The protein system
Nearly 18.6 million people live in the thirteen states we assess.[5] Weighted by protein consumed rather than by population, Papua New Guinea accounts for about 47 per cent of the region and New Zealand for about 38 per cent, with Fiji at 6 per cent and the remaining ten states sharing the last 9 per cent.[1][5] Any regional judgement is therefore mostly a judgement about two countries, while the ten smallest members carry under a tenth of the weight and present most of the fragility.
Protein supply runs from 54.5 grams per person per day in the Solomon Islands and 62.2 grams in Papua New Guinea to 129.3 grams in Tonga and 119.0 grams in the Marshall Islands, with New Zealand at 100.9 grams.[1] High supply and high dependence coincide in Polynesia and Micronesia: 67.8 of Tonga's 129.3 grams comes from meat, of which domestic production supplies 14 per cent.[1] Nine of the ten heaviest nations on earth are Pacific island states,[33] while Papua New Guinea carries child stunting near 50 per cent.[14]
Production is concentrated in New Zealand, which in 2024 produced 21.5 million tonnes of milk, 741,844 tonnes of beef, 449,606 tonnes of sheep meat and 230,307 tonnes of poultry for 5.29 million people,[2][5] set records in the 2024/25 season at 1.94 billion kilograms of milksolids from 4.68 million cows,[25] with NZ$28.6 billion of dairy export revenue forecast for the year to June 2026.[24] Island protein comes from root crops and village pigs, reef and lagoon fish, and imported meat. Papua New Guinea's 2.06 million pigs producing 77,843 tonnes of pork are the largest livestock protein source outside New Zealand,[2] and fish supply runs from 7.1 kilograms per person per year in Papua New Guinea to 72.3 kilograms in Kiribati against a global average below 21 kilograms.[1]
Above all of it sits the western and central Pacific tuna fishery, which landed 3,059,005 tonnes in 2024, the largest catch on record and 56 per cent of the global tuna catch, more than 80 per cent of it in coastal-state waters.[6] That resource is the region's principal sovereign asset, and almost none of it reaches Pacific plates.
Durable domestic capacity: 65
Quantitative score 63, adjusted up 2 points
New Zealand's capacity is close to the global frontier. Our protein-weighted calculation from FAOSTAT food balance sheets puts its production at 258 per cent of the protein its population eats and 368 per cent on animal protein alone.[1] Its fisheries are among the better governed in the index, with 87.2 per cent of the 149 stocks of known status above the soft limit (130 stocks), representing 97.0 per cent of assessed landings by weight.[27]
Papua New Guinea is the second pole and is routinely misread. Bourke and Harwood attributed 83 per cent of national food energy and 76 per cent of protein to local agriculture,[13] and our 2023 FAOSTAT calculation independently returns a protein self-sufficiency ratio of 80 per cent and an animal protein ratio of 84 per cent.[1] Built on sweet potato, sago, taro, banana and village pigs with almost no purchased inputs, that system is durable against trade shocks and nutritionally inadequate at the same time.
Below those two, capacity thins fast. Vanuatu and the Solomon Islands retain self-sufficiency ratios of 96 and 89 per cent on low absolute supply, and Fiji holds 73 per cent on poultry, with dairy reduced to a 13.7 per cent ratio.[1][2] Samoa has fallen to 23 per cent and Tonga to 34 per cent, at animal protein ratios of 10 and 14 per cent,[1] and the atoll states have no meaningful terrestrial capacity at all.
We apply +2. Food-balance-sheet ratios credit the tuna resource to nobody, because the fish is exported, yet all four target stocks were assessed as neither overfished nor subject to overfishing in 2024, with skipjack spawning biomass at 51 per cent of unfished levels.[6] A managed protein stock of that scale inside the entity's own waters is real capacity the consumption data cannot register.
Resource headroom: 59
Quantitative score 56, adjusted up 3 points
New Zealand has run out of room. The sheep flock fell from 29.80 million head in 2014 to 23.58 million in 2024 and cattle from 10.37 million to 9.52 million.[2] A statutory cap of 190 kilograms of synthetic nitrogen per hectare per year has applied to pastoral land since July 2021,[28] and nitrogen use has fallen with it.[3] Growth in dairy protein now comes from imported material: palm kernel expeller imports hit a record 2.6 million tonnes in 2025 against a six-year average near 1.9 million tonnes,[30] and roughly 63 per cent of grain and feed supply is imported.[31] The base pasture system is rain-fed and grass-fed, while the marginal litre of milk is fed on Indonesian and Malaysian by-product.
Papua New Guinea is the region's one reservoir of agronomic headroom, applying 20.75 kilograms of nitrogen per hectare of cropland against abundant land and rainfall.[3] On paper that reads as expansion room; in practice, customary tenure over approximately 97 per cent of the national land area,[57] thin road networks outside a few corridors and negligible extension and input-supply infrastructure mean it is not demonstrably mobilisable at the pace the raw indicator implies, and Papua New Guinea carries 47 per cent of the region's consumption weight.
The three evidence series, irrigation, agricultural credit orientation and cropland change, confirm that reading: Papua New Guinea has no FAOSTAT record of land equipped for irrigation at any date, and its agricultural credit orientation index has averaged roughly a tenth of parity across 2013-2024, both corroborating the tenure and infrastructure constraint.[58][59] Its cropland and arable land have genuinely grown, 33.8 and 61.0 per cent respectively since 2000, a faster rate than New Zealand posted over the same window, so this is not a zero-mobilisation resource.[58] But population grew 87.6 per cent over the same period, nearly doubling, so cropland per person fell 28.7 per cent even as the total expanded.[60] A World Bank and EU-financed smallholder programme, the Productive Partnerships in Agriculture Project, disbursed US$95.18 million between 2010 and 2021 for cocoa and coffee value chains and closed with an outcome rating of Satisfactory, 60,494 registered households and yields at 142 to 208 per cent of target,[61] evidence that financed mobilisation is achievable in Papua New Guinea under a partnership model built around customary landholders rather than around land reform. New Zealand's own irrigated area nearly tripled since 2000 to 780,500 hectares in 2023, exceeding its entire arable land base because irrigation is concentrated on dairy pasture, and its credit orientation index has stayed above parity throughout the same period at 1.60 in 2024,[58][59] confirming that New Zealand's constraint is regulatory rather than water or finance, exactly as already stated above.
Coastal fisheries have already passed their limit. Roughly half of reef fisheries across island countries are overexploited, and 16 of 22 Pacific island countries and territories are projected to be unable to meet current per-capita fish consumption from coastal stocks by 2030.[12] That matters more than land scarcity, because reef fish is what outer-island households actually eat.
We apply +3 net. The tuna resource earns +5: the largest and among the best-managed protein stocks on earth, inside the region's EEZs, and the only growth channel that requires no imported feed, held below the maximum because the same biomass is projected to shift eastward under warming, with a 13 per cent average decline across ten Pacific SIDS EEZs by 2050 under high emissions, a 20 per cent fall in purse-seine catch and US$90 million a year of lost access fees.[7] Against that we apply a realisability discount of -2 to Papua New Guinea's nominal cropland headroom, for the reasons set out above.
Import exposure: 61
Quantitative score 66, adjusted down 5 points
On trade ratios the region looks ordinary. Food is 12.5 per cent of New Zealand's merchandise imports, 16.5 per cent of Papua New Guinea's, 19.0 per cent of Fiji's and 23.7 per cent across Pacific island small states as a group, rising to 41.4 per cent in Kiribati.[4] On protein the picture splits. New Zealand's animal protein self-sufficiency ratio is 252 per cent, so a total import halt would remove bread, rice and around 60 per cent of the pork it eats[32] while leaving milk and red meat intact. Papua New Guinea would lose rice, at 1.9 per cent cereal self-sufficiency. Tonga and Samoa would lose almost all animal protein within weeks, at ratios of 14 and 10 per cent, and Nauru and Tuvalu produce effectively none, with cereal self-sufficiency of zero in six member states.[1]
The region's structural advantage over the Gulf is geography: no significant volume of Pacific food transits Hormuz, Malacca, Suez or Panama, and the suppliers are Australia and New Zealand, three to ten days away.
We apply -5. Consumption weighting places 85 per cent of the region's protein in the two insulated members and understates a genuine concentration risk. For the eleven smaller states the physical channel is one liner rotation out of Auckland or Suva, some receive only 40 to 50 container-ship port calls a year, and Pacific SIDS pay roughly twice what developed countries pay to move imports.[16] Tonga drew 25.1 per cent of all imports from New Zealand in early 2026,[39] and New Zealand takes 89 per cent of its wheat from one supplier.[31]
Upstream dependence: 52
Quantitative score 56, adjusted down 4 points
This is the region's weakest pillar. New Zealand made only 96,600 tonnes of the 368,000 tonnes of nitrogen it applied in 2023, imports all of its potash, and manufactures superphosphate from imported rock, including 110,500 tonnes taken from occupied Western Sahara by Ballance Agri-Nutrients in 2024.[3][29] Since Marsden Point closed in April 2022 it has imported 100 per cent of its refined fuel, 48 per cent by value from Korea and 33 per cent from Singapore,[23] and because those refineries run on Gulf crude, roughly 60 per cent of New Zealand's fuel originally transited Hormuz.[21] Statutory cover is 28 days of petrol and 21 of diesel.[22]
Papua New Guinea scores best in the region here for the same reason it scores modestly on capacity: its subsistence food system uses effectively no purchased inputs, total nitrogen imports were 25,215 tonnes in 2023,[3] and the Napa Napa refinery provides 32,500 barrels a day of domestic refining.[38]
The small island states are the floor. Oil products supply around 80 per cent of the region's total energy and imported diesel generates over 90 per cent of electricity in some island systems.[16] Outboard motors, ice plants and cold chains all run on that fuel, so the coastal fishery feeding these countries is itself an imported-input system.
We apply -4: the 2026 crisis has demonstrated this failure mode rather than modelled it, and no fertiliser or feed statistic captures it.
Access and affordability: 39
Quantitative score 42, adjusted down 3 points
Only one of the region's thirteen members has a published FAO/World Bank Cost and Affordability of a Healthy Diet headcount, and it is the region's hub rather than its poorest state: 58.0 per cent of Fijians could not afford a healthy diet in 2025, down from a peak of 70.1 per cent in 2021 and still above the 50.6 per cent of 2017, against a healthy diet costing 4.75 PPP dollars a day.[45] The World Bank's copy of the same series has no reading for the other twelve members, New Zealand and Papua New Guinea included.[45] Fiji carries only 5.0 per cent of the region's population, so its measured 58.0 per cent is not a regional composite; New Zealand (28.5 per cent) and Papua New Guinea (56.9 per cent), the two largest members, plus ten smaller island states (9.6 per cent) have no measured equivalent at all, a 95.0 per cent coverage gap.[5] Papua New Guinea's widely quoted 32 per cent is an unverified secondary rendering of FAOSTAT, and the country has no entry at all in the World Bank's copy of the FAO series;[43][45] it stands as unverified context, corroborated in direction by undernourishment of 28.7 per cent, child stunting near 50 per cent, and an IFPRI wage survey finding a healthy household diet costs 37 to 45 per cent of a low-skill worker's daily wage across seven PNG locations, worst in Port Moresby,[46][14][50] and carries no weight in the affordability figure or the indicators below. With no usable regional affordability headcount, we derive the pillar as a population-weighted judgement across members, pricing each population block on its own measured evidence: Papua New Guinea (56.9 per cent of the population) on undernourishment of 28.7 per cent, stunting of 49.5 per cent on a 2010 survey with UNICEF estimating 45 to 50 per cent in 2023, and the IFPRI finding that a healthy household diet costs 37 to 45 per cent of a low-skill daily wage, equivalent to the low 20s on this pillar;[46][14][50] New Zealand (28.5 per cent) by high-income analogue to Australia's measured 3.2 per cent unaffordability, equivalent to the mid 80s;[45] Fiji and the ten smaller island states (14.6 per cent) on Fiji's measured 58.0 per cent, positioned at 38 within its band, supported by stunting of 34.8 per cent in the Marshall Islands (2017), 31.7 per cent in the Solomon Islands (2015) and 28.9 per cent in Vanuatu (2013).[46][47] The population-weighted result is 42. The FAO's blended "Oceania" figure of 20.1 per cent unaffordable includes Australia and is context rather than evidence here.[44] New Zealand looks strong on the outcome checks that exist, with undernourishment of 2.5 per cent and a healthy diet costing 3.71 PPP dollars a day,[46][45] but its own Health Survey found 27.0 per cent of children lived in households where food ran out often or sometimes in 2023/24, up from 21.3 per cent, rising to 54.8 per cent of Pacific children and 34.3 per cent of Maori children.[49] Fiji is the sharpest warning against reading affordability off outcome data: undernourishment of 6.8 per cent and stunting of 7.2 per cent look mild while nearly three-fifths of Fijians cannot afford a healthy diet.[46][47][45] Cheap imported starch and oil hold the quantity indicators up while diet quality gives way, the pattern behind the region's obesity burden.[33] Solomon Islands and Vanuatu combine moderate undernourishment with stunting above 28 per cent,[46][47] while price volatility has run into double digits across Palau, Vanuatu, Tonga, Samoa and Kiribati since 2022.[48] Physical access compounds this: the same 40-to-50 port-call shipping rotations that constrain imports also limit what reaches outer islands and PNG's Highlands, where much of the region's stunting is concentrated.[16]
We apply -3, for New Zealand's documented divergence from the high-income analogue used above: its own Health Survey found children in households where food ran out often or sometimes rising from 21.3 to 27.0 per cent in a single year, reaching 54.8 per cent of Pacific children and 34.3 per cent of Maori children.[49] No separate coverage deduction is taken: the 95.0 per cent coverage gap in the affordability data is priced once, inside the block derivation itself, and is recorded in the data gaps.
Shock endurance: 63
Quantitative score 60, adjusted up 3 points
New Zealand's endurance is structural rather than stockpiled. It holds no strategic food reserve. A 2.5-fold protein surplus means a world price spike raises national income: food price inflation eased to 2.5 per cent in the year to June 2026, down from 4.5 per cent in February, through the worst of the fuel shock.[34] Its weaknesses are a supermarket duopoly holding about 90 per cent of grocery sales and fuel cover of 33.6 days of petrol and 30.0 days of diesel on 31 March 2026.[21]
Papua New Guinea's gardens absorb trade shocks and fail against drought. The 2015-16 El Nino affected around 2.7 million people, roughly 480,000 of them in the most severe food-shortage category,[15] and a strong El Nino is again considered possible in late 2026.[17]
The island states have the worst demonstrated record. Cyclone Pam destroyed almost the entire national banana crop in Vanuatu in 2015,[35] Cyclone Winston caused up to 100 per cent crop damage in the worst-hit Fijian districts in 2016,[36] and Cyclone Harold affected 65 per cent of Vanuatu's population in 2020.[37] In 2026 a fuel shock alone was sufficient: the Marshall Islands saw its fuel import bill rise about US$40 million, equal to 11.5 per cent of GDP, and cut FY26 growth from 4.1 to 2.0 per cent;[18][20] Tuvalu declared a state of emergency on 14 April;[19] Fiji's ministers took a 20 per cent pay cut.[17]
Crisis purchasing power is a net positive for most of the region. Papua New Guinea held US$4,062.2 million in gross reserves at the end of 2025, worth 7 months of total import cover and 11 months of non-mineral import cover against the IMF's three-month benchmark for its income group.[53][56] Fiji, Solomon Islands, Tonga, Vanuatu and Samoa hold reserve cover of between 4.4 and 12.6 months on the latest available data.[52] New Zealand's strength here is market access rather than reserve depth, because its own cover of 3.90 months in 2025 is the thinnest in the region;[52] what carries it is an S&P AA+ foreign currency and AAA local currency rating and a Moody's Aaa that let it borrow through a shock at a price the smaller states cannot reach.[54] Fitch affirmed New Zealand at AA+ in March 2026 while revising the outlook to Negative, the first sign that even this buffer has a limit.[55] Formal social protection that could scale is the weak link: World Bank ASPIRE records meaningful programme coverage for only two of the region's thirteen members, leaving fisheries rent, trust funds and external partners to do the work a domestic safety net would otherwise do.[51]
We apply +3 for two further funded, operating mechanisms the indicators miss. Sovereign fisheries rent gives the smallest states real purchasing power in a price shock: Kiribati collected US$210.2 million in licence fees in 2024, about US$1,560 per head and more than 70 per cent of fiscal revenue.[9][10][11][41] Alongside it sits a tested external response capability, with the Pacific Islands Forum emergency mechanism invoked in 2026,[17] World Bank rapid financing for the Marshall Islands in June,[20] and Australian and New Zealand disaster response that has delivered within days of every major cyclone. The adjustment is held below the maximum because the same rent is projected to fall by up to US$140 million a year as tuna move east.[8]
Trajectory
Three things decide where the region goes next. The first is domesticating tuna: the region controls 56 per cent of the world's tuna catch and eats almost none of it, and redirecting even a small share into island markets through nearshore fish aggregating devices, cold-chain investment and domestic-landing conditions in access agreements would raise capacity and import scores together. SolTuna shows the potential and the limit: about 25,000 tonnes processed a year, largely for export.[42] The second is energy: the 2026 shock has made electrification a food-security question, because in most of these states the cold chain, the fishing fleet and the freight all sit downstream of imported diesel. The third is New Zealand's input base, where domestic urea capacity, deeper fuel reserves and reduced palm kernel expeller dependence would each raise the upstream score, and with it the regional score, given New Zealand's 38 per cent weight.
Working against all three is climate, which is projected to move the tuna east, degrade the reef fisheries that feed outer islands, and raise cyclone intensity where one storm can remove a national crop. We expect the region's position to hold, with the internal spread widening.
Country notes
New Zealand. An exceptionally strong producer on its own: production at 258 per cent of protein eaten and 22.1 per cent of world dairy exports,[1][26] with land fully allocated and every intensifying input imported.[3][30][31]
Papua New Guinea. The heaviest weight and most misread member: 76 per cent of protein from local agriculture[13] and 84 per cent animal protein self-sufficiency,[1] against stunting near 50 per cent.[14]
Fiji. The regional hub, at 87.0 grams of protein per day and 75 per cent animal protein self-sufficiency on poultry, with dairy at a 13.7 per cent ratio,[1][2] and the region's only measured affordability headcount, at 58.0 per cent unable to afford a healthy diet in 2025.[45]
Solomon Islands. The lowest protein supply at 54.5 grams per day, yet 89 per cent self-sufficient on root crops, pigs and fish;[1] access fees fund about 10 per cent of revenue.[8]
Vanuatu. The strongest small-island performer on capacity at 96 per cent protein self-sufficiency,[1] undermined by shock frequency rather than structure.[35][37]
Samoa. The clearest production collapse: protein self-sufficiency 23 per cent, animal protein 10 per cent, pig herd halved since 2014.[1][2]
Tonga. The highest protein supply at 129.3 grams per day and 14 per cent animal protein self-sufficiency, despite the largest arable endowment at 0.19 hectares per person.[1][5]
Micronesia (Federated States of). Protein supply of 72.0 grams per day and zero cereal self-sufficiency, while 166,000 tonnes of tuna are landed from its waters and exported.[1]
Marshall Islands. Arable land of 0.01 hectares per person, effectively zero animal protein self-sufficiency, and the sharpest 2026 fuel shock at 11.5 per cent of GDP.[5][18][20]
Palau. No FAOSTAT balance sheet exists: it imports close to all its food[40] and has forgone tuna rent by closing 80 per cent of its EEZ from 2020.
Kiribati. Leverage without access: 72.3 kilograms of fish per person a year and US$210.2 million of licence revenue funding over 70 per cent of the budget.[1][9][10][4]
Nauru. Weakest on every terrestrial measure, with zero animal protein self-sufficiency and a mined-out interior; fishing licences are now the fiscal base.
Tuvalu. The most fragile member: zero animal protein self-sufficiency, A$33.7 million of a A$47.1 million domestic revenue base from licences,[11][41] and a state of emergency on 14 April 2026.[19]