The bill we haven't paid

NOAA's Billion-Dollar Weather and Climate Disasters tracker: 2024 produced 27 separate US disasters each causing $1B+ in damages, with total damages of approximately $182 billion. The 10-year average is $140B/yr, up from ~$55B/yr in the 1990s (inflation-adjusted). The rising count and cost are the single clearest, hardest-to-contest measurement of the attributable damage from climate change.

EIA / Our World In Data: US per-capita CO₂ emissions are 14.4 tonnes per year as of 2024, still the highest among G20 economies. China: 7.5 t. EU: 6.4 t. India: 2.0 t. US emissions fell from a 2007 peak (~20 t/person) as coal retired and efficiency improved, but the per-capita level remains roughly 2× the EU and 2× the world average.

Copernicus / WMO / NASA GISS: 2024 became the first calendar year in which the global surface temperature anomaly averaged above +1.5°C vs pre-industrial baseline, the threshold targeted by the Paris Agreement. Crossing it in a single year does not technically trigger the long-term 'breach' (which requires sustained multi-year averages), but every major climate body reported the threshold as functionally reached. Multiple subsequent months of 2025 have continued near or above the mark.

A record year. Historical average (1980-2000): ~4 billion-dollar US disasters/year. 2000-2020 average: ~13/year. 2020-2024 average: ~22/year. Insurance industry projections anticipate 30+ annual billion-dollar events as the new normal under current warming. NOAA's tracker was paused by the Trump administration in early 2025 'pending review'; multiple scientific societies and a bipartisan group of governors have petitioned for its continuation.

EIA: 21% of US electricity generation in 2024 came from renewables (wind 10%, solar 6%, hydro 5%, biomass/geothermal 0.5%). Natural gas: 43%. Coal: 16%. Nuclear: 18%. Renewable share rose from 6% in 2000 and 10% in 2015. China's 2024 share: 33% renewables. EU: 47%. Germany: 57%. The US is the world's largest economy and among the slower-decarbonizing of major grids.

Lazard 2024 Levelized Cost of Energy: unsubsidized utility-scale solar at $29-$38/MWh, onshore wind at $27-$73/MWh. Combined-cycle gas: $45-$108/MWh. Coal: $68-$166/MWh. New nuclear: $142-$222/MWh. Solar + 4-hour battery storage is competitive with new gas plants in most of the US. The economics of new-generation decarbonization flipped in favor of renewables around 2018-20; subsequent deployment has been gated by permitting, transmission, and (in 2025-26) policy headwinds rather than cost.

Kelly Blue Book / Cox Automotive: battery-EV share of US new-vehicle sales was ~8% in 2024, up from 2% in 2020. Global average: ~18%. China: ~40%. Norway: >90%. US adoption is gated by policy, charging infrastructure, and automaker-lineup choices rather than price (EV total cost of ownership has been below ICE for midsize vehicles since ~2023). OBBBA's 2025 repeal of the $7,500 federal EV credit is projected to slow adoption.

The Inflation Reduction Act (August 2022) established the largest federal climate-spending program in US history: $370 billion over 10 years in clean-energy tax credits, manufacturing subsidies, and rebates. Most of the provisions were repealed or phased out in OBBBA (July 2025), including EV credits, the residential solar ITC, heat-pump rebates, and the wind/solar production tax credit for new projects. Rhodium Group projects US emissions ~500 million tons higher through 2035 than they would have been under full IRA implementation.

CDC Heat Illness Tracker: emergency-department visits for heat-related illness have risen 117% between 2018 and 2024, with the largest increases in Southern and Southwestern states. Heat is now the deadliest weather-related cause of death in the US (more than hurricanes, floods, tornadoes combined). Counts are widely understood to be undercounts because heat exacerbation of cardiovascular and kidney disease is rarely coded as heat-related on death certificates.

US Census Bureau Household Pulse Survey + Internal Displacement Monitoring Centre: roughly 1.2 million Americans per year are displaced by climate-related events (hurricanes, wildfires, floods). Displacement durations range from days to permanent. The term 'climate refugee' is not recognized in US law; displaced residents rely on FEMA Individual Assistance ($700M-$4B/year) and SBA disaster loans. Florida, Louisiana, and California account for ~60% of annual displacements.

CRS / OMB: federal disaster-relief spending (FEMA, SBA, HUD CDBG-DR, Agriculture) totaled approximately $1 trillion across 2017-2024, averaging ~$125B/yr. The 2017 and 2022 hurricane seasons and the 2025 wildfires were the largest single-year drivers. FEMA's Disaster Relief Fund has required supplemental congressional appropriations in 6 of the last 7 fiscal years. The fund is projected to be insolvent for peak-season-scale events without structural reform.

Florida avg home insurance premium is ~$4,200/yr (3× national average); state-backed Citizens Property holds 772,000 policies as of Sept 2025. California's FAIR Plan (insurer of last resort) filed in October 2025 to raise premiums by ~36% due to wildfire claims. Major private insurers have exited or restricted underwriting in both states. See the housing page for the full insurance data; the story is the private insurance market's repricing of climate risk, which shifts loss exposure first to state backstops and ultimately to taxpayers.

Dominion Energy's 2024 Integrated Resource Plan disclosed that data centers consume approximately 25% of Virginia's electricity, the highest share in any US state. Northern Virginia is the world's largest data-center cluster; new builds added ~3 GW of contracted load in 2024 alone. Dominion's IRP projects 6-7× growth in data-center demand by 2039, requiring new gas plants the utility had not previously planned. Virginia residential customers have seen rate increases approved in 2024 and 2025 explicitly tied to grid build-outs serving hyperscaler facilities.

The 2025/26 PJM capacity auction cleared at $269.92/MW-day vs $28.92 the year before, a 9× jump driven by data-center load colliding with coal retirements. The aggregate cost of the auction outcome is roughly $14.7 billion, billed back to retail customers across PJM's 13-state footprint (Mid-Atlantic + much of the Midwest, ~65 million people). Monitoring Analytics, PJM's independent market monitor, attributed the spike to capacity scarcity created by AI-driven demand growth outpacing new generation interconnection. The cost-shift is the cleanest single number on what residential ratepayers are paying for the AI capex cycle.

Sherwood News (Apr 29 2026) reports 14 states are actively considering moratoriums on new data-center builds, with residential electricity prices projected to rise 15-40% over the next 5 years across affected grids. Local pushback has run ahead of federal policy: Indianapolis IN extended its 2024 moratorium through 2025; Prince William County VA paused approvals on the 23,000-acre 'Digital Gateway' corridor and Fairfax, Loudoun, and Stafford counties tightened zoning rules; Atlanta-metro suburbs Dunwoody, Peachtree Corners, and Norcross GA passed full moratoriums while Marietta GA imposed a study pause; Tucson AZ effectively blocked water-intensive builds via a 2024 water ordinance, with Chandler and Mesa AZ raising water-use thresholds; Memphis TN has open public-health complaints over xAI's Colossus gas-turbine emissions. State consumer advocates in MD, NJ, and OH have filed FERC complaints arguing the PJM auction cost-shift is unjust. The externality is land + water + grid, all governed by the city, county, and state authorities federal AI policy doesn't touch.

Sherwood News (Apr 29 2026): Alphabet, Amazon, Meta, and Microsoft are spending >$650B on data-center construction in 2026, with sector electricity demand projected to roughly double by 2030, enough to power France + Germany combined. ~2,300 GW of generation and storage is stuck in interconnection queues, more than the entire installed US power-grid capacity. The bottleneck is no longer capital; it's grid + transformers + GPUs + memory + CPUs. The mismatch creates the political opening for state moratoriums (see above): hyperscalers have been making 500 GW of community commitments while only ~50 GW of physical buildouts is actually achievable through 2030.

Roughly 3.6 GW of coal generation that utilities had previously committed to retire by 2025-2027 has been deferred into the 2030s, with hyperscaler load growth cited as the operational driver. AEP Ohio's Mitchell plant, Indianapolis Power & Light's Petersburg, and Dominion's Clover all pushed retirement dates. Each year of additional coal operation produces ~5-8 Mt CO₂ per GW; the cumulative emissions cost of the 3.6 GW deferral is roughly 18-30 Mt CO₂ per year, equivalent to adding ~4-6 million cars to US roads.

After a 13-month public-records lawsuit by The Oregonian, the City of The Dalles disclosed in 2022 that Google's data centers consumed approximately 355 million gallons of water in 2021, about 25% of the city's total water use. The disclosure is the cleanest documented case of a hyperscaler's share of a municipal water supply. Similar conflicts are open in Goodyear AZ (Microsoft, Meta), Mesa AZ (Google, Meta), and across the Texas Hill Country. Hyperscalers typically negotiate water rights and tax abatements through non-disclosure-protected economic development agreements; The Dalles case set the precedent that the rights are public information.

CDP / Climate Accountability Institute Carbon Majors database (2017): 100 fossil-fuel producers (state and investor-owned) are responsible for ~71% of global industrial greenhouse-gas emissions since 1988. ExxonMobil, Shell, BP, Chevron, and a handful of state oil companies (Saudi Aramco, Gazprom, NIOC) sit at the top of the list. The producer side of the externality is much more concentrated than the consumption side, which is why supply-side policy (production caps, fossil-fuel subsidy reform) is structurally more efficient than demand-side policy at the same political cost.

Supran, Rahmstorf & Oreskes (Science, 2023): of 16 internal ExxonMobil climate projections from 1977-2003, 63%-83% were consistent with subsequently observed warming, with a median projected warming rate of 0.20°C/decade ± 0.04, at least as skillful as the contemporaneous independent academic and government models. The same period was characterized by ExxonMobil's public communications questioning the reality, attribution, and certainty of climate change. The internal-versus-external gap, documented in their own technical archive, is the cleanest single piece of evidence that the public-facing doubt was a strategic position rather than a scientific one.

Brulle (Climatic Change, 2014): 140 US foundations made 5,299 grants totaling roughly $558 million to 91 organizations comprising the climate-countermovement infrastructure (denial think tanks, contrarian media, advocacy groups). The combined annual income of those organizations averaged ~$900 million across the period. Direct ExxonMobil and Koch family contributions are documented but have decreased as a share over time, replaced by donor-advised funds and 'dark money' channels (Donors Trust, Donors Capital Fund) that obscure the original source. The infrastructure is what made the public communications gap above operational; the funding is what made the infrastructure persist.

OpenSecrets: oil & gas industry federal lobbying totaled approximately $154.5 million in 2024, with 745 registered lobbyists and ~70% of those lobbyists being former federal employees. The American Petroleum Institute (API) is the largest single industry-funded vehicle. The lobbying spend has run between $120M and $155M annually for the past decade, peaking with each major energy-policy or tax-policy fight (Obama-era cap-and-trade, IRA, the 2025 IRA-rollback fights). It is one of the most consistent industry-lobbying spends in Washington and substantially exceeds combined renewable-energy industry lobbying.

IMF Working Paper 2023/169 (Black, Liu, Parry, Vernon): the US 'total subsidy' figure for fossil fuels in 2022 was approximately $757 billion ($3B explicit, $754B implicit). The implicit component prices the externality (climate damage, local air pollution, road accidents, congestion) at IMF social-cost estimates. Direct US tax-code subsidies (intangible drilling cost deductions, percentage depletion, foreign tax credit gaming, accelerated depreciation, MLP tax treatment) run an additional ~$20B per year per EESI tracking. Globally the total reached $7 trillion in 2022, ~7.1% of world GDP. The producer industry is the largest single recipient of unpriced public subsidy in the global economy.

ITEP (2024) analysis of corporate filings: the US oil-gas-pipelines sector paid an average federal effective tax rate of 2.0% on US profits across 2018-2022, against a statutory rate of 21%. Specific 2021 figures: ExxonMobil 2.8%, Chevron 1.8%. Combined 2022 IFRS net income for the five Western majors (ExxonMobil, Chevron, Shell, BP, TotalEnergies) was approximately $152 billion (the chart above plots the IFRS-attributable series; BP's reported 2022 figure was dragged $25B+ by a Rosneft impairment, so underlying-profit headlines circulate at ~$199B). 2023 fell to ~$113B and 2024 to ~$84B as crude prices receded. Even as profits compressed, those same five paid out roughly $114 billion in dividends and share repurchases in 2023 (IEEFA), more than their reported net income. Profits at the top, payouts to shareholders, taxes near zero, and the implicit externality on the previous row is what the public is paying instead.

The US first exited in 2020 under Trump's first-term executive order (signed 2017, took effect 2020); rejoined under Biden in 2021; re-exited January 20, 2025 under Trump's second-term executive order. The off-on cadence, which is unique among major emitters, makes multi-decade private investment in US-sited decarbonization capital costly to underwrite. Most credible multi-decade capital allocation decisions now assume US federal climate policy is zero until proven otherwise.

July 2025: the Interior Department revoked more than 3.5 million acres of federal waters previously designated for offshore wind leasing. The Department separately paid TotalEnergies $982 million to cancel two approved wind projects in New York and North Carolina, with terms requiring the company to redirect the funds into US fossil-fuel investments. It is an explicit policy of moving private capital from one energy type to another using taxpayer money as the transfer agent.

EPA Administrator Zeldin announced in May 2025 a plan to reduce EPA staffing to "Reagan-era levels," roughly 1-in-3 current career staff. 300+ had departed via resignation, firing, or buyout by mid-2025. Project 2025 called for EPA to be "largely gutted," for all grants to advocacy groups to stop, and for specific regulations (the endangerment finding, PFAS limits, methane rules) to be rescinded. Most are in process or finalized as of late 2025.

The Trump administration paused NOAA's 45-year-old Billion-Dollar Weather and Climate Disasters database in early 2025. It is the authoritative historical record of US climate-attributable economic damage and the source most commonly cited in this space. A bipartisan group of governors and multiple scientific societies have petitioned for its restoration; the administration has cited 'review' without a timeline. Insurance industry and academic users have begun building private mirrors, which fragments the standard measurement that underwriters and policymakers rely on.