Some 83% of companies in a Harvard database that announced an emissions-reduction target revised it before the target year, adding to the difficulty of judging what businesses have promised and whether they delivered.
An October working paper by Joseph Aldy, Michael Toffel, Alison Ong and Lavender Yang examines how voluntary commitments interact with government policies. It argues that separating additional corporate action from reductions driven by regulation, subsidies or carbon prices requires more than a comparison of headline emissions figures.
The research introduces the Salata Institute Corporate Climate Targets Database, assembled from public company documents covering 2000 to 2024. Its historical results are not a count of companies changing their targets in 2026.
Revisions do not all move in the same direction
According to Harvard’s September 21 launch announcement, about 15% of revisions increased ambition in the target level, about 42% reduced it and 44% maintained it. The rounded percentages total slightly more than 100%.
The 83% figure concerns companies that had announced at least one emissions-reduction target. The percentages describing ambition concern revisions, so they should not be read as the proportions of all listed companies strengthening or weakening their commitments.
The number of companies in the database with climate targets fell from 1,135 in 2023 to 1,056 in 2024. A changing target can also involve different deadlines, emissions categories or measurement methods, which complicates comparisons across reports.
The database draws on sustainability reports, annual reports, regulatory filings and press releases. It covers firms in the Russell 3000, an index representing most US publicly traded equity by market capitalisation, rather than every business operating in the country.
A lower footprint does not identify its cause
An absolute emissions target concerns the total quantity released. An intensity target measures emissions relative to an activity, such as output or sales. Intensity can fall while total emissions rise if the business expands sufficiently.
The October paper also explains how a company’s reported footprint can fall when it sells an operation or when the electricity supplying its facilities becomes cleaner. Those changes need to be separated from reductions attributable to its voluntary target.
Carbon pricing places a cost on emissions through a tax or an emissions-trading system. Regulations and subsidies can also affect investment. A business operating in several jurisdictions faces different incentives at each location.
Some targets mirror public-policy goals, and some internal carbon prices used by companies to assess investment are benchmarked against existing policy prices, the authors report. A voluntary announcement and a legal requirement can influence the same business decision.
Disclosure still leaves gaps
In an October 8 discussion of the project, the researchers described incomplete corporate emissions records as an obstacle to evaluating results. Some companies disclose emissions, others require estimates, and information for some smaller listed firms remains limited.
The database allows users to compare target histories and download records. It does not, by itself, establish that adopting a target caused a particular reduction. The working paper remains research circulated for discussion.
Our coverage of packaging and carbon footprints examined how the boundaries of an environmental assessment affect its result. Corporate targets require similar attention to what is included, the baseline and whether the reported reduction extends beyond the company’s own accounts.
Harvard plans annual database updates. For investors, customers and procurement teams assessing a claim, the original target, subsequent revisions and comparable emissions records need to remain visible together.
