cases is Te Pāti Māori president and former co-leader John Tamihere. If his name recognition and standing in this west Auckland community can pull a combined 10% from Labour, Green, and NZ First voters, it could just do enough.

We think it’s a very, very remote possibility, but if it happens then National’s Angee Nichols might make a surprise trip to Wellington. At 49^th and 39^th on the list, neither major party candidate is electable from there.

Tamihere’s list position at 26^th puts him well out of contention, so his presence here is largely part of TPM’s attempts to lift their party vote with the general roll. Current ACT list MP Simon Court is on his way back at 8^th.

Te Tai Hauāuru

🟠Toss-up Māori GAIN

Candidate vote

🟠TPM – Ngarewa-Packer (list 1st) 56% – odds 63%

🔴Lab – Peke-Mason (n/a) 37% – 37%

🔵Nat – Hipango (31st) 6% – 0%

History

⚫🔴(🔴/🟠)🟠🟠🟠🔴🔴🔴

Party vote

🔴Lab 39

🟠TPM 30

🟢Gre 11

⚫NZF 8

🔵Nat 4

🔣Oth 4

Commentary

Te Pāti Māori’s best chance of a pick up in 2023 sees co-leader Debbie Ngarewa-Packer take on Labour list MP Soraya Peke-Mason. Given the expected swing towards TPM and Ngarewa-Packer’s near miss (4%) in 2020, she is probably marginally ahead.

Labour incumbent and Speaker Adrian Rurawhe’s decision to go list-only adds to Ngarewa-Packer’s chances, but given the uncertain split vote loyalties between Labour and TPM in the Māori seats, Peke-Mason still has a solid chance.

As with Tāmaki-Makaurau, the entry of a National candidate (list MP Harete Hipango) adds another complication (albeit a small one given National and ACT only poll an expected 5% combined). She won’t win, but could hurt Ngarewa-Packer’s chances.

Dame Tariana Turia’s win in the 2004 by-election was what gave TPM their start following the break with Labour over the Foreshore and Seabed legislation, so the party’s roots here run fairly deep. However, Turia has endorsed National’s Hipango.

If that matters to TPM voters, then it could make a difference (not modelled for).

Opinion polling sees a very different race, with Peke-Mason 5% ahead. (Curia, 19 Sep: 🔴Lab 34 🟠TPM 29 🔵Nat 12). Given the party vote in the poll showed National on a very surprising 12% (which would be a Māori seat record) it’s tempting to be sceptical.

At 1^st on the TPM list, Ngarewa-Packer will get a list spot if there any to go around (which there may not be depending on other races). Peke-Mason is list-only, and Hipango is in a marginal (marginally out) 31^st.

Te Tai Tokerau

🔴Lean Labour hold

Candidate vote

🔴Lab – Davis (list 2nd) 47% – odds 71%

🟠TPM – Kapa-Kingi (7th) 38% – 28%

🟢Gre – Lyndon (10th) 9% – 0%

History

⚫🔴🔴🟠(🟠/🟤)🟤🔴🔴🔴

Party vote

🟠TPM 41

🔴Lab 30

⚫NZF 10

🟢Gre 8

🔵Nat 4

🔣Oth 4

Commentary

Labour Deputy Leader Kelvin Davis faces a stiff challenge from Te Pāti Māori’s Mariameno Kapa-Kingi. On the one hand, the swing towards TPM will probably put them in front on the party vote. But in Davis’ favour, he has historically done well with TPM voters.

This effect should only increase as voters swing away from Labour. The sizeable number of expected NZ First voters (their best electorate in 2020 with 7.1% of the vote) probably also boost Davis, given they’ve tended to favour Labour over TPM.

An additional uncertainty is the two-tick campaign from Green candidate Huhana Lyndon. Difficult to say which candidate this will benefit, although Green voters have historically marginally preferred Labour to TPM or MANA candidates.

At 2^nd on the list, Davis is safe in any case. Lyndon should also be unless the Greens have an incredibly bad night. At 7^th, Kapa-Kingi will miss out.

Te Tai Tonga

🔴Likely Labour hold

Candidate vote

🔴Lab – Tirikatene (list 21st) 51% – odds 82%

🟠TPM – Ferris (5th) 44% – 18%

History

⚫🔴🔴🔴🟠🔴🔴🔴🔴

Party vote

🔴Lab 33

🟠TPM 25

🟢Gre 16

🔵Nat 8

⚫NZF 8

🔝TOP 4

Commentary

The Tirikatene whanau in Te Tai Tonga, there probably isn’t a lot more to say. Their standing in the electorate is near-unimpeachable, but Te Pāti Māori candidate Tākuta Ferris will try anyway.

Ferris did manage a respectable 25% on just 9% of the party vote in 2020. If TPM do as well as is anticipated at nearer 25% this time around, there is some remote chance he’s able to do it. At 5^th on the list, he’s next cab off the rank on current polling, so may not need to.

The seat is National’s strongest among the Māori seats, with an expected 9% (and even 6% in the depths of 2020), and National voters favoured Ferris 2:1 over Tirikatene in 2020.

Tukituki

🔵Solid National GAIN

Candidate vote

🔵Nat – Wedd (list 23rd) 51% – odds 98%

🔴Lab – Lorck (40th) 36% – 2%

History

🔴🔴🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 42

🔴Lab 21

🟡ACT 12

🟢Gre 10

⚫NZF 7

🟣Con 3

Commentary

Labour incumbent and vitamins enthusiast Anna Lorck was able to tip out National incumbent Lawrence Yule off the back of the red wave. She faces an incredibly difficult task repeating the effort against National challenger Catherine Wedd.

Lorck’s performance among National voters was an OK 3% (3^rd best out of Labour candidates challenging incumbents). Even if she is able to improve on these numbers as an incumbent and with swing voters moving back to National, she’ll struggle on a net-40% swing.

While it has an urban core of Labour-leaning territory around Hastings, the seat is a very bluish purple at PVI+7 for National and PVI-4 for Labour. By contrast, neighbouring Napier (where Stuart Nash was able to win in the lean years) only favours National by +5 on net.

Wedd would be in on the list on current polling at 23^rd, but a swing against National on polling day and a few wins for candidates deeper down the list could change that.

Upper Harbour

🔵Solid National GAIN

Candidate vote

🔵Nat – Brewer (list 62nd) 47% – odds 99%

🔴Lab – Walters (30th) 32% – 1%

🟢Gre – Doyle (33rd) 6% – 0%

⚫NZF – Wilson (9th) 6% – 0%

🟡ACT – Chhour (6th) 5% – 0%

History

➖➖➖➖➖➖🔵🔵🔴

Party vote

🔵Nat 43

🔴Lab 24

🟡ACT 11

🟢Gre 8

⚫NZF 6

🔝TOP 3

Commentary

Like Tukituki, Upper Harbour a seat Labour can hope to win open in a good year, or even against an incumbent in a great one. As mentioned repeatedly in this thread: 2023 is at best mediocre, so even with incumbency and facing a new challenger Walters’ chances are dire.

2020 National candidate Jake Bezzant is replaced by former (non-local) councillor Cameron Brewer, a switch which can only improve National’s chances.

At 30^th on the list, Walters’ is largely out of contention. NZ First’s David Wilson at 9^th stands some chance, and ACT MP Karen Chhour is assured a place at 6^th.

Waiariki

🟠Safe Māori hold

Candidate vote

🟠TPM – Waititi (list 2nd) 66% – odds 100%

🔴Lab – Boynton (39th) 32% – 0%

🟣FNZ – Hunia (12nd) 3% – 0%

History

⚫🔴🔴🟠🟠🟠🟠🟠🟠

Party vote

🟠TPM 41

🔴Lab 33

🟢Gre 8

⚫NZF 7

🔣Oth 3

🔵Nat 3

Commentary

Te Pāti Māori candidate Rawiri Waititi’s surprising (but not shocking) 2020 win in Waiariki was responsible for reviving Te Pāti’s fortunes following their stint in the wilderness. In 2023, anything other than a solid victor for him would be a surprise.

Waiariki is TPM’s best seat overall (part of the reason Waititi won when Ngarewa-Packer and Tamihere didn’t) and there should be a significant swing to the party based on nationwide polling.

Waititi did well with Labour voters in 2020 (34% of splits) but that could go either way in 2023: on the plus side he’s now an incumbent and TPM more align with Labour, on the downside, many of those split voters will have moved to TPM on the party vote.

Polling is consistent with the expected outcome (Curia, 9 Oct: 🟠TPM 50, 🔴22)

No candidates will be elected from the list.

Waikato

🔵Safe National hold

Candidate vote

🔵Nat – Van De Molen (list 58th) 68% – odds 100%

🔴Lab – Toko (73rd) 24% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 49

🔴Lab 14

🟡ACT 14

⚫NZF 9

🟢Gre 6

🟣Con 3

Commentary

Even in 2020, National’s Tim Van de Molen won by 13%. It will come as a shock to no one that he will win it again.

No candidates will be elected from the list.

Waimakariri

🔵Safe National hold

Candidate vote

🔵Nat – Doocey (list 8th) 60% – odds 100%

🔴Lab – Rosewarne (32nd) 27% – 0%

🟣LBP – Baker (2nd) 5% – 0%

History

🔴🔴🔴🔴🔴🔵🔵🔵🔵

Party vote

🔵Nat 44

🔴Lab 19

🟡ACT 13

🟢Gre 10

⚫NZF 7

🟣Con 3

Commentary

A seat that National did well to hang onto in 2020, given that in past waves Labour were able to win it. National incumbent Matt Doocey has an easier time of it in his third rematch with Labour’s Dan Rosewarne.

Leighton Baker Party leader (Leighton Baker) will probably get his deposit back given the seat is one where outsider conservatives do well.

At 32^nd on the Labour list, Rosewarne won’t be returned to Parliament.

Wairarapa

🔵Toss-up National GAIN

Candidate vote

🔵Nat – Butterick (list 61st) 45% – odds 52%

🔴Lab – McAnulty (16th) 43% – 48%

🟢Gre – Wade-Brown (15th) 5% – 0%

History

🔵🔴🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 41

🔴Lab 21

🟡ACT 12

⚫NZF 10

🟢Gre 10

🔝TOP 2

Commentary

The closest race in the country. Class-of-2020 incumbent Labour’s Kieran McAnulty is in an incredibly close fight to retain his seat against in a rematch with National’s Mike Butterick. We’ve got this as a toss-up National gain because you’ve gotta take a position but 🤷‍♂️🤷‍♂️🤷‍♂️.

The race isn’t just close in terms of margin (2%) but is uncertain on top of that.

Starting with the basics: the seat leans National by PVI+6 on the party vote, and there’s also a significant NZ First PVI+3 contingent. Butterick starts with a net +15% lead between National and ACT vs Labour and the Greens on the party vote.

In McAnulty’s favour, he was the best performer for Labour in an open seat, picking up 4.4% of National voters. Incumbency and retaining voters shifting to other parties on the party vote should lift that figure.

There’s a large uncertainty about the strength of the incumbency effect, so a significant win by McAnulty isn’t actually out of the question.

Added to this is the impact of other candidates (or lack thereof). First, the Green Celia Wade-Brown could act as a mild spoiler: although she was only mayor over the hill so that might not count for much. In 2020 Green voters seemed to understand the FPP dynamic (19% two-ticks).

The absence of an NZ First candidate probably works in McAnulty’s favour. NZ First voters historically tended to favour Labour candidates by an average of 14%, or much more then there’s either no NZF candidate or one not called Ron Mark.

Given shifting voter bases and partisan alignments though, this may not hold true in 2023, so wide confidence intervals.

If McAnulty does miss out, his list position at 16^th is shaky. He’d miss out on current estimates, but as with many marginal candidates, a polling miss in Labour’s favour or some losses further down the list could see him in.

Butterick is all-or-nothing in the seat race given he’s 61^st on the list. Wade-Brown would be in on current numbers at 15^th, but a polling miss against the Greens could still see her out.

Waitaki

🔵Safe National hold

Candidate vote

🔵Nat – Anderson (list 59th) 56% – odds 100%

🔴Lab – Reille (76th) 22% – 0%

History

🔵🔵🔴🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 44

🔴Lab 16

🟡ACT 15

🟢Gre 12

⚫NZF 6

🔝TOP 3

Commentary

National newcomer Miles Anderson (taking over from retiring incumbent Jacqui Dean) is a shoe-in. Many moons, boundary shifts, and demographic shifts ago Labour were able to pull off a win here (and got within 7% of an incumbent in 2020). Not any more.

Riding in the caboose at 76^th, Labour challenger Ethan Reille won’t be elected from the list.

Wellington Central

🟢Likely Green GAIN

Candidate vote

🟢Gre – Paul (list n/a) 41% – odds 82%

🔴Lab – Omer (37th) 31% – 18%

🔵Nat – Sheeran (45th) 24% – 0%

History

🟡🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🟢Gre 38

🔴Lab 26

🔵Nat 17

🔝TOP 8

🟡ACT 6

⚫NZF 3

Commentary

Like its neighbour Rongotai to the south, Wellington Central sees a close contest between Labour and the Greens to replace a Labour incumbent (Grant Robertson, now list-only) in a deeply red-green seat that’s been getting greener overtime.

Again, there’s no good precedent for a Labour-Green head-to-head, so there’s a lot of uncertainty about what the two party’s voters will do. But given how baseline Green the seat is, we have local councillor Tamatha Paul ahead of Labour list MP Ibrahim Omer.

The Green’s strongest (PVI+25) seat in the country by a wide margin, and one where their current nationwide polling would have them winning the party vote (even if the assumption that the long-term +2% per cycle trend abates). Labour also do narrowly better here at PVI+1.

What this would mean for the candidate race is that unlike colleagues in the other red-green races, Paul is not relying on doing better with Labour voters than Omer does with Green ones (and conversely: Omer is).

Polling had a much closer race with Omer ahead (Reid, 14 Sep: 🔴Lab 31 🟢Gre 27 🔵Nat 28) but given it also showed a -4 swing away from the Greens on the back of a nationwide swing of +4 towards them, I’d be cautious about relying on it too much.

So advantage Paul, but if the Greens underperform their polls (either locally or nationally) and Omer does well with Green (and National – Sheeran is not a factor in his own right here) voters, then he still has a shot.

None of the candidates will get in on the list: Omer and Sheeran are too low at 37^th and 49th respectively, and Paul is unranked.

West Coast-Tasman

🔴Solid Labour hold

Candidate vote

🔴Lab – O'Connor (list 10th) 46% – odds 95%

🔵Nat – Pugh (26th) 37% – 5%

🟣FNZ – Grey (2nd) 4% – 0%

History

🔴🔴🔴🔴🔵🔴🔴🔴🔴

Party vote

🔵Nat 32

🔴Lab 27

🟢Gre 13

🟡ACT 12

⚫NZF 7

🔣Oth 3

Commentary

Another 2020 rematch, but one that should see incumbent (Damian) O’Connor hold on. However, Pugh’s chances are significantly better than being struck by lightning a fourth time.

The seat is purple purple, leaning narrowly towards Labour at PVI+1 and slightly against National at PVI-2. On current polling and adding in ACT and the Greens (who both over-poll here) it’s currently estimated at a narrow +3 on net for the right.

However, in O’Connor’s favour he’s historically done well with National votes (averaging 13% - albeit only 4% in 2020 when many of those split voters moved to Labour). There’s some uncertainty about whether he recovers this position on the swing however.

The other wild card is FreedomsNZ’s Sue Grey. We have her on 4%, but with wide uncertainties. WCT has been (comparatively) favourable territory for outsider parties like Ban1080 or Advance/NZPP, and her profile in the anti-mandate community may draw voters from other parties.

O’Connor is at a marginal 10^th on the list, but should be safe on current figures. Pugh is currently last one-on at 26^th, but a small swing against National or surprise results lower down would have her gone.

Whanganui

🔵Lean National GAIN

Candidate vote

🔵Nat – Bates (list 47th) 44% – odds 68%

🔴Lab – Lewis (50th) 42% – 32%

History

🔴🔴🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 37

🔴Lab 27

⚫NZF 10

🟡ACT 10

🟢Gre 8

🔣Oth 2

Commentary

Another knife-edge defensive effort by a 2020 Labour incumbent. On current polling we have National’s Carl Bates narrowly ahead of Labour’s Steph Lewis, but that could easily switch.

On the party vote the seat is very close in normal times: both National (PVI+2) and Labour (PVI+1) slightly over-performing. On current polling this gives National a comfortable 10% lead locally.

In Lewis’ favour, and making this a closer contest than other +10% races, she performed the best out of all Labour candidates challenging an incumbent at 3.4% with National voters and 88% of Labour one (with Labour voters she was 6^th overall, impressive as a newcomer).

She should be able to build on both those figures as an incumbent and with swing voters voting back home to National but sticking with her for the second tick. Whether she can win will depend on how much. Currently we think just not enough.

At 50^th and 47^th on the list, neither candidate has that as a backup plan.

Whangaparāoa

🔵Safe National hold

Candidate vote

🔵Nat – Mitchell (list 11st) 65% – odds 100%

🔴Lab – Muller Pallarès (55th) 19% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 51

🟡ACT 15

🔴Lab 10

🟢Gre 9

⚫NZF 7

🟣Con 4

Commentary

Profoundly blue-yellow exurban Auckland seat: an easy win for incumbent Mark Mitchell.

Other random observation: this is a seat where the more religiously-inclined outsider right parties can expect to do well on the party vote given historic patterns.

Whangārei

🔵Safe National GAIN

Candidate vote

🔵Nat – Reti (list 4th) 50% – odds 100%

🔴Lab – Warren-Clark (43rd) 31% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 37

🔴Lab 19

⚫NZF 12

🟢Gre 11

🟡ACT 11

🔝TOP 3

One of the reach seats that Labour managed to win for the first time under MMP in 2020, this was always going to be struggle for Labour to defend. The loss of incumbency with the departure of Labour’s Emily Henderson and former incumbent Shane Reti will give it to National.

Reit performed about par with Labour voters in 2020 at 13%, a figure that might slip somewhat as those voters return to National. On the other hand, in some ways he’s more like an incumbent now, so might retain those figures.

At any rate, he only lost by a narrow 1% in 2020. Replacement Labour candidate and incumbent list MP Angie Warren-Clark will miss out on the list at 43^rd.

Wigram

🔴Safe Labour hold

Candidate vote

🔴Lab – Woods (list 5th) 50% – odds 100%

🔵Nat – Summerfield (37th) 33% – 0%

History

⚪⚪⚪⚪⚪🔴🔴🔴🔴

Party vote

🔵Nat 32

🔴Lab 30

🟢Gre 15

🟡ACT 7

⚫NZF 7

🔝TOP 5

Commentary

An easy one to round things out. Labour’s Megan Woods will be safely returned in Wigram over National challenger Tracy Summerfield. While the seat leans Labour overall (PVI+4, with PVI+3 for the Greens for good measure), on current polling National should be narrowly ahead here.

However, strong support from Labour voters, respectable levels from Green ones, and a modest showing with National ones give Woods a safe buffer.

No candidates will be elected from the list.

Note that all numbers are as-of 7 October. A (mercifully shorter) updated will follow on Friday.

If you’d rather just look at a map, here you go:

Auckland Central

🟢Solid Green hold

Candidate vote

🟢Gre – Swarbrick (list 3rd) 42% – odds 90%

🔵Nat – Muralidhar (43rd) 36% – 10%

🔴Lab – Sims (63rd) 16% – 0%

History

 🔴🔴🔴🔴🔵🔵🔵🔵🟢

Party vote

🔵Nat 32

🟢Gre 26

🔴Lab 20

🟡ACT 10

⚫NZF 4

🔝TOP 4

Commentary

Green incumbent Chlöe Swarbrick pulled off a somewhat surprising win in 2020. With the benefit of incumbency likely to improve her performance among voters from all parties, a reasonable (+7) expected local party vote swing towards the Greens, and the perception of a two-horse race with National’s Mahesh Muralidhar should see her retain the seat.

Downside risk comes from the even stronger (+11) expected party vote swing towards National in the seat, and the chance that a strong performance from Labour’s Oscar Sims splits the left vote.

Polling had the race closer than the model does (Curia 24 Sep: 🟢Gre 38, 🔵Nat 36, 🔴Lab 17) but Swarbrick’s 2020 over-performance vs polling and the difficulty of polling electorates leave us reasonably confident the margin will be wider than that suggests.

If Swarbrick loses, her position at 3^rd on the list will see her re-enter Parliament regardless. Given their rankings, neither Muralidhar at 43^rd or Sims at 63^rd will enter off the list.

Banks Peninsula

🔴Solid Labour hold

Candidate vote

🔴Lab – McLellan (list 27th) 46% – odds 99%

🔵Nat – Weenink (40th) 32% – 1%

🟢Gre – Pham (6th) 13% – 0%

🟡ACT – Trask (10th) 4% – 0%

History

🔵🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔵Nat 29

🔴Lab 27

🟢Gre 21

🟡ACT 9

⚫NZF 5

🔝TOP 4

Commentary

Labour incumbent Tracey McLellan took over the seat from long-time incumbent Ruth Dyson in 2020. Despite a -23 expected swing away from Labour on the party vote, incumbency and support from Green voters should see her retain the seat.

Very remote chance for National’s Vanessa Weenink would require a combination of a further party vote swing to National, a weak incumbency effect for McLellan, and an over-performance from new Green candidate Lan Pham who takes over from veteran (and comparatively well-performing) Green candidate Eugenie Sage

At 27^th, McLellan would not return on the list if she somehow loses the seat. Weenink at 40^th is also entirely reliant on winning the seat. Pham’s high placing (6^th) will see her entering Parliament, and ACT’s Trask is probably safe at 10^th.

Bay of Plenty

🔵Safe National hold

Candidate vote

🔵Nat – Rutherford (list, n/a) 60% – odds 100%

🔴Lab – Taikato (54th) 19% – 0%

⚫NZF – Murfitt (11th) 9% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 47

🟡ACT 13

🔴Lab 12

⚫NZF 10

🟢Gre 8

🔝TOP 4

Commentary

New National candidate Tom Rutherford takes over from retiring incumbent Todd Muller. While Muller held the seat by 7% in the 2020 red wave, the seats solid National lean (PVI +13) and large (+12) party vote swing towards National put this beyond doubt.

Labour’s Pare Taikato at 54^th will not enter on the list, but NZ First’s Kristen Murfitt at 11^th is an outside chance if the party over-performs its current polling nationwide.

Botany

🔵Safe National hold

Candidate vote

🔵Nat – Luxon (list 1st) 66% – odds 100%

🔴Lab – Singh (n/a) 28% – 0%

Party vote

🔵Nat 49

🔴Lab 27

🟡ACT 9

🟢Gre 6

⚫NZF 5

🟣Con 2

History

➖➖➖➖🔵🔵🔵🔵🔵

Commentary

National incumbent and leader Christopher Luxon took over the seat from scandal-plagued rebel Jami-Lee Ross in 2020. The “favourite son” effect (a boost in support across the board party leaders see), incumbency, and a natural blue lean (PVI +14) will see him retain the seat with a greatly increased margin.

Christchurch Central

🔴Safe Labour hold

Candidate vote

🔴Lab – Webb (list 20th) 42% – odds 100%

🔵Nat – Stephens (29th) 33% – 0%

🟢Gre – Carter (14th) 16% – 0%

History

🔴🔴🔴🔴🔴🔵🔵🔴🔴

Party vote

🔵Nat 30

🔴Lab 28

🟢Gre 19

🟡ACT 8

⚫NZF 6

🔝TOP 5

Commentary

Labour incumbent Webb has held the seat since 2017 after winning it back from National’s Wagner. Despite the likely swing away from Labour (-16) towards National (+9) and the Greens (+9), his own above-average performance will return him.

At 29^th on the list, National’s Dale Stephens has a slightly below 50% chance of a place in parliament. Barring a significant party vote under-performance by the Greens’, Kahurangi Carter at 14th should be elected from the list.

Christchurch East

🔴Safe Labour hold

Candidate vote

🔴Lab – Davidson (list 57th) 49% – odds 100%

🔵Nat – Stock (50th) 31% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 34

🔵Nat 28

🟢Gre 15

🟡ACT 8

⚫NZF 8

🔝TOP 3

Commentary

Labour newcomer Reuben Davidson takes over the seat from retiring incumbent Poto Williams. Despite the overall swing and the loss of incumbency advantage, Davidson is certain to win the solidly red (PVI+8) seat. No candidates will be elected on their parties’ lists.

Coromandel

🔵Safe National hold

Candidate vote

🔵Nat – Simpson (list 55th) 57% – odds 100%

🔴Lab – Riley (n/a) 22% – 0%

History

🔵🟢🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 44

🔴Lab 15

🟡ACT 13

⚫NZF 11

🟢Gre 10

🟣Con 2

Commentary

National incumbent Scott Simpson held the seat by a less-than-comfortable 7% margin in 2020. In 2023, with National’s overall improvement he will be safely returned. No candidates will on their parties’ list.

Dunedin

🔴Likely Labour hold

Candidate vote

🔴Lab – Brooking (list 23rd) 33% – odds 90%

🟢Gre – Hernandez (17th) 24% – 8%

🔵Nat – Woodhouse (n/a) 21% – 1%

🔣Ind – O’Malley (n/a) 10% – 1%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 31

🟢Gre 28

🔵Nat 20

🟡ACT 7

⚫NZF 6

🔝TOP 5

Commentary

One of the more interesting/strange candidate races. Labour list MP Rachael Brooking takes over from retiring Labour incumbent David Clark. Under normal circumstances, Brooking should expect to comfortably win, and may in fact do so. However…

Events have conspired to make the race less certain. First, National list-MP and long-term local candidate Michael Woodhouse – dissatisfied with his offered list ranking – has all-but conceded the race and is not actively campaigning.

Second, the Green’s Francisco Hernandez is running an active (if informal) campaign in a strongly Green-leaning (PVI +15, third highest in the country) seat. Finally, a late independent entry by centre-left local councillor O’Malley may attract a mix of Labour voters disatistifed with the party and National voters left without a candidate.

If she somehow misses, Brooking has a just below-even chance of a list place at 23^rd, dependent on other results. On current polling, Hernandez would just miss out at 17^th on the Green list, but even a tiny improvement would see him in.

East Coast

🔵Solid National GAIN

Candidate vote

🔵Nat – Kirkpatrick (list 44th) 46% – odds 100%

🔴Lab – Coffey (36th) 32% – 0%

History

🔴🔴🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 37

🔴Lab 24

🟢Gre 11

⚫NZF 10

🟡ACT 10

🔝TOP 2

Commentary

Retiring incumbent Labour MP Kiri Allen won the seat in the 2020 red wave with a defendable margin of 15%. The seat looked to be competitive at the start of the year, but following Allen’s scandals and resignation, National’s Dana Kirkpatrick is now safe.

Labour have tagged current list MP Tamati Coffey to stage a rear-guard action, but given the swing against Labour and the loss of Allen’s incumbency, his chances are non-existent. At 36^th on the list, he will get to retire as originally planned.

East Coast Bays

🔵Safe National hold

Candidate vote

🔵Nat – Stanford (list 7th) 68% – odds 100%

🔴Lab – Chen (33rd) 21% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 53

🟡ACT 13

🔴Lab 12

🟢Gre 10

⚫NZF 5

🟣Con 3

Commentary

In one of the bluest seats in the country, and as the best performer with non-National voters, National’s Erica Sanford is perhaps the safest MP in the country. An unassailable 23% margin in 2020 should grow to nearing 50% in 2023.

Current Labour list MP Naisi Chen’s forecast 21% is flattered by the absence of a Green candidate, and at 33^rd on the Labour list absent a miracle for the party or a catastrophe for her colleagues in other races, she is likely leaving Parliament.

Epsom

🟡Safe ACT hold

Candidate vote

🟡ACT – Seymour (list 1st) 52% – odds 100%

🔵Nat – Goldsmith (5th) 22% – 0%

🔴Lab – Belich (26th) 13% – 0%

🟢Gre – Xu-Nan (16th) 8% – 0%

⚫NZF – Unkovich (8th) 1% – 0%

History

🔵🔵🔵🟡🟡🟡🟡🟡🟡

Party vote

🔵Nat 49

🟢Gre 15

🟡ACT 14

🔴Lab 14

🔝TOP 4

⚫NZF 3

Commentary

ACT leader David Seymour will be safely returned in his Epsom anchor seat. Notwithstanding ACT’s recent slide, the party has still improved nationwide relative to 2020, and he performs will with voters from all parties.

The decline of the Hipsters for Goldsmith vote (Labour and Green voters trying to elect National’s Paul Goldsmith against his will) leaves the seat essentially uncontested. However, there’s a realistic prospect of the seat having five list MPs.

Goldsmith at 5^th on the list is assured of a spot in all but the strangest of scenarios. The Greens’ Lawrence Xu-Nan (16^th) and NZ First’s Tanya Unkovich (8^th) are both (just) in on current numbers. Finally, Labour’s Camilla Belich is an outside chance at 26^th.

On the party vote, picking the Greens at 2^nd might raise eyebrows, and it may not come off. Seymour’s success masks a long-term (relative) trend away from ACT, and conversely the Greens’ PVI position has improved every year since 2002.

Hamilton East

🔵Likely National GAIN

Candidate vote

🔵Nat – Hamilton (list n/a) 49% – odds 87%

🔴Lab – Dansey (31st) 38% – 13%

🟡ACT – Parmar (13th) 2% – 0%

History

🔵🔵🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 36

🔴Lab 26

🟢Gre 13

🟡ACT 10

⚫NZF 5

🔝TOP 4

Commentary

The first genuine contest. A bellwether seat on the party vote (second only to Maungakiekie in 2020), Hamilton East sees Georgie Dansey (Labour’s 2022 Hamilton West byelection candidate) face aptly-named local councillor National’s Ryan Hamilton.

Retiring incumbent and ANZAC Anschluss advocate Jamie Strange took the seat with a defendable 7% margin in 2020. The loss of incumbency and the swing to National should see Hamilton win comfortably.

However, the absence of a Green candidate and the presence of an ACT one gives Dansey a modest chance if she can over-perform with swing voters moving to National.

At 31^st , Dansey has a slim chance of entering on the Labour list (the only non-incumbent MP in this position). Hamilton is unranked, so is reliant on the seat. ACT candidate Himanshu Parmer at 13^th is just in on current numbers.

Hamilton West

🔵Safe National hold

Candidate vote

🔵Nat – Potaka (list 24th) 47% – odds 100%

🔴Lab – Williamson (62nd) 33% – 0%

🟢Gre – Doyle (18th) 6% – 0%

History

🔵🔴🔴🔵🔵🔵🔵🔵(🔴/🔵)

Party vote

🔵Nat 34

🔴Lab 28

🟡ACT 10

🟢Gre 9

⚫NZF 8

🔝TOP 4

Commentary

Ever-so-slightly more left-leaning than its neighbour across the river (the seventh most representative seat in 2020), Hamilton West will see 2022 byelection winner National’s Tama Potaka safely returned.

Erstwhile Labour rebel Gaurav Sharma won comfortably by 15% in 2020 and would have been a 50/50 shot to return. However, with Potaka taking over the incumbency advantage following the byelection, the seat is safely National’s.

Replacement Labour candidate Myra Williamson will not be elected at 62^nd on Labour’s list. However, Green candidate Benjamin Doyle could be if the Greens improve nationally.

Hauraki-Waikato

🔴Solid Labour hold

Candidate vote

🔴Lab – Mahuta (list n/a) 55% – odds 91%

🟠TPM – Maipi-Clarke (4th) 41% – 9%

🟣FNZ – Pokere-Phillips (4th) 5% – 0%

History

➖🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 39

🟠TPM 36

🟢Gre 8

⚫NZF 6

🔵Nat 4

🔣Oth 4

Commentary

Labour’s safest Māori seat should be an easy win for Nanaia Mahuta (the only candidate to hold the seat on its current boundaries) returned. However, the swing away from Labour and towards Te Pati Māori, combined with unpredictable splitting give Hana-Rawhiti Maipi-Clarke a chance.

While TPM have seen only a small increase in nationwide support (up 2% from 1.2% in 2020 to nearer 3% on current polling), this support tends to be highly concentrated in the Māori seats. Because of this, the swing in Hauraki-Waikato is likely to be 15x as large.

Polling has shown a closer race than expected (Curia, 27 Sep: 🔴Lab 36, 🟠TPM 32) but the high number of undecideds and frankly oddly-reported numbers make this less reliable. The presence of 2017 TPM candidate Donna Pokere-Phillips now with Freedom’s NZ only adds to the uncertainty.

On current party vote polling, Maipi-Clarke would just be returned on the list at 4^th, but given the unpredictability of other Māori seat results, could miss out.

Hutt South

🔵Toss-up National GAIN

Candidate vote

🔵Nat – Bishop (list 3rd) 45% – odds 65%

🔴Lab – Andersen (17th) 41% – 35%

History

🔴🔴🔴🔴🔴🔴🔴🔵🔴

Party vote

🔴Lab 31

🔵Nat 31

🟢Gre 15

🟡ACT 8

⚫NZF 6

🔝TOP 4

Commentary

Among the closest and most difficult to predict races, Hutt South sees a 2020 rematch between Labour incumbent Ginny Andersen and former incumbent Chris Bishop. On balance, Bishop is slightly ahead but the seat will probably be decided on specials.

The seat is favourable for Labour in normal times (PVI+6). However, on current polling the parties should be tied on the local party vote. On top of this, Bishop over-performs with Labour voters (20% in 2020, vs an average of 9%).

With swing voters returning to National, that 20% figure will drop. But likely neither candidate will benefit from an incumbency bump given they have both held the seat. At 17^th Andersen may just miss out on the list, other results depending. At 3^rd on his party’s list, Bishop is guaranteed to return.

Ikaroa-Rāwhiti

🟠Toss-up Māori hold

Candidate vote

🟠TPM – Whaitiri (list 3rd) 51% – odds 71%

🔴Lab – Tangaere-Manuel (n/a) 47% – 29%

🟣Con – Tuhakaraina (35th) 2% – 0%

History

⚫🔴🔴🔴🔴🔴🔴🔴(🔴/🟠)

Party vote

🔴Lab 45

🟠TPM 32

🟢Gre 9

⚫NZF 5

🔣Oth 4

🔵Nat 3

Commentary

What was already a close contest on current polling is even more complicated given Meka Whaitiri’s defection from Labour to Te Pāti Māori. We have this as a TPM hold, but it’s borderline impossible to model.

Fundamentals would normally favour a TPM “incumbent”: an estimated +20% swing in TPM’s favour, greater two-tick loyalty among TPM voters than Labour ones, and the absence of a Green candidate following Elizabeth Kerekere’s resignation.

However, whether personal loyalty and name recognition trump the sense of betrayal Labour party voters will feel is difficult to say. It’s very possible Labour voters stick with (or is that switch to??) Cushla Tangaere-Manuel, handing her victory.

In the event Whaitiri does miss out, she would return at 3^rd on the TPM list, but upsets in other Māori seats (Your Tais Tokerau/Tonga or Tāmaki Makaurau) could see her miss out. As a late replacement, Tangaere-Manuel is unranked.

Ilam

🔵Likely National GAIN

Candidate vote

🔵Nat – Campbell (list 63rd) 40% – odds 82%

🔝TOP – Manji (1st) 29% – 18%

🔴Lab – Pallett (51st) 22% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 39

🔴Lab 19

🟢Gre 16

🟡ACT 12

🔝TOP 7

⚫NZF 4

Commentary

The high-water mark of the red tide in 2020 saw Labour’s Sarah Pallett defeat eight-term incumbent Gerry Brownlee. In 2023, former local councillor then 2017 independent candidate, and now TOP leader Raf Manji is trying to use the seat to anchor his party’s chances.

The overall swing back towards National and Brownlee’s retirement would have seen (and probably still will see) National’s Hamish Campbell easily take the seat back, given its National PVI+4 and Labour PVI-7 leanings.

Using Manji’s numbers from his 2017 independent run still show him behind without an even implicit nod from National or Labour. An overperformance with National, Labour, and (the sizeable number of) Green voters might just be enough.

Polling was not favourable (Curia, 23 Aug: 🔵Nat 43, 🔴Lab 20, 🔝 TOP 18) but atypical cases like this are hard to model with certainty.

With Campbell at 63^rd and Pallett at 51^st on their party’s lists, they are both dependent on an electorate win. Despite being 1^st on his party’s, given TOP’s nationwide polling of ~2.5%, Manji is too.

Invercargill

🔵Safe National hold

Candidate vote

🔵Nat – Simmonds (list 16th) 62% – odds 100%

🔴Lab – Craig (44th) 33% – 0%

History

🔴🔴🔴🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 41

🔴Lab 22

🟡ACT 12

⚫NZF 9

🟢Gre 8

🔣Oth 2

Commentary

Proving that candidate quality still matters even if you can’t reliably predict it, in 2020 National’s Penny Simmonds managed to outrun her party for a narrow 0.6% victory. A rematch with 2020 Labour candidate Liz Craig won’t be that close.

A predicted net swing of 36% towards National on the party vote, and the benefits that come with incumbency should see Simmond’s margin increase 50-fold to nearer 30%. At 44^th on the party list, Craig will not be returned to Parliament.

Kaikōura

🔵Safe National hold

Candidate vote

🔵Nat – Smith (list 56th) 54% – odds 100%

🔴Lab – DeWhirst (n/a) 20% – 0%

⚫NZF – Arbuckle (6th) 7% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 44

🔴Lab 16

🟡ACT 14

🟢Gre 10

⚫NZF 9

🔝TOP 3

Commentary

In a sign of how wild 2020 was, National’s Stuart Smith saw his margin whittled to just 5%. The overall picture in 2023 will see that lead grow to nearer 35%. Labour’s Emma DeWhirst is unranked. NZ First candidate Jamie Arbuckle is as safe as any NZ First candidate at 6^th on the list.

Kaipara ki Mahurangi

🔵Safe National hold

Candidate vote

🔵Nat – Penk (list 18th) 56% – odds 100%

🔴Lab – Wishart (n/a) 20% – 0%

⚫NZF – Marcroft (5th) 4% – 0%

History

➖➖🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 45

🟡ACT 15

🔴Lab 14

🟢Gre 11

⚫NZF 7

🔝TOP 3

Commentary

Another safe, uninteresting National seat for Twitter’s own Chris Penk. Fellow bipartisan Posters’ Caucus member Labour’s Marja Lubeck makes way for Guy Wishart in taking a 30%+ hiding.

NZ First’s Jenny Marcroft looks set to return to Parliament at 5^th on their list.

Kelston

🔴Safe Labour hold

Candidate vote

🔴Lab – Sepuloni (list 3rd) 54% – odds 100%

🔵Nat – Schaumkel (42nd) 23% – 0%

🟢Gre – Ghahraman (7th) 8% – 0%

History

🔵🔵🔴🔴🔵🔵🔴🔴🔴

Party vote

🔴Lab 43

🔵Nat 23

🟢Gre 13

⚫NZF 7

🟡ACT 6

🔝TOP 2

Commentary

Labour Deputy Prime Minister (but not Deputy Leader) Carmel Sepuloni is set to retain the safe (as of the 2014 boundary change) Labour seat of Kelston.

At 42^nd, National’s Ruby Schaumkel has no chance of election from the list, but the Green’s Golriz  Ghahraman will be safely returned at 7^th.

Mana

🔴Safe Labour hold

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Candidate vote

🔴Lab – Edmonds (list 18th) 53% – odds 100%

🔵Nat – Hughes (33rd) 30% – 0%

Party vote

🔴Lab 37

🔵Nat 26

🟢Gre 15

🟡ACT 8

⚫NZF 6

🔝TOP 4

Commentary

Labour’s Barbara Edmonds will be comfortably returned in deep-red Mana. More important than Mana’s pro-Labour lean is its tendency not to swing quite as much. Labour loyalty means only a -20ish swing versus a -25% swing nationwide.

On top of that, Edmonds was one of the better 2020 performers among National voters at 5% (3^rd highest among non-incumbents). With voters swinging back to National from Labour, expect that figure to improve.

At 33^rd on the list, National’s Frances Hughes would need a combination of good party vote performance and poor (other) candidate performances to be elected.

Māngere

🔴Safe Labour hold

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Candidate vote

🔴Lab – Sosene (list 53rd) 71% – odds 100%

🔵Nat – Bourke (32nd) 13% – 0%

Party vote

🔴Lab 69

🔵Nat 11

⚫NZF 6

🟢Gre 5

🔣Oth 3

🟠TPM 2

Commentary

Māngere is not only deeply favourable to Labour (PVI+43): it’s also profoundly loyal to the party. For every 1% of swing nationwide, you only see a 0.4% swing in the seat (the second stickiest seat in the country for the party).

All of which is to say this will be an easy hold even for new candidate Lemauga Sosene. At 32^nd, National’s Rosemary Bourke is probably beyond the list line (currently sits at 26/27) at this point, but the right combination of other National electorate candidates further down the list missing out could see her in.

Manurewa

🔴Safe Labour hold

Candidate vote

🔴Lab – Williams (list 48th) 62% – odds 100%

🔵Nat – Kilari (30th) 16% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 60

🔵Nat 18

⚫NZF 8

🟢Gre 4

🟡ACT 3

🔣Oth 3

Commentary

Third seat in Labour’s M-block heartland, and another easy hold for sophomore candidate Arena Williams. As with similar seats, Labour does both well (PVI+34) and are consistent (swing coefficient of just 0.6).

Williams was a strong performer compared to other new Labour candidates in 2020 (2^nd with National voters on 6%, 3^rd among Labour voters on 87%) so the controversy around her selection battle with Louisa Wall did not seem to affect her performance.

At 30^th on the list, National’s Siva Kilare will probably be just outside contention but has a slightly better shot than the abovementioned Bourke and Hughes.

Maungakiekie

🔴Likely Labour hold

History

🔵🔴🔴🔴🔵🔵🔵🔵🔴

Candidate vote

🔴Lab – Radhakrishnan (list 15th) 47% – odds 82%

🔵Nat – Fleming (65th) 38% – 12%

Party vote

🔴Lab 33

🔵Nat 33

🟢Gre 12

🟡ACT 10

⚫NZF 5

🔝TOP 3

Commentary

One of the fiercest contests in the country. In 2020, Maungakiekie was the best bellwether on the party vote (most closely resembled the nationwide party vote). Labour’s Priyanca Radhakrishnan managed only a 1.7% win in 2020 despite the nationwide tailwind.

This year the race looks set to be close again. The swing against Labour (though smaller than the nationwide swing) means on the party vote it’s likely a level 33/33 contest (although the Green vote is slightly larger than the ACT one).

Radhakrishnan is now running as an incumbent, potentially offsetting the swing. More important though is 2020 National candidate Denise Lee choosing not to stand again, given her very strong performance among Labour voters (18% in 2020).

If Radhakrishnan sees a generic level of incumbency advantage (the assumption made here) it should see her home fairly comfortably. Whether that happens is uncertain though, so Fleming still has a real chance. At 65^th on the list he will need it.

A potential (unmodelled) factor could be an under-performance by controversial Green candidate Sapna Samant. Our current assumption is she’ll pick up 28% of the Green party vote, but if she doesn’t that may benefit Radhakrishnan.

Mount Albert

🔴Likely Labour hold

Candidate vote

🔴Lab – White (list 47th) 37% – odds 80%

🟢Gre – Menendez March (8th) 31% – 18%

🔵Nat – Lee (13th) 25% – 2%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 29

🟢Gre 27

🔵Nat 25

🟡ACT 8

🔝TOP 5

⚫NZF 4

Commentary

2020 saw then-Prime Minister Jacinda Ardern wrack-up one of the largest individual candidate majorities of all time. Three factors will lead to a very different race in 2023. First – obviously – is Ardern’s retirement.

Second is the overall swing to National and the Greens (estimated +7 a piece). This would see the three parties within ~4 points of one another on the party vote.

The third factor – which will take some explaining – is the Greens’ decision to run a two-tick campaign.

To estimate the impact this might have on the race, we use Green candidate split performance in past two-tick campaigns. Specifically:

• Coromandel in 2002,

• Helensville* and Nelson in 2017; and

• Auckland Central and Tāmaki Makaurau in 2020.

*Not strictly-speaking a formal two-tick campaign, but Green candidate Haley Holt’s performance resembled one.

The pattern across these races is that Green candidates do do better when they ask for the electorate vote. Unsurprisingly the effect is strongest with Green voters (80%, vs 37% across all races). But it also holds for Labour (30% vs 7%) and National (5% vs 1%) voters.

A complication with Labour voters is we don’t have any prior examples for head-to-head Labour vs Green contests. To account for this, we arbitrarily nerf the Green candidate share of Labour vote down to 15%.

Right, with that out of the way: new Labour candidate Helen White is likely to replace Ardern. The seat is baseline Labour/Green friendly (PVI +3 and +14 respectively), and Labour should narrowly lead the Greens on the party vote.

However, any one of three factors could tip it for Green candidate MP Ricardo Menendez-March. First, if the Greens improve their nationwide total the balance on the party vote will tip in the Greens’ favour (in Mt Albert the Greens get 1.3% for every 1% swing nationwide).

Second, the concerted campaign could lift the Green party vote more than expected (as it did in neighbouring Auckland Central in 2020). Finally, Labour voters may be less loyal to their candidate than Green voters are to theirs.

There is also an outside possibility of National’s Melissa Lee running up the middle. Earlier this year when the Greens decided to go for two-ticks and Labour were polling significantly better, this looked impossible. It’s still unlikely, but a 33/33 split might just do it.

On the list, Lee and Menendez March are reasonably safe at 13^th and 8^th respectively. White on the other hand will miss out at 47^th if she repeats her 2020 Auckland Central result.

Mount Roskill

🔴Safe Labour hold

Candidate vote

🔴Lab – Wood (list 45th) 55% – odds 100%

🔵Nat – Cheung (48th) 34% – 0%

History

➖🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 39

🔵Nat 33

🟢Gre 11

🟡ACT 7

⚫NZF 5

🔝TOP 3

Commentary

Back into the conventional M-block, Mount Roskill should see incumbent Michael Wood easily hold on. Even if there’s a significant dip due to his scandals earlier this year (not modelled for here) an expected 20% buffer will be enough.

No candidates will be elected off the list.

Napier

🔵Safe National GAIN

Candidate vote

🔵Nat – Nimon (list 22nd) 51% – odds 100%

🔴Lab – Hutchinson (60th) 34% – 0%
🟡ACT – Milewski (34th) 3% – 0%

History

🔴🔴🔴🔵🔵🔵🔴🔴🔴

Party vote

🔵Nat 40

🔴Lab 25

🟡ACT 11

🟢Gre 10

⚫NZF 7

🟣Con 3

Commentary

Like its neighbour East Coast, Napier sees an incumbent resigning amid scandal turn a close race into an easy National gain. Retiring incumbent Stuart Nash (an over-performer with National and NZ First voters) would have struggled but may have got there even swimming against the tide.

Without that head start, Labour replacement Mark Hutchinson will not be able to retain the seat against new National candidate Katie Nimon.

Polling is broadly in line with this outcome (Curia, 20 Aug: 🔵Nat 35, 🔴Lab 27, 🟡ACT, 17), but saw a much higher than expected share for ACT’s Pawel Milewski. Given the decline in ACT support since mid-August, it would be very surprising to see their canddiate run that far ahead of the party vote.

No candidates will be elected off the list.

Nelson

🔴Likely Labour hold

Candidate vote

🔴Lab – Boyack (list 42nd) 47% – odds 90%

🔵Nat – Cameron (35th) 37% – 10%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 31

🔴Lab 27

🟢Gre 16

🟡ACT 10

⚫NZF 7

🔝TOP 4

Another race which sees a Labour class-of-2020 candidate defending their seat against a new National challenger, Nelson will probably see Labour incumbent Boyack hold on. However, if Boyack sees a worse-than-average incumbency bonus, National’s Cameron could take it.

A baseline Labour-Green leaning seat (combined PVI+5), National were able to hold the seat consistently from 1996-2020 thanks to immovable object Nick Smith’s superlative performances with Labour and Green voters.

The general trend away from National in 2017 would probably have seen even these efforts fail, were it not for a (reasonably successful) two-tick campaign by the Greens. In 2020, even a respectable 17% with Labour voters wasn’t enough.

With Smith retiring to the mayoralty, Boyack faces an easier but not guaranteed path. No candidates are electable list positions.

New Lynn

🔴Safe Labour hold

Candidate vote

🔴Lab – Russell (list 22nd) 51% – odds 100%

🔵Nat – Garcia (34th) 34% – 0%

🟢Gre – Abel (9th) 9% – 0%

History

🔴🔴🔴🔴🔴🔴🔴🔴🔴

Party vote

🔴Lab 35

🔵Nat 30

🟢Gre 13

🟡ACT 8

⚫NZF 6

🔝TOP 2

Another safe red seat that will see the Labour incumbent (Russell) re-elected. At 34^th on the list, National’s Garcia will in all likelihood miss out, but at 9^th on his party’s list Green candidate Abel should be elected after a narrow miss in 2020.

New Plymouth

🔵Likely National GAIN

Candidate vote

🔵Nat – MacLeod (list n/a) 52% – odds 85%

🔴Lab – Bennett (29th) 41% – 15%

History

🔴🔴🔴🔴🔵🔵🔵🔵🔴

Party vote

🔵Nat 43

🔴Lab 21

🟡ACT 11

🟢Gre 9

⚫NZF 8

🔝TOP 4

Yet another of Labour’s new incumbents in a struggle to hold on, New Plymouth will most likely see National’s David MacLeod take over from Labour’s Glenn Bennett. Labour’s precipitous drop over the course of the year probably dooms Bennett, but there is some cause for hope.

Firstly, the absence of a Green candidate and the presence of an ACT one boosts his chances. Secondly, if he can over-perform as an incumbent with National and ACT voters then he might just pull it off.

If he does, National’s MacLeod is unranked so will miss out. Similarly, if MacLeod wins as expected, Bennett will likely miss out at 29^th on the Labour list.

North Shore

🔵Safe National hold

Candidate vote

🔵Nat – Watts (list 17th) 58% – odds 100%

🔴Lab – Hampton (41st) 21% – 0%

History

🔵🔵🔵🔵🔵🔵🔵🔵🔵

Party vote

🔵Nat 47

🔴Lab 14

🟡ACT 14

🟢Gre 13

⚫NZF 5

🔝TOP 4

A deeply blue (PVI+12) seat with a National incumbent holds little drama. No candidates will be elected from the list.

Northcote

🔵Lean National GAIN

Candidate vote

🔵Nat – Bidois (list 60th) 44% – odds 77%

🔴Lab – Halbert (28th) 40% – 23%

History

🔵🔴🔴🔵🔵🔵🔵🔵🔴

Party vote

🔵Nat 39

🔴Lab 24

🟢Gre 14

🟡ACT 10

⚫NZF 6

🔝TOP 4

Another replay of the 2020 contest with a former National incumbent (Bidois) looking to reclaim the seat from current Labour incumbent Halbert.

Neither candidate was an especially strong performer with the other party’s voters in 2020, and given both have held the seat, neither will probably benefit from incumbency advantage. The seat is slightly blue-leaning overall (PVI+4) and in a blue year Bidois should win out.

Neither candidate is in an especially electable list position, nor are any minor party candidates.

Northland

🔵Solid National GAIN

Candidate vote

🔵Nat – McCallum (list n/a) 47% – odds 99%

🔴Lab – Prime (9th) 24% – 1%

⚫NZF – Jones (2nd) 16% – 0%

🟡ACT – Cameron (7th) 3% – 0%

🟣DNZ – King (1st) 3% – 0%

History

🔵🔵🔵🔵🔵🔵(🔵/⚫)🔵🔴

Party vote

🔵Nat 37

🔴Lab 18

🟡ACT 12

⚫NZF 12

🟢Gre 11

🟣Con 3

A three-way race in a very red year saw Labour’s Prime just manage to claim the seat from National’s King by less than 1%. Her odds of a repeat are vanishingly small, but a complicated four-ish way race leaves the remotest path to victory.

First, NZ First deputy leader and TikTok sensation Shane Jones is again trying to claim the seat in more favourable circumstances for the party. Nevertheless, he is no Winston Peters when it comes to ability to attract National and Labour voters, so will not win the seat.

The other wildcard is 2020 National candidate now DemocracyNZ leader Matt King. Polling and fundamentals modelling show he won’t really be a factor, but outsider candidates with name recognition are very hard to predict.

Polling (Curia, 10 Sep: 🔵Nat 49, 🔴Lab 20, ⚫NZF 15, 🟣DNZ 5, 🟢Gre 5, 🟡ACT 3) is broadly in-line with expectations.

Prime (9^th), Jones (2^nd) and ACT’s Cameron (7^th) are all in what should be electable positions. National’s McCallum is unranked.

Ōhāriu

🔴Lean Labour hold

Candidate vote

🔴Lab – O'Connor (list n/a) 39% – odds 68%

🔵Nat – Willis (2nd) 36% – 32%

🔝TOP – Hammond (6th) 12% – 0%

🟢Gre – Rodgers (20th) 8% – 0%

History

🟣🟣🟣🟣🟣🟣🟣🔴🔴

Party vote

🔵Nat 31

🔴Lab 27

🟢Gre 19

🟡ACT 9

🔝TOP 6

⚫NZF 4

Ōhāriu sees two-term Labour incumbent (Greg) O’Connor face off against National deputy leader Nicola Willis, who has moved up the hill from Wellington Central for a more realistic shot at an electorate seat.

Fundamentals favour O’Connor. The seat is baseline Labour-Green leaning (combined PVI+9, and net +6% over National/ACT even in a bad year), and O’Connor has averaged a respectable 9% with National voters in 2017 and 2020.

Despite her ascent within the party, Willis was not an especially strong candidate with Labour or Green voters in Wellington Central, although now in a race where she stands a real chance that may change.

The final complicating factor is repeat TOP candidate Hammond, who managed 11% in 2020 attempting to occupy the centrist niche once held by United-Future’s Peter Dunne.

If she loses, Willis is safe on the list at 2^nd. O’Connor is unranked so relies on the seat. At 20^th on the Green list, Rodgers is the last Green candidate with a mathematical (but very small) chance to election.

Ōtaki

🔵Solid National GAIN

Candidate vote

🔵Nat – Costley (list 64th) 48% – odds 90%

🔴Lab – Ngobi (46th) 40% – 50%

History

🔴🔴🔴🔴🔵🔵🔵🔵🔴

Party vote

🔵Nat 38

🔴Lab 26

🟢Gre 11

🟡ACT 9

⚫NZF 9

🔝TOP 3

Another bellwether seat (6^th most representative in 2020) Ōtaki sees a rematch between now-incumbent Ngobi of Labour and unsuccessful 2020 National candidate Costley. In 2020, an underwhelming 1% with National voters wasn’t enough to cost Ngobi, but 2023 will need improvement.

Given the swing back to National (and Ōtaki’s close following of the nationwide result, on current polling the seat is Costley’s to lose, but if she can hang on to enough departing swing voters, Ngobi has a small chance.

Welcome to Issue 1.1 of the Overhang I guess?

I had planned to do some kind of proper introductory post. To lay out what my goal is with all this and what you can expect. But that was too much like hard work. Something else grabbed my attention. So we’re starting here.

tl;dr

Support for the minor parties is historically high. The gap between Labour and National is historically low. This is historically unusual.

The last time this happened was the late 1990s, in the early days of MMP. Those elections (1996 and 1999) came after a period of upheaval and radical change, but were then followed by a period of major party consolidation.

Ultimately, 1996 and to a lesser extent 1999 are better templates for the 2023 election than 2002 or 2005.

Chaotic coalitions

Both commentators and politicians have been calling out coalitions of chaos or of cuts or of colonisers. Support for the major parties remains low, and the outcome of the election remains in doubt. This isn’t unusual. It’s an election year. Even after 30 years, people still pretend not to know how MMP works.

I’m not the only one to have noticed this (four people did while I’ve been working on the draft) and it’s kind of…meh.

But the thing is we’re not wrong to notice something's up.

Both the left and the right can credibly claim their opponents might form a government of three or more parties, and with significant minor party representation. This is unusual. It’s only really happened twice before and is the jumping-off point for this post.

Where we are

Since the ascension of the two Christophorae, polling has settled into a stable pattern:

• The gap between Labour and National remains low.

• The share held by minor parties remains high.

Post-Luxon, the gap between National and Labour has averaged a little over 3%. Post-Hipkins, it has averaged just 1%. This is well below the long-run average of 12%.

Over this whole period, the share held by minor parties has averaged around 30%. In recent months it has climbed as high as 35%. This is largely on the back of respectable numbers for ACT, the Greens, and more recently Te Pāti Māori (but with a non-negligible share taken by the rest). Again, this is above the long-run average of 22%.

On their own neither of these situations is novel or interesting. National and Labour have run close before. The minor parties have done well before.

What is weird is both happening at the same time.

Boundaries of the chaos zone

I first tried making this point a while ago with a scatter plot comparing the major party gap and the minor parties’ share of polling since the the start of the MMP era.

Apparently that was hard to follow. What it does is break the polling space down into four quadrants to help define partisan electoral dynamics at any given point.

The quadrants are, going clockwise:

• Major party collapse one of either National or Labour collapsing and shedding support to their opponent and to the minors (above average gap, above average minor share)

• Single party dominance one of the majors dominating their opponent and the minor parties (above average gap, below average minor share)

• Two party contest National and Labour in a close race at the expense of the rest (below average gap, below average minor share)

• The Chaos Zone a close contest between National and Labour, high percentages for minor parties (below average gap, above average minor share)

The right-hand quadrants can be further broken down by which of National or Labour is collapsing/dominating.

The amount of time we’ve spent in each quadrant is not equal. The largest share of the time since 1993 (37%) has been spent in a two-party contest: National and Labour both looking credible and absorbing most of the support. The next most common are phases where either one major party is collapsing (27%) or is sweeping the field (24%).

The rarest situation - and the one we’re in now - is an all-round close contest (14%).

Where we’ve been

Plotting the data above over time might make things clearer.

Combining this view with the quadrants (or “phases”) above periodises recent political history and the partisan dynamics we’ve experienced.

That’s still too many colours at once, so we’ll take each in turn…

Major party collapse

Starting in the top-right corner, we have periods of major party collapse.

Technically there have been seven intervals of this, but really only four (two for each major party)…

February 1995 - September 1996 The lead up to the 1996 general election, where it really looked for a minute like Labour was going to lose it’s status as a major party, and become just one of several (along with the Alliance and NZ First) opposed to the Fourth National government. Clark’s subsequent rebuild in the final stretch of 1996 and the lead up to 1999 is one of recent political history’s underrated achievements.

July 2002 - December 2002 This covers the 2002 election and its immediate aftermath, after both National’s collapse under English and Peter Dunne’s tango with The Worm. Worth noting (in contrast to the 2020 episode below) that peripheral National support had four places to flee to (Labour, NZ First, United Future, and ACT) rather than just two.

On and off from December 2011 - July 2017 Labour’s Era of Many Davids, from Phill Goff’s defeat at the 2011 election through to the resignation of Andrew Little (spiritually a David). Despite a few glimpses of a close contest, this was a period where Labour shed support to the Greens on the left and to NZ First on the right. Not as dramatic as the pre-1996 edition, but much more extended.

October 2020 - October 2021 The back-half of National’s COVID/Collins collapse, where National remained in the depths, but more to the benefit of ACT than of Labour (as noted, unlike in 2002 there were fewer places for the looser National vote to go).

Single-party dominance

Next up are periods where one of the major parties lords it over everyone else.

Five-ish main instances, and much longer for National than for Labour…

February 2000 - May 2000 The Clark government’s honeymoon period, ended by a period of Shipley revival during the ‘winter of discontent’ (sidenote: what we thought was a serious disturbance in the early 2000s seem so quaint now).

February 2002 - June 2002 The early lead-up to the 2002 general election, around the time of the implosion of the Alliance and Bill English’s well…

December 2007 - November 2011 The Key government’s brief engagement to the electorate and then their long, long honeymoon.

Intermittently from June 2012 - December 2015 The bits of the back-end of the 5NG where Labour were still a basket case, but the minor parties weren’t featuring as much.

April 2020 - September 2020 COVID and all it entailed.

Two party contest

Where we’ve spent the plurality of the time (and where it felt like we were heading post-Luxon until we just… didn’t): acting like it’s still the FPTP days.

Five (but really only four) instances…

May 1997 - July 1999 The lead-up to the 1999 general election where somehow the contest remained close, despite Helen Clark reviving her party’s fortunes, National going through a coup, and the National-NZF coalition blowing up in Loony Toons fashion.

June 2000 - November 2001 The abovementioned false dawn of the second Shipley era, during the early Clark years.

January 2004 - November 2007 Ōrewa to ‘08: the tense end to the Clark years.

Two brief, fleeting gasps in March 2013 and September - December 2013 A short blip where Shearer appeared to be getting his shit together, and then Cunliffe’s early bump after rolling him.

August 2017 - March 2020 Jacindamania through the start of the pandemic.

The Chaos Zone

Okay (after a long diversion through a bunch of history that was maybe besides the point but that I find interesting) the actual thing I’m supposed to be writing about.

Things fall apart, the centre cannot hold, the falcon cannot hear the falconer, the old world is dying, the new world struggles to be born, now is the time of backbench MPs acting up and acting out.

November 1993 - January 1995 The beginning of the MMP era. The nation awoke bleary-eyed from the 1993 general election and realised not only had Irish Jim managed to pull it off, but we’d also chosen a new electoral system. Okay in reality this episode was a continuation of the pre-MMP, post Roger and Ruth crack-up of the two-party system. (And yes I know the result wasn’t clear the morning after).

October 1996 - April 1997 A transitional phase after the 1996 general election, as rising support for Labour crossed over with declining support for NZ First and the Alliance. Unclear whether this should really count.

August 1999 - January 2000 The lead up to and aftermath of the 1999 general election where Labour had not yet consolidated the left, with a freshly-independent Green Party and a bloodied but unbroken NZ First asserting themselves alongside the Alliance and ACT.

November 2021 - Present *gestures around*

What does it mean?

Things could get better, or they could get worse. It’s hard to say.

The 1996 and 1999 elections followed a decade of reform, disruption, and austerity. We looked like we were headed down a European path: genuine multi-party democracy, a Pasokifying social democratic party, and a rising populist right.

Low support for the then incumbent National party showed clear dissatisfaction with the status quo. At the same time, the initial lack of a coherent alternative from a Labour party still getting over what they did to themselves and to the country in the 80s meant the major opposition party didn’t benefit.

The electorate were casting about for something - anything - as an alternative. Minor parties rose, fell, split, and merged. The majors cracked-up with defections, in-fighting, and coups (attempted and successful).

Sound familiar?

But then it didn’t happen.

Contrary to a subsequent pattern of major parties shedding support over election campaigns, Labour’s position improved over 1996. Then National stitched together a largely continuity government despite all Winston Peters’ bluster.

Longer term, first Clark then Key forged a modified Westminster equilibrium on the back of a third-way mixed economy powered by the China boom, milk money, and low global interest rates.

So maybe that’s where this goes. Things get better, or at least more Normal. Like Bolger in 1996, Labour keep the nine-year cycle going by cobbling together one more win. Then, after a period of disquiet, one of the majors (probably National, given where we are in the cycle) is able to consolidate around a new set of modifications to the consensus. The institutional fundamentals (more on this in a future post) that dictate a two-and-three-thirds party system will win out.

But maybe they won’t.

Firstly, at a micro level, neither Hipkins nor Luxon seem to have the talents of Clark or Key. The pattern of minor parties rising with greater attention during the election campaign could reassert itself, and we could end up with our most diverse (in partisan terms) parliament yet.

Secondly, there’s no centrist minor party exercising an amphoteric effect on the majors helping to maintain the status quo between governments (unless TOP can pull it off).

Finally, the macro forces of the 2020s and 2030s will be destabilising rather than stabilising. While the Clark-Key consensus was partly down to their policies and personal abilities, it was also a product of the end of history. Hipkins, Luxon, and whoever follows them will not be so fortunate.

Events have and will continue to transpire.

The disruption we’ve seen through the COVID pandemic, supply chain disruption, geopolitical instability, inflation, and now climate emergencies isn’t going away. If the major parties insist things are still Normal, the electorate will look for answers in other (sometimes weirder) places.

All of which is to say:

Congratulations to anyone who made it this far. I didn’t plan on writing 2000 words on this, but we’re a one-man-show, so lack editorial discipline. Future posts will have a shorter, more current-events style focus in the lead-up to October. But I can’t promise I won’t also do rabbit-hole pieces like this one.

Any feedback/critique/character assassination is welcome, I’m still working out what this is for, and what other (more normal) people are interested in.

Ka kite anō (hopefully)

Rustie

Subscribe now