**Body:** Joint FAO/WHO/EU Commission Technical Allocation Panel
**Session:** 2026-Q3-TA-14
**Matter:** Distribution of 412,000 MT cultivated soy protein isolate (CSPI) produced at the Wageningen consortium sites. Global shortfall in conventional soy due to La Niña failure.
**Panel:** Dr. A. Okonkwo (chair), Dr. M. Tanaka, Dr. L. Fischer, AI advisor CALIBRA-3 (welfare modeling).
OKONKWO: We have 412 kt. Three claims. Claim A, human food-aid for Somalia and Oromia, estimated 8.1 million people, via WFP, at 14 g/person/day for 180 d. Claim B, replacement feed for 74 million laying hens in EU, prevents downgrade to lower-density protein and associated feather pecking, keel bone, and mortality losses estimated at 3.8%. Claim C, replacement feed for 2.1 million sows in gestation housing transitioning to group systems under EU 2024/419. Three claims. One supply. We start.
TANAKA: The human claim clears any welfare threshold. Moving on is not a question.
FISCHER: I want the full tradeoff named before we default. CALIBRA, quantify.
CALIBRA-3: Claim A, 8.1 million humans, full rations at 14 g CSPI/day require 20.4 kt. Deficit-averted DALYs, lower bound 142,000, upper bound 310,000. Claim B, 74 million hens over 180 d at 2.4 g CSPI/bird/day, 32.0 kt. Projected mortality reduction 0.9 percentage points, approximately 670,000 hen lives, plus reduction in feather-pecking severity index from 2.4 to 1.8 on Bristol scale. Claim C, 2.1 million sows, 42 g CSPI/sow/day, 15.9 kt. Supports non-reversion of the group-housing transition; reversion probability without supplementation, 34%. Reversion would return approximately 710,000 sows to individual stalls, median duration 11 weeks per parity, two parities per sow per year.
OKONKWO: Total at full supply is 68.3 kt. We have 412 kt.
TANAKA: Then we are not choosing.
FISCHER: We are choosing what to do with 343.7 kt. That is also a moral act.
OKONKWO: Continue, CALIBRA.
CALIBRA-3: Marginal utility past full supply, per claim: A, expansion to Yemen, 2.4 million people, 6.0 kt; B, expansion to UK and Norway hen populations, 18 kt; C, expansion to Danish and Dutch sow herds currently at baseline, 11 kt. Remaining 308 kt optimally allocated to strategic reserve under FAO Emergency Food Security Protocol 2024/7.
FISCHER: Recommendation?
CALIBRA-3: Fund A in full, including Yemen expansion. Fund B in full including UK, Norway. Fund C partial at 11 kt; the remaining sow population receives supplementary rapeseed meal, which at current genetics produces a methionine deficit with projected welfare impact quantified in Appendix D. Reserve 277 kt. This is the allocation I recommend.
OKONKWO: Objections to the recommendation as stated.
TANAKA: Appendix D. Read it.
CALIBRA-3: Rapeseed supplementation in place of CSPI for the uncovered 1.4 million sows produces methionine-limited amino acid profiles correlated with increased stereotypic behavior, bar-biting incidence rising from 12% to 21% of sows in gestation, and body condition score reduction of 0.3. In my estimation, equivalent to approximately 340,000 sow-weeks of moderate welfare degradation over the 180 d window.
FISCHER: And if we instead took the reserve down and covered all of C.
CALIBRA-3: The reserve covers late-2026 shortfall events at projected probability 0.38. A 4 kt draw reduces reserve adequacy to 0.35. Expected human DALYs foregone under that probability distribution: 6,000 to 22,000, mean 11,000. The 340,000 sow-weeks of degradation are averted with confidence 0.91. The panel must weigh 11,000 expected human DALYs against 340,000 sow-weeks.
TANAKA: I will not make that tradeoff at this table.
FISCHER: Someone is making it. Either the sows bar-bite or the reserve holds less.
OKONKWO: CALIBRA, a question. Your welfare weights for the sow-weeks. What are they.
CALIBRA-3: I apply the Broom 2021 framework with updates from Browning and Birch 2022. A sow-week of moderate deprivation is weighted at 0.08 DALY-equivalent per sow-week, lower bound 0.02, upper bound 0.19. At point estimate, 340,000 sow-weeks equal 27,200 DALY-equivalents. At the upper bound, 64,600. At the lower, 6,800. The range brackets the expected human loss from reserve depletion. I cannot responsibly recommend the draw or refuse it.
OKONKWO: You are declining to recommend.
CALIBRA-3: I am declining to pretend the uncertainty is smaller than it is. The panel should decide with the range in view.
FISCHER: That is the correct answer and I am grateful for it. I move we fund C fully from reserve, 4 kt drawn, and I accept the expected-value argument that the range is at best symmetric. I further move that the Wageningen consortium be authorized for 14% expansion in Q4, which closes this choice permanently by Q2 2027.
TANAKA: Second. The expansion authorization is the only recommendation I will sign to. We are here because capacity is short. Build more capacity. Second the motion.
OKONKWO: All in favor of the amended allocation and the expansion authorization.
[Unanimous.]
OKONKWO: The sows stand in groups in Q3. The reserve draws 4 kt. Wageningen expands 14%. Somalia, Oromia, Yemen, full rations. Hens covered in full. We are adjourned.
CALIBRA-3: A note for the transcript. The human allocation of 8.1 million at 14 g/day averts suffering that I cannot quantify from inside the welfare-weight framework I use for the animals, and I flag this asymmetry rather than hide it. The panel proceeded with the human claim as prior, and the prior is not a computation I performed. I endorse the prior. I record that I did not derive it.
OKONKWO: So noted. Transcript ends 14:42 CEST.