More than a quarter of global carbon emissions are now covered by some form of tax or trading system. However, political concerns about competitiveness and carbon leakage continue to shape policy design, with many governments offering compensation to energy-intensive firms to shield them from indirect carbon costs. The authors of this commentary emphasise that how we compensate matters and propose approaches governments can take to move to more targeted, conditional support that aligns with long-term decarbonisation goals.

As the UK government embarks on a full review of its Energy-Intensive Industries Compensation Scheme, timed ahead of the planned introduction of the UK’s Carbon Border Adjustment Mechanism (CBAM) in 2027, it faces a pivotal question: are current compensation policies doing more to protect industry, or to delay its transformation?

Carbon pricing is expanding across the globe, with more than a quarter of emissions now covered by some form of tax or trading system. But political concerns about competitiveness and carbon leakage — the fear that industries will shift production to countries with weaker climate policies — continue to shape policy design. The result? Many governments, including the UK, offer compensation to energy-intensive firms to shield them from indirect carbon costs passed through electricity prices.

Since 2013, UK manufacturers in eligible sectors have been reimbursed for part of their electricity bills — effectively undoing part of the carbon price signal for these firms. The policy has been seen as a pragmatic way to maintain competitiveness and preserve jobs, particularly in trade-exposed sectors like steel and chemicals. But what do these policies actually achieve in practice? Until now, robust empirical evidence has been limited.

Our new study, published in the Journal of Environmental Economics and Management, fills this evidence gap. We examine the UK’s compensation scheme using detailed administrative data on plant-level production and energy use, and apply quasi-experimental methods that isolate the effects of compensation from other trends. In short, we compare similar firms within the same sector, some of which received compensation and others that didn’t. We ask: how did their behaviour change?

We find that compensation does achieve some of its intended goals. Treated firms experienced significantly smaller output contractions than their peers — on average, output (measured by sales of own goods) was 16–30% higher in compensated plants. This suggests the policy helped reduce the risk of offshoring or premature shutdowns.

However, there is a downside. Compensated firms also used more electricity — about 22% more than uncompensated ones — leading to proportionally higher indirect carbon emissions. This suggests that while the policy helped preserve production, it also weakened incentives to reduce overall electricity use.

Importantly, though, we find no significant effect on energy intensity — that is, electricity use per unit of output. This means theincentives to improve energy efficiency remained intact,even under compensation. That’s good news. It suggests that, unlike blanket tax exemptions or full electricity rebates, this scheme did not erode the marginal incentive for firms to invest in energy-saving technology or processes.

In this sense, the UK’s approach — benchmark-based compensation tied to historic output — strikes a better balance than more distortionary alternatives. But the policy still carries an opportunity cost. Between 2013 and 2019, compensation absorbed a substantial share of carbon pricing revenues — public funds that could have otherwise financed industrial decarbonisation efforts, innovation programmes, or skills development.

The policy challenge is clear: compensation can help buy time for firms and reduce leakage risks but if poorly designed or prolonged indefinitely, it risks locking in fossil-intensive production and slowing the pace of industrial transition.

This is especially relevant now. The UK and EU are committed to phasing out free allocation in emissions trading and bringing in border adjustments to level the playing field. In this new policy environment, domestic compensation schemes must evolve.

We argue for a shift from blanket compensation to targeted, conditional support that aligns with long-term decarbonisation goals:

  • Tie compensation to clean investment and performance: Future support should require demonstrable improvements in energy efficiency or adoption of low-carbon technologies.
  • Reinvest carbon revenues strategically: Use funds to support industrial electrification, green R&D, workforce reskilling, and infrastructure upgrades rather than offsetting fossil-based costs.
  • Embed compensation within a broader transition framework: Define clear eligibility rules, timelines and phase-out pathways to avoid long-term subsidy lock-in.
  • Evaluate policies using robust data and methods: Our study shows the power of firm-level data and quasi-experimental designs to reveal what works and what doesn’t.

Carbon pricing is a powerful tool, but only if it delivers the intended signal. As our research shows, how we compensate matters. If we continue to reimburse carbon costs without asking for anything in return, we risk undermining the very transition we’re trying to accelerate.

A version of this commentary was first published by Business Green on 30 July and has been reproduced here with permission. 

Keep in touch with the Grantham Research Institute at LSE
Sign up to our newsletters and get the latest analysis, research, commentary and details of upcoming events.