What’s the value of disaster protection?

Why do governments tend to undervalue projects that protect us?

Imagine a leader deciding whether to invest in a new railway or a stockpile of vaccines. Each project is expected to deliver the same overall benefit over time. But the railway’s benefits depend on the economy continuing to function much as it does today. A vaccine stockpile’s value, by contrast, becomes apparent only during a pandemic or some other major disruption.

Even though the two projects perform very differently when society comes under stress, most governments fail to account for this. This is because they evaluate them using a single discount rate, the standard tool for converting future costs and benefits into today’s terms.

The result is a systematic bias. Projects that protect us against rare but severe crises look less attractive than they really are, while projects whose benefits arrive mainly in good times look more valuable than they should. This can lead to chronic underinvestment in preparation for rare catastrophes*.

* In “The welfare cost of ignoring the beta” (2024), Christian Gollier finds that using a single discount rate generates huge losses, equivalent to at least 15% of permanent consumption.

Why does a crisis change the value of our assets?

Think about insurance. Most years, your home insurance seems like a poor investment. You pay the premiums and get nothing back. But if your house burns down, that same policy suddenly becomes enormously valuable. It’s not just that your policy pays out, it’s that every €1 means far more to you after a major loss than it does when everything is going well.

The same logic applies to society. A billion euros of benefits delivered in a crisis is likely to generate far more collective value than a billion euros delivered in prosperous times. By acting like insurance, projects that pay off in bad states of the world carry an extra, often invisible value. If bad outcomes are possible, we should account for how valuable these projects would be if they occurred.

If economists agree on this, why haven’t governments responded?

Economists already have a gold-standard approach for pricing risk. Known as consumption-based asset pricing, it values every project according to how its benefits change across different possible futures. But putting this idea into practice is extraordinarily difficult. It requires analysts to make complex, often debatable judgments about how performance will vary in good times and bad.

A few countries, including France, Norway and the Netherlands, have experimented with risk-adjusted approaches. But these methods have struggled to gain traction because they are difficult to explain, cumbersome to implement and leave too much room for arbitrary assumptions.

The challenge is not finding the right theory but finding a method governments can actually use. A method that is theoretically perfect but impossible to implement is of limited use. A method that people can understand, explain and apply may lead to better decisions.

Is there a simpler way to put a price on resilience?

Our proposal is deliberately simple. Instead of trying to model every possible future, we ask decision-makers to consider just two. The first looks broadly like today, with steady economic growth and no major disruption. The second experiences a severe and lasting catastrophe.

We then ask a straightforward question: what is the project worth in each of these futures? The final valuation combines the two answers, weighted by the likelihood of each scenario.

This is not a perfect representation of reality. The future is obviously more complicated than two possible worlds. But our results suggest this simple shortcut captures most of the value of far more elaborate methods, while remaining transparent and practical. Crucially, our method results in a much better allocation of capital than the standard use of a single discount rate.

What difference does this make in the real world?

Governments and firms make decisions under deep uncertainty all the time. Our goal is not to assign a precise probability to societal collapse. It is to offer a radically simple and practical way to recognize the true value of projects that protect us. 
To illustrate our approach, we looked at one of France’s longest-term decisions: where to store highly radioactive nuclear waste. Surface storage can work well, but only if sites are maintained. The long-term risk is that a future crisis leaves us lacking the resources or expertise to keep these facilities safe.

Deep geological storage, as planned by France’s Cigéo project, can continue to isolate waste if society breaks down. Using plausible assumptions, our simple evaluation method suggests that its insurance value is large enough to justify the investment.

The broader lesson of our analysis reaches far beyond nuclear waste. Every long-term investment contains an implicit bet, not just on whether we will be richer or poorer, but on whether future societies will remain stable and secure.

Our proposal can help to improve real decisions about flood defenses, earthquake-resistant construction, backup electricity systems, and other investments whose benefits become fully visible only in times of disaster. If we account for catastrophic scenarios, resilience acquires real economic value.

KEY TAKEAWAYS

• Blind spot – Conventional cost-benefit analysis can systematically undervalue investments that protect us against rare but catastrophic events.
• Rainy day value – Benefits delivered during a crisis are worth more than in good times. Protective projects carry a hidden insurance value that standard accounting misses.
• Less can be more – A simple two-scenario approach can capture most of the benefit of far more complex risk models, while staying practical enough for real-world use.
• Disaster ready – The same insurance logic reveals hidden value in decisions on pandemic preparedness, nuclear waste, and other long-term investments in critical infrastructure.

FURTHER READING

Seeking to inform policy decisions and the public debate, researchers at TSE Infrastructure & Network Center offer expertise on complex challenges that require long-term planning and investment such as construction, responsible finance, transport, and telecoms. “Stress discounting”, forthcoming in the Journal of Risk and Uncertainty, and other publications by Frédéric Cherbonnier and Christian Gollier are available to read on the TSE website.

Article published in TSE Reflect, October 2026