Three signatures in four months describe a shift the "US-versus-China" frame cannot hold. New Zealand and the Cook Islands signed a defence and security declaration in April 2026, the Cook Islands naming Wellington its "partner of choice" and repairing a rift that opened when Rarotonga signed cooperation deals with Beijing in early 2025. On 22 July, in Port Moresby, New Zealand and Papua New Guinea signed a statement of intent to expand joint maritime surveillance, exercises and personnel exchanges. Two weeks earlier, on 6 July, Australia and Fiji had already gone further, sealing a mutual-defence treaty — Fiji's first alliance. The desk's read is structural: these are middle powers organising among themselves, thread by bilateral thread, rather than waiting to be organised by a great-power patron. Analysts, and only analysts, supply the connecting motive — that the common object is to keep China from embedding security and policing arrangements across the islands. The signatories themselves speak of regional cooperation and studiously avoid naming Beijing. That gap — between what is signed and what it is said to mean — is the story, and the reason to log a prediction against it rather than assert one.
Start with the chart, not the communiqué. Draw the sea lanes that run north from Australia and New Zealand through Melanesia and out into the wide Polynesian triangle, mark the exclusive economic zones no island state can patrol with the boats it owns, and the pattern of the past four months stops looking like coincidence. It looks like a perimeter being drawn — not from Washington, and not for the first time in the region's history, but this time by the middle powers who live inside it.
1. Three signatures, four months
The individual facts are firm. In April 2026, in Rarotonga, New Zealand and the Cook Islands signed a defence and security declaration in which the Cook Islands named New Zealand its "partner of choice" on defence and security, and Wellington resumed development funding — reported at NZ$29.8 million — that it had paused a year earlier. The rift being repaired was concrete: it opened when the Cook Islands signed a set of cooperation agreements with China in early 2025, including on deep-sea minerals, which Wellington read as a strategic tilt in a state with which it shares a constitutional relationship. Established
On 22 July 2026, in Port Moresby, New Zealand's Defence Minister Chris Penk and Papua New Guinea's Defence Minister Elias Kapavore signed a statement of intent on defence cooperation for 2026–2029: expanded joint maritime surveillance, the deployment of assets into PNG waters, military exercises, disaster response, search and rescue, officer training and personnel exchanges. Established
And on 6 July 2026, in Suva, Prime Ministers Anthony Albanese and Sitiveni Rabuka signed two instruments: the Vuvale Union, deepening economic integration, and the Ocean of Peace Alliance, a mutual-defence treaty committing each to the other's aid — described as Fiji's first alliance and Australia's fourth. Established We flag this because a security lattice assembled from mutual-defence commitments is a heavier object than one assembled from surveillance memoranda, and the Australia–Fiji tie is the load-bearing one.
2. What the map shows that the frame hides
Read as a set, three things stand out, and none of them is visible from inside a two-body "US-versus-China" model.
The first is who is doing the drawing. The lead actors here are New Zealand and Australia — middle powers, not the superpower — and the counterparties are small island states with their own agency, not clients being assigned to a camp. The United States is not absent from the wider Pacific contest, but it is not in the frame of any of these three instruments; none was signed in Washington, brokered by Washington, or announced alongside it. Assessed A structure in which the second-tier powers self-organise, rather than deferring to the hegemon to organise them, is a different kind of order than the alliance-spoke model the region is usually described through.
The second is the shape of the thing. This is not a bloc and not a treaty organisation; it is a mesh of overlapping bilateral ties — Wellington–Rarotonga, Wellington–Port Moresby, Canberra–Suva, with Wellington now eyeing the Canberra–Suva frame. Lattice, not alliance. A mesh has no single door to guard and no headquarters to capture, which makes it cheaper to extend and harder to unwind — but also looser, more deniable, and slower to act in concert than a formal pact. Assessed
The third is the currency the lattice trades in, which is not warfighting but presence: maritime-domain awareness, surveillance flights, personnel exchanges, port and training access, the resumption of aid. The scarce commodity in the island Pacific is the ability to see and patrol vast maritime zones, and it is precisely that capacity these agreements pool. Assessed Whoever supplies the boats, the radar and the training becomes the default security partner — which is the quiet prize both Beijing and the incumbents are competing for.
3. The motive everyone reads in — and no one signs
Here the labelling has to be strict, because this is where reporting most often launders inference into fact. The claim that the common thread of these pacts is to counter China's expanding security and policing links in the Pacific is an analytical inference. It is not a stated position of any government that signed them.
The people making that read are named and credentialed. Estelle Denton-Townshend, a teaching fellow in international security at the University of Waikato, told the South China Morning Post that "security agreements and treaties in the Pacific are attempts to ensure that China is unable to establish security agreements with Pacific states." Al Gillespie, an international-law professor at the same university, told regional outlets that more such partnerships are likely and would move "throughout the region," and that the trend was "unlikely to be welcomed by Beijing." Assessed
The signatories say something narrower. Chris Penk framed the PNG agreement in the language of resilience — "in an increasingly volatile world, Pacific nations like ours are stronger when we work closely together" — and did not name China. Established That is not evasion to be ignored; it is data. Small island states have their own reasons to avoid being cast as anyone's front line: they trade with China, take its infrastructure finance, and guard their non-aligned standing carefully. The distance between the analysts' framing and the ministers' is itself a fact about how much weight the "countering China" reading can bear. The desk's position is that the China-hedging interpretation is the most plausible single explanation for the timing and the clustering — but "most plausible" is an assessment, and the honest label is Assessed, not Established. Assessed
4. What to watch
The tell for whether this is a genuine lattice or a run of unrelated bilateral housekeeping is extension. A lattice grows nodes and cross-links; a coincidence does not. Watch three things. First, whether New Zealand formalises a tie into the Australia–Fiji framework, turning three separate threads into a connected mesh. Second, whether the next island state to sign — Tonga, Vanuatu, Solomon Islands, Tuvalu are the candidates most often named — does so with a Canberra or Wellington counterparty rather than a Beijing one. Third, whether any of these instruments acquires teeth beyond surveillance and training: basing rights, standing joint patrols, or a mutual-assistance clause of the kind the Australia–Fiji treaty already carries. Assessed
The counter-signal worth equal attention: a Pacific state signing a fresh security or policing arrangement with China in the same window would show the lattice is porous — that the middle powers are drawing a perimeter the islands themselves decline to hold. The contest is not settled by the three signatures already on paper. It is settled by the fourth, the fifth, and by whose name is on them.
Three defence signings in four months — Wellington with the Cook Islands and with Papua New Guinea, Canberra with Fiji — are individually modest and collectively a structural signal the standard frame misses: the Pacific's middle powers are assembling their own security lattice, bilaterally and on their own initiative, rather than waiting on Washington to organise the region for them. What they are building is established fact. That they are building it to fence out China's policing reach is a credible inference by named analysts — not a purpose any signatory will state, and the islands' own careful silence is part of the evidence. Judge the pattern by whether it spreads.
By 31 March 2027, the lattice will visibly extend by at least one of two moves: (a) New Zealand formally joins, or signs an instrument tying it into, the Australia–Fiji defence framework; or (b) Australia or New Zealand signs a new bilateral defence, security or maritime-surveillance agreement with a further Pacific Island state that has not previously held one with them (candidates including Tonga, Vanuatu, Tuvalu or Solomon Islands). At least one of the two occurs within the window.
Confidence: Medium-to-High (Assessed). The basis is the cadence already on display — three signings inside four months — plus named analysts' explicit expectation that more partnerships are coming, and New Zealand's own floated interest in the Australia–Fiji arrangement. The principal ways this resolves wrong: a change of government in Canberra or Wellington that cools the push; an island state instead deepening ties with Beijing in the same window, stalling the Western lattice; or the existing pacts proving to be discrete housekeeping with no appetite for extension. Resolution date: 31 March 2027, scored against announced and signed agreements as reported by primary government sources and wire services.