Carbon pricing covered nearly 30% of global greenhouse gas emissions across 87 operating policies in 2026, according to the World Bank. For businesses, that brings emissions into operating budgets and investment decisions, while purchases of carbon credits raise a separate question: how much pollution was actually avoided or removed?
The bank’s 2026 report says direct carbon pricing generated more than $107 billion for public budgets in 2025. Those figures include carbon taxes as well as emissions trading systems. They are not a measure of the voluntary carbon-credit market.
Regulated trading can constrain the emissions of covered businesses through a declining supply of allowances. Project-based credits require evidence that an activity delivered an additional reduction or removal elsewhere. Buying one does not automatically satisfy a company’s legal obligations under an emissions trading system.
In the World Bank’s May announcement, Paschal Donohoe, its managing director and chief knowledge officer, said: “When designed well, they can help to drive efficiency and innovation, while mobilizing resources for development priorities.”
What are carbon markets?
Carbon markets allow participants to trade emissions allowances or credits representing claimed emissions reductions or removals. They put a financial value on emissions, but the instrument being traded determines what a buyer receives.
In an emissions trading system, often called an ETS, a regulator sets a cap on covered emissions. Businesses must surrender allowances for their reported emissions. A company that needs fewer allowances can sell its surplus, while another can buy allowances to meet its obligations.
A carbon credit typically represents one metric ton of carbon dioxide equivalent reduced or removed by a project. Carbon dioxide equivalent expresses the warming effects of different greenhouse gases in a common unit.
Credits can come from activities such as methane capture, forest protection or carbon removal. They are purchased voluntarily, and some are eligible for specified regulatory programs. Eligibility depends on the rules of the program concerned. A carbon tax, by comparison, charges a specified price for emissions without creating a market in allowances.
Allowances affect equipment and production decisions
The European Union’s ETS, launched in 2005, covers electricity and heat generation, industrial manufacturing, aviation and maritime transport within its defined scope. One allowance permits one metric ton of carbon dioxide equivalent.
As the cap falls, fewer allowances are supplied. Trading lets businesses with relatively inexpensive opportunities to cut emissions do more of the work, while companies facing higher costs can purchase allowances. The cap and enforcement constrain the total; trading determines where reductions occur.
For a manufacturer, the decision extends beyond the price of a replacement furnace or boiler. Avoided allowance purchases become part of the expected return on cleaner equipment. Electricity costs, installation downtime, financing and the equipment’s remaining life still affect whether an investment pays.
Free allowances can reduce the immediate financial burden on some industries. Price volatility and expectations about future rules also affect planning. A tighter cap does not guarantee that allowance prices rise every year, since demand changes with production, fuel choices and other economic conditions.
Funding can remain unresolved even after a project receives permission. In a separate example, we reported that Drax’s carbon-capture permit did not settle the funding needed to operate its proposed facility.
Europe’s emissions fell, but the causes need separating
The European Commission’s April 2026 release reported a 1.3% decline in EU ETS emissions in 2025 compared with 2024. It said emissions in covered sectors had roughly halved since the system began. The release used reports submitted by March 31, with aviation and maritime reporting still ongoing.
That long-term decline cannot all be assigned to carbon trading. Renewable generation, fuel prices, efficiency measures and changes in industrial output also affect emissions. The Commission linked part of the 2025 industrial decline to weaker construction and other economic activity.
Research has separately tried to isolate the trading system’s effect. A 2020 study published in PNAS, by Patrick Bayer and Michaël Aklin, estimated that the EU ETS avoided about 1.2 billion metric tons of carbon dioxide between 2008 and 2016 relative to a modeled world without the system. That historical estimate provides evidence of an effect beyond the observed downward trend; it is not a measurement of the system’s impact in 2025.
Costs can also move through supply chains when producers pass allowance costs into prices. How much reaches customers depends on competition and the ability to substitute products or energy sources. Lower emissions and higher costs for particular buyers can occur together.
Auctions create public revenue. The Commission says EU ETS auctions raised more than €43 billion in 2025, taking the total since 2013 above €258 billion. Member states must use relevant revenues, or an equivalent financial amount, for climate and energy purposes, with an exception for permitted compensation of electricity-intensive industries’ indirect carbon costs.
Credit buyers face a different measurement risk
Project-based credits depend on additionality: the credited reduction or removal would not have occurred without the incentive from carbon-credit revenue. A project can produce environmental benefits while generating too many credits if it exaggerates what would have happened without the project.
A 2024 Nature Communications review synthesized rigorous evaluations of 2,346 credited projects, covering almost one billion metric tons of issued credits. The researchers estimated that less than 16% of the credits examined corresponded to actual emissions reductions. Performance differed between project categories, and the estimate applies to the assessed projects, not every credit available today.
A separate 2026 Nature Communications study combined six independent evaluations of 44 first-generation projects intended to reduce deforestation. Its estimates indicated that most reduced forest loss, but their aggregate claims were 10.7 times the avoided deforestation supported by independent estimates. The reported 95% confidence interval for that ratio was 5.4 to 26.5.
The authors identified problems with comparison areas and models of future deforestation. Their analysis assessed earlier projects and relied on estimates of an unobservable alternative: what would have happened without protection. It does not establish the performance of every newer forest-credit methodology.
Procurement needs evidence beyond a price per credit
For corporate buyers, additionality is only part of the assessment. Forest projects also face permanence risk, since stored carbon can be released by fire or later clearing. Leakage occurs when an activity prevented inside a project’s boundaries shifts elsewhere.
The Integrity Council for the Voluntary Carbon Market’s ten Core Carbon Principles address additionality, permanence, conservative measurement, independent verification and protection against double counting, among other requirements. Its approval system assesses both crediting programs and categories of credits.
A buyer consequently needs to examine the project’s methodology, monitoring results, treatment of reversals and the claim it plans to make. A low purchase price says little about the amount of additional emissions reduction delivered.
Compliance budgets and voluntary credit purchases should remain separate in company reporting. Allowances meet obligations within a regulated system. Credits finance claimed reductions or removals elsewhere. If those benefits are overstated, the company’s own emissions remain unchanged and its public claim exceeds the evidence behind its purchase.
