How should the UK government be changing its approach to discounting in policy appraisal?

‘Discount rates’ are used in policymaking to allow the comparison of investments occurring across different timeframes. Considering what discount rate to use is particularly important in the context of transformational projects, such as climate-related interventions, where benefits are expected to occur long in the future. Stefania Cerruti examines the UK Government’s current approach to discounting and suggests where to make adjustments and focus attention for further review.

In 2025, the UK Government carried out a review of the Treasury’s ‘Green Book’, the policy appraisal handbook used by public organisations in appraising, evaluating and making decisions about spending proposals to ensure value for money. The review highlighted that the handbook’s approach to discounting required attention and a further review into that aspect was commissioned. We submitted a response on that question and share our recommendations below, after explaining the use of discount rates.

Understanding the use and importance of discount rates

Discounting is a key concept in economic appraisal. It captures the idea that having £1 in the pocket today tends to be preferred to £1 in the future. Discount rates are applied to the estimated costs and benefits of an investment to allow policymakers to compare proposals occurring across different timeframes.

Selecting the right discount rate in government appraisal is crucial for all policies, but it is particularly relevant for transformational projects that are expected to have impacts in the long term, such as climate related investments.

To see why this might be the case, we can consider two separate policies.

Suppose the first policy costs £100 a year for five years in real and undiscounted terms. The policy returns £100 per year for 15 years in benefits to society, in real and undiscounted terms, starting from the second year post-intervention. If we consider three different discount rate scenarios, including the current rate in the Green Book of 3.5%, we find the benefit-to-cost ratios to be all quite similar in this case – see Table 1.

We can then consider a second policy, also costing £100 a year for five years and returning £100 a year for 15 years in benefits to society, in real and undiscounted terms. However, in this instance benefits would start being realised from the fiftieth year post-intervention. In this case, even with the same discount rates as in the first policy, benefit-to-cost ratios look very different across scenarios – see Table 2.

In the absence of any other considerations around the costs and benefits of this spending proposal, if the current Green Book discount rate were employed, the policy would not appear to be good value for money, as the costs outweigh the benefits. In general, when policies are expected to return benefits to society in the long term – as in the case of climate change mitigation and adaptation – the choice of a discount rate plays a much stronger role in the appraisal.

How, then, can the right discount rate be chosen to better reflect the benefits of transformational policies?

Recommendations for the UK government

The four recommendations below offer some practical solutions for a discount rate that better reflects the value of transformational investments. Getting the discount rate right is critical for unlocking the long-term benefits of investments in climate change mitigation and adaptation, and nature restoration, which our society urgently needs. A fuller set of recommendations can be found in our survey response report.

1. Government should continue to prioritise the use of the ‘social time preference rate’ – with a few changes.

Generally, there are two primary ways to discount. The first uses a social time preference rate (STPR). STPR has two components: a time preference element, which measures how we value consumption in the present compared with consumption in the future, and a wealth effect, which captures the idea that future generations will be wealthier than present generations. The second broader category of approaches uses rates that reflect market returns, capturing the opportunity cost of investing in a project instead of keeping the same funds in the market.

Although STPR has limitations, it is better suited than other approaches to government decision-making. STPR allows a government to account for factors that are broader than just market returns, such as trade-offs between present and future generations, levels of inequality aversion, and growth considerations.

Furthermore, for other market-related discount rates to work, we need to assume that actors in the market operate under perfect competition. The literature has shown extensively that, in a world affected by climate change and nature degradation, market failures are widespread. Therefore, failing to take this into account would lead to biased interest rates.

Nevertheless, we recommend two key changes to the way STPR is currently derived in the Green Book:

• In the time preference component, we would endorse setting the rate of pure time preference to 0. Any value greater than 0 favours consumption of present generations over future ones, which is difficult to justify morally.
• A key element of the wealth component is the growth rate of future real per-capita consumption, which is proxied by gross domestic product (GDP) per capita. GDP is a metric that does not adequately account for the health and value of the natural environment or resource availability. Investment decisions affect the natural environment, and we propose this should be accounted for in the discount rate, perhaps by using a different metric. One could be to choose a growth rate that accounts for resource use and the natural environment, such as the Office for National Statistics’ Gross Inclusive Income or Net Inclusive Income. Even if the growth in real GDP per capita were retained as the preferred growth metric, whether the current 2% rate is the appropriate one is questionable. In the latest forecasts from the Office for Budget Responsibility, growth in real GDP per capita is consistently below 2%, and at no point higher than 1.8% in the forecast period.

2. The degree to which environmental costs are systematically included in government appraisal should be investigated.

The Treasury’s previous review of discounting concluded that “…the Green Book should not change the discount rate for environmental impacts” and favoured “…improved valuation for environmental impacts and updating these estimates to reflect latest evidence”.

If the Treasury chooses not to incorporate environmental impacts into the discount rate as part of the current review, it is important that environmental costs and benefits are systematically accounted for in appraisals. This should be the case both for policies with a direct environmental focus and for those without, as it is likely that most government interventions will impact the environment directly or indirectly.

It would be beneficial to investigate the degree to which environmental costs are systematically included in government appraisal or if they may be underestimated. This could be achieved by requesting access to a sample of business cases across different policy domains and budget envelopes.

3. Instead of using alternative discount rates for transformational projects, a single, lower discount rate across the portfolio of government investments should be considered.

Typically, arguments in favour of adjusted discount rates for transformational projects recommend using lower discount rates. This assumes that transformational investments will return significant long-term benefits to society that get discounted heavily if the discount rate is high. Therefore, when compared with other projects, transformational projects are unfairly overlooked.

However, employing a different discount rate would fail to acknowledge two cost-related issues:

• Non-transformational projects with spending happening in the short term will appear cheaper in present terms than transformational projects, which will reduce the incentive to invest in transformational projects even further.
• Non-transformational projects might have long-term costs on the environment and natural resources. However, if the discount rate applied to non-transformational projects is high, the present value of costs will be unjustifiably low.

We recommend employing a single, lower discount rate across the portfolio of government investments. This would ensure that the long-term benefits of transformational projects are not overly discounted, while allowing a fair comparison across projects.

4. Risks need to be considered carefully when appraising transformational projects.

Parameters in the current formulation of the Green Book discount rate do not include project-specific risks, which are instead to be considered in the main cost–benefit analysis. This is typically done by calculating a certainty-equivalent value of risks and incorporating it into the financial contingency of the project. Serious consideration needs to be given to the extent to which appraisal practitioners effectively carry out this stage.

To better assess whether this is currently an issue in government appraisal, it would be sensible to investigate how many Treasury business cases accurately quantify certainty-equivalent values of risks, and include it in contingency calculations.

The author thanks Daisy Jameson, Sini Matikainen, Lilia Akatova, Jack Pepin-Hall and Georgina Kyriacou for their review of this commentary.

Some of the numbers in Table 2 were amended on 31 March.