Climate Costs as the World's Largest Unpriced Externality

The William Nordhaus carbon-price work, the Stern Review, the IPCC mitigation scenarios, and EPA's social cost of carbon all converge on the same uncomfortable fact: greenhouse gas emissions impose costs that markets do not price, at a scale that conventional Pigouvian remedies have struggled to match.

Climate change is the largest unpriced externality in the global economy. Carbon emissions impose costs on every other country and on future generations that the emitter does not pay. The economics is straightforward, and the political economy is the part that has been unsolvable for forty years. The conventional Pigouvian fix — tax carbon at its social cost globally — has been impossible to implement at the scale the problem requires.

The Scale of the Externality

Total global CO2 emissions in 2024 ran above 37 billion metric tons. Methane and other greenhouse gases add a roughly equivalent CO2- equivalent burden. The atmospheric stock has risen from ~280 parts per million pre-industrial to ~420 ppm today, and the temperature consequences are already producing measurable damages — heat-related mortality, agricultural disruption, sea-level rise, ocean acidification, extreme weather frequency.

Estimating the dollar value of these damages requires a damage function (how much harm does each additional ton of CO2 cause?) and a discount rate (how do we compare damage today against damage in 2080?). Different modeling choices produce wildly different numbers. The US Interagency Working Group's 2021 central estimate was $51 per ton at a 3% discount rate. The EPA's 2023 revision raised it to $190 at a 2% discount rate. The European Investment Bank uses €100 as a working estimate. Nicholas Stern's 2006 review used closer to $200. Each estimate is defensible; the choice among them is the policy.

Why It Is the Hardest Externality

Climate change has features that make it the most difficult Pigouvian problem ever posed:

Global pool: every emission contributes to a single atmospheric stock. Unilateral action by one country benefits all countries.
Free-rider incentives: each country has an incentive to underinvest in abatement because the benefits are diffuse and the costs are local.
Long time horizons: emissions today produce damages decades later, which makes the political-economy calculus extraordinarily difficult.
Distributional asymmetry: countries that emitted historically are not the same as countries that will bear the damages, which creates fairness disputes that disrupt collective action.
Irreversibility: certain damages (sea-level rise, permafrost thaw, biodiversity loss) cannot be undone once they occur.

Each of these features individually would make policy hard. The combination has produced thirty years of largely failed international negotiations.

What Has Been Tried

The UNFCCC framework (1992) and the Kyoto Protocol (1997) attempted binding international targets. The US never ratified Kyoto; the agreement covered only roughly 15% of global emissions. The Paris Agreement (2015) replaced binding targets with nationally determined contributions, which are weaker but politically achievable. Every country submits its own plan; nobody is bound to ambitious action.

Sub-national carbon markets — the EU ETS, the California program, the RGGI northeast US system — have produced measurable emission reductions in covered sectors but have not extended to global scale. National carbon taxes — Sweden's $130/ton, the UK's carbon price floor, the BC carbon tax — have worked where implemented but have not spread to most countries.

Subsidies for clean alternatives — the US Inflation Reduction Act, the EU Green Deal, China's massive support for solar and EV manufacturing — have arguably done more for emission reduction than the carbon-pricing track. They work by making low-carbon technology cheaper rather than by making fossil fuels more expensive, which has turned out to be politically much more durable.

The Carbon-Pricing vs. Subsidy Debate

Economists historically favored carbon pricing as more efficient. The political record has shown that subsidies are more durable. William Nordhaus's Nobel-Prize-winning work pioneered the carbon-pricing framework; his policy recommendations have been implemented in only a few jurisdictions. Mariana Mazzucato and others have argued that the subsidy and industrial-policy track has actually produced more measurable emission reductions because it sidesteps the political-economy obstacles to pricing.

The pragmatic accommodation that has emerged in the major economies is a mix: modest carbon pricing where politically feasible, substantial subsidies for clean alternatives, regulatory requirements for emission intensity in specific sectors. None of these alone reaches the level of action that the climate science says is necessary, but the combination is the most that the politics has so far supported.

The Border Adjustment Question

If some countries price carbon and others do not, production relocates to non-pricing jurisdictions ("carbon leakage"). The solution is border-adjustment tariffs — tax imports from non-pricing countries at the carbon-content equivalent. The EU CBAM (Carbon Border Adjustment Mechanism) launched in transition in 2023 and goes to full implementation in 2026. The US has not adopted a similar framework but bipartisan interest is unusual for climate policy.

CBAMs change the global political-economy calculus. They give high-pricing jurisdictions political cover (domestic industries are not disadvantaged), and they create incentives for other countries to implement their own pricing to capture the revenue rather than ceding it to importers. Adele Morris of Brookings has been one of the leading academic voices for US adoption.

The Distributional Concerns

Carbon pricing is regressive in most implementations — lower-income households spend a larger share of income on energy. The Swedish and Canadian carbon-tax-and-dividend models recycle the revenue back as flat per-capita payments, which makes the net effect progressive. The political durability of carbon-pricing reforms depends substantially on whether the revenue is recycled in ways that protect lower-income households from the cost increases.

The energy-transition employment effects are concentrated in specific geographies — Appalachian coal counties, the Permian Basin, certain rust-belt manufacturing zones — that bear local costs of policy benefits that are diffuse. The political-economy lesson from the China-shock literature applies here: transitions that ignore concentrated local damage produce backlash that constrains future climate policy. The transition policy that works probably has to include substantial place-based investment in affected regions.

The Honest Reading

Climate change is the externality the Pigouvian framework was made for, and the conventional framework has failed for forty years to produce the level of policy response the science requires. The incremental progress that has been made — sub-national markets, national taxes, the IRA subsidies, the EU Green Deal, CBAM — is real but inadequate. The political constraints are not going away. The question for the next decade is whether the combination of subsidy-driven cost reductions in clean technology, regulatory requirements for emission intensity, and slowly expanding carbon- pricing coverage can compound to the level of action required. The optimistic case is that clean technology becomes cheap enough fast enough that the political economy reverses. The pessimistic case is that progress remains incremental while damages accelerate. Both are plausible; the next decade of policy and technology development will determine which prevails.