Pollution & Extraction Wealth & Labor Tech & AI Submit a Question
Global data — three topics — no spin

The world
by the
numbers.

Public data from verified sources, in one place. No agenda. The figures that shape daily life — and where they come from.

Explore by topic
🌍 Pollution & Extraction Emissions, water, food waste, extraction 📊 Wealth & Labor Inequality, wages, water access, labor Tech & AI Energy, water, automation, ownership
The cost
of production.
Animal ag. emissions / year
9.8Gt
CO₂ equivalent — 57% of all food system emissions globally
US share of global transport emissions
20.2%
US produces over a fifth of all global transport emissions
Food wasted globally
1.05B t
35% of all food produced in the US goes unsold or uneaten — every year.
Water per kg beef wasted
25,000L
The embedded water discarded every time one kilogram of beef goes uneaten
US food waste annual cost
$218B
Spent growing, processing, and disposing of food Americans never eat
Top 20 corps — global emissions
35%
Estimated share of all industrial greenhouse emissions attributable to 20 companies since 1965
Agriculture's freshwater share
70%
Of all global freshwater withdrawals go to agriculture — animal farming accounts for the dominant share in most regions
Water to produce 1kg of beef
15,400L
Freshwater required on average to produce 1kg of beef — vs. 1,250L for 1kg of wheat. The gap is embedded in every meal.
Water use by sector — global annual withdrawal (km³)
Agriculture total
2,700 km³ — 70% of all use
Animal farming (est.)
~1,700 km³
Industry
~800 km³ — 20%
Municipal / domestic
~400 km³ — 10%
All data centers
~0.3 km³
Agricultural emissions trajectory — 2000–2040 projection (GtCO₂eq/yr)
Historical
Projected (current trajectory)
Projected (reform scenario)
Context

National emissions figures measure where production happens — not who profits from it. Climate Trace now attributes emissions to the operating entity rather than the geography, producing a meaningfully different picture of accountability. Water tells the same story: agriculture withdraws 70% of global freshwater, but water scarcity risk falls hardest on communities that didn't choose those supply chains. The waste built into retail supply chains is a structural feature, not an anomaly — suppliers guarantee full shelves by design, making overproduction the default outcome. Every kilogram of beef that goes uneaten also discards roughly 15,400 liters of embedded water.

Who keeps
what.
Top 1% US wealth share
32%
Of all US wealth held by the top 1% of households
Bottom 50% US wealth share
2.5%
Of all US wealth held by the bottom half of the population
CEO-to-worker pay ratio
350:1
Compared to 21:1 in 1965. Worker productivity has also roughly doubled over the same period.
Billionaire wealth growth 2020–23
+$2.2T
Increase in combined billionaire wealth during COVID-19
Racial wealth gap — median
8:1
White median household wealth (~$188k) vs. Black median (~$24k)
Productivity vs. wage growth 1979–2020
+64%
US productivity growth since 1979. Real wages for the bottom half: essentially flat.
Workers without retirement savings
~55%
Estimated share of Americans with no meaningful retirement savings — across all racial groups
Americans without safe water access
~2M
People in the US lacking running water or basic indoor plumbing — concentrated in low-income, rural, and indigenous communities
US wealth distribution
Top 1%
32% of all wealth
Next 9%
~38% of all wealth
Next 40%
~27%
Bottom 50%
2.5%
Context

The ultra-wealthy can access loans at 0.5–3% interest using appreciated assets as collateral — living on borrowed money that isn't taxable income, while the assets continue to grow. When those assets transfer at death, the accumulated gains are permanently excluded from capital gains tax. This strategy is legal, documented, and structurally unavailable below a certain wealth threshold. The racial wealth gap compounds the picture: decades of documented housing, lending, and labor policy exclusions created gaps that market participation alone does not close.

The energy
behind
the screen.
Infrastructure & Energy
Projected data center demand 2030
945TWh
Terawatt hours — roughly equivalent to Japan's total national electricity consumption today
US data center emissions 2023
105Mt
Million tons CO₂ equivalent — 2.18% of all US emissions, with 56% from fossil fuels
Global nuclear investment 2024
$80B
Doubled from $40B in 2020 — driven by AI energy demand from major tech companies
Work tasks at risk from automation
~30%
McKinsey Global Institute, 2024. The economic gains flow primarily to whoever owns the technology.
Ownership & Labor
Big 3 cloud market share
66%
AWS, Microsoft Azure, and Google Cloud control roughly two thirds of global cloud infrastructure
AI vs. animal ag. emissions
~100×
Animal agriculture currently produces ~100× more CO₂ equivalent than AI data centers — but AI is growing faster
Data center water use — US 2023
600B gal
Used annually for server cooling across US data centers — Google alone used ~5.6B gallons in 2022, enough for 100,000 homes
Congo mineral wealth
$24T
DRC contains most of the world's cobalt — essential for EV batteries and smartphones. One of the poorest nations by GDP per capita.
AI energy footprint vs. other industries (CO₂eq/yr)
Animal agriculture
9.8Gt
Transportation
8.4Gt
All data centers
~0.9Gt
AI specifically
~80Mt — growing
Data center electricity demand — 2020–2035 projection (TWh/yr)
Historical
Projected — high AI growth
Projected — moderate growth
Context

AI's current environmental footprint is considerably smaller than animal agriculture or transportation — but the trajectory is the story. Data center demand is projected to roughly quadruple by 2034, and water consumption for cooling scales with it. Whether that growth runs on renewables and nuclear or on fossil fuels — and whether cooling systems shift to air or water — are open policy questions with real local consequences. Three companies control two thirds of the infrastructure and are now investing directly in nuclear capacity. The Congo cobalt figure follows the same pattern: the physical cost of the digital economy falls heaviest on the people least compensated by it.

Ground Truth
A question.
A correction.
A source.

This site is built on public data and human judgment — both of which have limits. If you've spotted an error, have a better source, or want to ask something, use the form.

Submissions are reviewed periodically. No account required. Your email is only used to respond — never shared or stored beyond that.