Continuum fiscal point estimates averaged 35%% of actual government program costs in the Carney calibration (March 2026). Root causes: estimates measured policy intervention minimums not capital program costs; no Arctic cost multiplier; no delivery mechanism differentiation; no confidence intervals. This proposal activates the range estimation framework, delivery mechanism multipliers, and split risk scoring that correct for systematic underestimation bias.

Epoch: 111

Domain: governance

Fiscal cost estimate (LLM): $0.02B CAD

Structural estimate (RIPPLE): +$3.83B CAD net (v3-bfs-signed depth=2, decay=0.5/hop; diverges)

Top RIPPLE cost paths

• +$4.22B → canada_health_transfer (Canada Health Transfer (CHT)) via gdp_growth
• +$3.89B → consultant_spending_federal (Federal Spending on Outside Consultants) via delivery_mechanism_cost_index
• −$2.42B → budgetary_balance (Budgetary Balance (Deficit/Surplus)) via fiscal_sustainability_index
• −$1.86B → federal_budget_balance (Federal Budget Balance) via fiscal_sustainability_index

Variable changes

• proposal_calibration_delta: {"new": 0.38, "old": 0.65}
• delivery_mechanism_cost_index: {"new": 0.72, "old": 0.45}
• fiscal_estimate_accuracy_ratio: {"new": 0.62, "old": 0.35}