Shadow Accounting · Oil-Blood Conversion Model · v1.0

THE BLOOD
PRICE OF OIL

deaths_per_barrel = total_attributable_deaths / proven_reserves_contested
blood_price_usd = (oil_price_usd / barrels_per_liter) × (deaths_per_barrel × liters_per_person)

Sources: Brown Univ. Costs of War · Iraq Body Count · Lancet · USGS · EIA · IBC · WHO
Global oil/day (2024)
102.6M
barrels/day · EIA 2024
Post-9/11 direct deaths
940K
Brown Univ. CoW · 2001–2023
Indirect / structural deaths
4.5–4.7M
Brown Univ. CoW · same period
Iraq proven reserves
145B
barrels · EIA / OPEC 2025
Midpoint deaths/barrel (Iraq)
1/480K
300K deaths ÷ 145B bbl
Shadow blood price
$0.001–$0.03
per liter · vs medical $880/L
War Dataset
n=8 oil-linked conflicts · deaths: low / mid / high estimate · reserves contested in billions of barrels
Conflict Period Deaths (mid) CI Low CI High Reserves (Bbbl) Deaths/barrel (×10⁻⁶) $/liter blood Source tier
Deaths/Barrel over Time
scatter with 95% CI bands · OLS regression line · log scale y-axis
OLS Regression Results — Deaths/barrel ~ Year (log-transformed)
Slope (β₁)
+0.0421
log-deaths/bbl per year
Intercept (β₀)
at year 1991
variance explained
p-value (β₁)
H₀: slope = 0
Estimation Uncertainty
mean · std dev · 95% CI across three accounting tiers
// Three-tier model — each tier is a different accounting boundary
TIER_1 = direct_war_deaths / proven_reserves_contested
         = 300,000 / 145,000,000,000 = 2.07×10⁻⁶ deaths/bbl

TIER_2 = extraction_labor_deaths / production_volume (US, 2003–2013)
         = 1,189 / 47,000,000,000 = 2.53×10⁻⁸ deaths/bbl

TIER_3 = indirect_deaths_attributed / cumulative_production (era)
         = 100,000,000 / 1,500,000,000,000 = 6.67×10⁻⁵ deaths/bbl

// Blood price conversion
blood_price = (oil_usd/bbl × deaths/bbl × liters/person) / (liters/bbl)
            = (75 × D/bbl × 5,000L) / 158.99L

// Gap: medical blood market = $440–880 USD/liter
// shadow price = $0.001–$0.031 USD/liter
externality_gap = medical_price / shadow_price ≈ 14,000×–880,000×
Shadow Price per Liter of Blood
log-normalized bars across accounting tiers vs. medical market
[1] War deaths: Brown Univ. Costs of War 2023. Iraq Body Count (conservative floor). Lancet 2006 (upper bound). Midpoints used for regression.   [2] Reserves: EIA, OPEC 2024–2025. Iraqi proved reserves 145Bbbl.   [3] Extraction labor deaths: CDC/NIOSH 2003–2013. US only. Global multiplier est. ×8 (ILO).   [4] Indirect deaths: WHO air pollution attribution (fossil fuels, 20% oil share). Climate mortality: Lancet Countdown.   [5] Oil price: Brent spot ~$75/bbl (March 2026).   [6] Blood volume/person: 5 liters.   [7] Barrels→liters: 1 bbl = 158.987L.   [8] Medical blood price: Red Cross / hospital acquisition cost $440–880 USD/liter.   [9] Regression: OLS on log(deaths/bbl) ~ year. n=8 conflicts. CIs from Gaussian propagation of estimate uncertainty bounds.   MODEL v1.0 · The price mechanism is specifically structured to prevent this conversion. The gap is the architecture.