What Is Carbon Pricing? Making Emissions Cost Something

Two instruments do the same job from opposite directions, and the choice between them is a choice about which variable to fix:

• •A carbon tax sets the price and lets the quantity of emissions fall out of it, which gives businesses a predictable cost to plan around and does not guarantee any particular emissions outcome
• •An emissions trading scheme, or cap and trade, sets the quantity by issuing a fixed number of allowances and lets the market determine the price, which guarantees the environmental outcome and produces a price that can swing wildly
• •Under trading, firms that cut emissions cheaply sell their surplus allowances to firms for whom cutting is expensive, so the reductions happen wherever they cost least, which is the efficiency argument
• •Hybrid designs add a price floor and ceiling to a trading scheme, which limits the volatility at the cost of loosening the cap
• •Allowances can be auctioned, which raises revenue, or given away free, which does not, and free allocation is the standard political concession to affected industries
• •Offsets, which allow an emitter to pay for a reduction elsewhere instead of cutting, are the most criticised component, since verifying that a reduction would not have happened anyway is genuinely difficult and frequently done badly

Around eighty schemes operate worldwide, covering something over a fifth of global emissions. The European Union's trading system, running since 2005, is the largest and longest-established and has been through three major redesigns after an early period of over-allocation left prices near zero; a market stability reserve introduced in 2019 withdraws surplus allowances automatically and prices have since been high enough to influence decisions. China launched a national scheme in 2021 covering its power sector, which is the largest by emissions covered and which regulates emissions intensity rather than absolute volume. Sweden has had a carbon tax since 1991 and has raised it to among the highest in the world while its economy grew and emissions fell. Canada operates a federal price with provincial variation, and California and several north-eastern American states run trading schemes. Coverage is uneven by sector: electricity and heavy industry are usually included, while agriculture, aviation, shipping and land use frequently are not, for reasons of measurement difficulty and political resistance.

Four difficulties recur and each has produced real failures. The price level in most schemes is far below the estimated social cost of carbon, so the signal is too weak to change investment decisions, and the political ceiling on the price is usually well under what the economics implies. Carbon leakage is the concern that production simply moves to a jurisdiction without a price, which produces no environmental gain and loses the industry, and the responses are free allocation to exposed sectors, which weakens the incentive, or a border adjustment charging imports at the domestic rate, which the European Union began phasing in from 2023 and which trading partners regard as protectionism. Distributional effects are regressive if nothing is done, since energy is a larger share of a poor household's spending, and the standard fix is to return the revenue as an equal per-person dividend, which makes most households better off in cash terms while preserving the incentive. And political durability is the hardest, since a visible price on fuel is an easy target, as the French fuel tax protests of 2018 and the Australian repeal of its scheme in 2014 both demonstrated.

The evidence is better than the political debate suggests and more modest than advocates claim. A substantial body of empirical work, including a 2024 meta-analysis covering dozens of schemes, finds that carbon pricing reduces emissions by something in the range of five to twenty-one percent relative to what would have happened, with most estimates clustering around the lower half of that, and finds no detectable negative effect on employment or economic growth in the jurisdictions studied. The British carbon price floor for electricity, introduced in 2013, is the clearest single case: it pushed coal generation from around forty percent of supply to almost nothing within six years, faster than any other policy in the sector. What pricing has not done is deliver the pace of reduction that stated targets require, which is why almost every serious analysis now describes it as one instrument among several rather than as the policy, with regulation, public investment, standards and research funding doing work that a price alone does slowly, particularly where consumers respond weakly to price or where the necessary technology does not yet exist.

Carbon pricing charges emitters for a cost they impose on others, either by setting a price through a tax and letting the quantity follow, or by setting a quantity through tradable allowances and letting the price follow. Around eighty schemes cover over a fifth of global emissions, with the European trading system the largest established one and China's the largest by volume. The recurring problems are prices below the social cost, leakage to unpriced jurisdictions, regressive effects best fixed by returning revenue per person, and political fragility. The measured reductions are real and smaller than targets require.