Phantom forests: How low-integrity offsets lead to policy failure

19 Nov 2026

Australia's Safeguard Mechanism (SGM) is failing to reduce emissions and drive decarbonisation in the industrial and mining sectors because of an overreliance on cheap, low-integrity offsets, according to a new report.

The Australian National University (ANU) Centre for Environmental Markets' Co-Directors Professors Andrew Macintosh and Don Butler, with researchers from Oxford University, UNSW, Rotterdam School of Management, Erasmus University, Haizea Analytics and the Centre for Environment and Integrity have assessed the extent to which low-integrity carbon credits pose a risk to the effectiveness of the carbon-pricing mechanism.

The report, entitled Phantom Forests and the Safeguard Mechanism: How Low-integrity Offsets are Obstructing Climate Action, was undertaken with financial support from the Australian Conservation Foundation.

The SGM is a centrepiece of Australia's climate policy. The intensity-based emissions trading scheme applies to facilities emitting over 100,000 tonnes of CO2-equivalent annually (collectively around 29 per cent of national emissions).

"The Australian Government is administering the Australian Carbon Credit Unit (ACCU) scheme in a way that results in a large supply of cheap, low-integrity credits," Professor Macintosh said.

"This is artificially suppressing the carbon price faced by SGM facilities, including large gas and coal entities, which account for 55% of emissions covered by scheme."

The study found the suppressed carbon price was stifling decarbonisation by prompting heavy reliance on cheap, low-integrity ACCUs.

"The result is higher emissions and negligible transformation in the industrial and mining sectors," Macintosh said.

The report includes a new analysis of the performance of human-induced regeneration (HIR) projects, the largest project type in the ACCU market and the world’s largest pure carbon removal project by credit issuances.

Last year, HIR projects accounted for 40 per cent of all units issued in the ACCU market and will soon reach 50 per cent of annual issuances.The report includes a new analysis of the performance of human-induced regeneration (HIR) projects, the largest project type in the ACCU market and the world’s largest pure carbon removal project by credit issuances.

HIR projects were found to be failing to regenerate native forests and, where regeneration was occurring, there was generally a large disparity between observed and expected regeneration based on credit issuances.

The projects cover 38 million hectares and they are claiming to regenerate native forests across an area of at least 7.4 million hectares – an area bigger than the whole of Tasmania. 

On the current trajectory, the claimed area of regeneration is likely to exceed 15 million hectares, more than two-thirds the size of Victoria.

According to the analysis, undertaken with datasets derived from 10m resolution Sentinel-2 imagery and very high-resolution (30-50cm) imagery, most of the projects have failed.

"We analysed the 50 most highly credited HIR projects on a per hectare basis and found there has been no, negligible or negative change in tree cover across the majority of the credited areas," Professor Butler said.

"For years, the Australian Government and Clean Energy Regulator have denied that there are integrity issues with HIR projects, releasing a series of government and government-commissioned reports they claim show the projects are performing as expected. 

"Our analysis proves this is wrong – the majority of the projects have not resulted in significant additional regeneration and are not regenerating new forests."

HIR projects have received over 60 million ACCUs, worth between $1.5 billion and $2 billion.