How should we weigh up the costs and benefits of public policies? Most economists believe that discount rates should reflect a project’s contribution to the aggregate risk faced by society. Yet most countries stubbornly refuse to adopt risk-sensitive evaluation of their investments. In a new working paper supported by the SCOR Risk Markets and Value Creation Chair, TSE director Christian Gollier calculates the eye-watering inefficiencies of this short-sighted practice.
What is the social discount rate and why is it important?
The social discount rate is a measure of the relative importance of consequences that occur at different points in time. By putting a present value on the future costs and benefits of projects such as schools, hospitals or solar power subsidies, it allows policymakers to allocate limited resources according to the merits of each initiative. Although this is a widely accepted tool for evaluating investment projects and public policies, there is still much controversy about which discount rate should be used in practice, particularly for the distant future.
Economics cannot provide a complete answer to this question, which involves deep ethical issues. However, the emergence of major challenges to the sustainability of our societies – from climate change and biodiversity to pension-fund liabilities and the reduction of public debt – has put pressure on economists to guide and inform humanity in its choices about how to value long-run costs and benefits. These choices can have huge implications. The present value of $1 million received in 200 years is equal to either $137,000 dollars or $1 depending upon the use of a discount rate of 1% or of 7%, as suggested by different experts.
Why do we need to adjust discount rates for specific projects?
Because Society is risk-averse, any sensible investment valuation system needs to account for a project’s risk profile. Investing in a new hospital, for example, offers insurance value as the hospital will be most useful in the next health crisis, when the economy will suffer. Like other investments which reduce risk – such as earthquake-resistant construction norms, strategic petroleum reserves, or supplies for the next pandemic – the hospital project should therefore have a discount rate that is lower than the interest rate.
In contrast, building a railway line exacerbates macroeconomic volatility because its greatest benefits will be felt during the next boom, and its smallest benefits in a recession. The discount rate for evaluating pro-cyclical investments such as energy or transport projects, especially those that depend on future tax revenues, should therefore include a positive risk premium. The large market risk premium observed over the last century adds support to the idea that risk adjustment should play a crucial role in investment evaluation.
In a 2023 survey, my coauthors and I find that more than 75% of professional economists agree that discount rates should be adjusted to a project’s consumption beta, which measures its contribution to the aggregate risk. The Consumption-based Capital Asset Pricing Model (CCAPM) provides a theoretical framework for doing so.
Why do most countries use a single discount rate?
France is currently the only country in which public investments must be evaluated using a discount rate that is sensitive to the project’s risk profile. While Norway and the Netherlands have experimented with a simplified CCAPM, all other countries that have published discounting guidelines use a single discount rate that fails to account for risk, aversion to risk, and hedging possibilities. As I noted in my 2023 paper with Frédéric Cherbonnier, this means that no insurance value is recognized for policies that hedge the macroeconomic risk. Symmetrically, no penalty is imposed to policies involving benefits that mostly materialize in good times.
The dogma of a single discount rate for the public sector has long been supported using the Arrow-Lind theorem (1970), which argues that “the government invests in a greater number of diverse projects and is able to pool risks to a much greater extent than private investors”, thereby washing out risk completely. This claim has often been used to argue that all public investment projects should be discounted at the risk-free interest rate. But, as widely recognized by economists, it is valid only for projects with a zero CCAPM beta. Because a vast majority of projects have a positive beta, using a risk-free rate implies that projects with a positive net present value (NPV) will exceed the capacity of public funding, forcing governments to impose capital rationing on top of the valuation process.
One of the most puzzling features of the debate on the public discount rate is its reliance on the Ramsey rule (1928). Adjusted for the uncertainty affecting economic growth, this rule provides the right basis to estimate the rate at which risk-free benefits and costs should be discounted. However, using it to recommend an all-purpose discount rate is highly dangerous and undermines constructive debate about how to value the future. The continuing stalemate over the social cost of carbon is a vivid illustration of our collective inability to transform consensual asset pricing theory into practical evaluation rules. From climate disasters to pandemics, this has led to a catastrophic undervaluation of policies that protect society.
What does your paper reveal about the social cost of this failure?
My modelling suggests that the economic consequences of the implied misallocation of capital are severe. I first measure the welfare loss incurred by an isolated agent who uses a single discount rate to value assets and determine portfolio choices. If this agent uses the average cost of capital prevailing in the rational equilibrium as the unique discount rate to value all projects, the welfare loss is equivalent to an immediate reduction of this agent’s wealth by 27%. If all agents use the same inefficient discounting system with a single discount rate equaling their average cost of capital, the welfare loss is around 15% of global wealth.
If all agents use the equilibrium interest rate as their single discount rate, capital rationing is required. With a scheme that only implements 60% of projects with a positive NPV, the welfare loss is then equivalent to a 45% drop in initial global wealth compared to the optimal discounting system. This is a reminder of the importance of capital allocation in generating collective prosperity.
Does the private sector use more efficient rules?
Standard textbooks in finance strongly recommend the CCAPM rule to evaluate investment projects and most CFOs claim to use it. However, there is ample evidence that the CCAPM pricing rule is only partially able to explain observed asset prices. The Security Market Line – which links expected returns to betas – is too flat. This generates a problem similar to the one observed in the public sector, with undervalued low-beta projects and overvalued large-beta projects. Another common misunderstanding is that an institution can use its weighted average cost of capital (WACC) as a single discount rate to reliably evaluate its investment opportunities. This “WACC fallacy” is the private-sector version of the misinterpretation of the Arrow-Lind theorem.
KEY TAKEAWAYS
• Failure to adjust discount rates to a project’s risk profile generates huge inefficiencies, with potentially catastrophic consequences for society.
• The welfare loss of using a single discount rate is equivalent to a permanent reduction in consumption of between 15% and 45%. This reminds us of the importance of the allocation of capital for our collective prosperity.
FURTHER READING ‘The welfare cost of ignoring the beta’ and other publications by Christian, including ‘Risk-adjusted social discount rates’, are available to read on his TSE webpage. For research conducted as part of the TSE-SCOR initiative, see the partnership’s dedicated web page.
