> For the complete documentation index, see [llms.txt](https://research.ambitiousimpact.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://research.ambitiousimpact.com/explainers-and-house-views/discount-rates.md). # Discount Rates Discount rates used across our calculations A discount rate is “used to put a present value on costs and benefits that will occur at a later date” [(Grantham Resarch Institute, 2018, paragraph 1)](https://www.lse.ac.uk/granthaminstitute/explainers/what-are-social-discount-rates/). AIM has historically deferred to GiveWell on their discount rates. Rethink Priorities [(Rethink Priorities, 2023)](https://rethinkpriorities.org/wp-content/uploads/2023/11/Discountrate.pdf) and SoGive have published reviews on their discount rate [(SoGive, 2022)](https://paperpile.com/c/nGACB5/hyGR). Our current discount rates are: * Monetary discount rate (costs and benefits): 6.4% (before April 2026: 4.3%) * Health discount rate (for humans and animals): 3.5%(before April 2026: 1.8%) Following Rethink Priorities’ suggestions for GiveWell [(Rethink Priorities, 2023)](https://rethinkpriorities.org/wp-content/uploads/2023/11/Discountrate.pdf), we calculate our monetary discount rate following this formula:

r= δ + (η − 1)g

Where δ is the sum of several factors (which includes all aspects apart from the wealth effect, see below) and (η − 1)g is the wealth effect. δ=temporal uncertainty + pure time preference + compounding non monetary benefits Our health and animal welfare discount rate is lower than the one for monetary costs and benefits. This is due to the fact that we only use temporal uncertainty for those benefits. ## Wealth effect (aka improving circumstances) Rethink Priorities suggests GiveWell tweak their way of calculating the wealth effect and further increase their consumption growth rate from 3% to 3.3%. Following this, we land on 2% as the wealth effect.

Wealth effect=(η − 1)g

Where η is an isoelastic utility function with η = 1.59. And g is the consumption growth rate. Wealth effect=(1.59-1) 3.3 Wealth effect=\~2 We will revise projections in 2028. ## Other factors δ=temporal uncertainty + pure time preference + compounding non monetary benefits δ = 3.5 + 0+0.9 δ= 4.4 ### Pure time preference We use 0% for pure time preference. See discussion of this in Rethink Priorities. We do not intend to review this in the near term. ### Temporal uncertainty GiveWell uses 1.4% for temporal uncertainty. SoGive [(SoGive, 2022)](https://paperpile.com/c/nGACB5/hyGR) suggested it should be 2.3%, breaking down components as shown in Table 1 at the bottom of the document. In 2026, we decided to use the temporal uncertainty component to also factor in uncertainty about how the future will affect growth and prosperity. Noting increased “noise” about how AI will affect the future, we increased the discount rate we use from 1.4% to 3.5%, somewhat arbitrarily. We decided to use 3.5% for temporal uncertainty. We will review this, in more detail, in 2028. ### Compounding non-monetary benefits We defer to GiveWell and Rethink Priorities on the rationale for including compounding non-monetary benefits at 0.9%, which we agree with. ## References and Annex * * * * * Table 1: Breakdown of temporal uncertainty (source: [(SoGive and Sanjay 2022)](https://paperpile.com/c/nGACB5/hyGR)) | Risk | Contribution to discount rate | Comments/sources | | ----------------------------------------------- | ----------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Nuclear weapons usage | 0.5% | 1.1% chance of a nuclear exchange per year (source: [Rethink Priorities](https://forum.effectivealtruism.org/posts/PAYa6on5gJKwAywrF/how-likely-is-a-nuclear-exchange-between-the-us-and-russia)) ✖ relevance probability (i.e. probability that such an exchange would affect low income countries materially enough to undo the intervention being modelled). A relevance probability of 50% was chosen here, somewhat arbitrarily. | | Other conflict/ severe geopolitical instability | 0.5% | This was estimated very quickly. Given that there are currently more than [35 armed conflicts in Africa](https://geneva-academy.ch/galleries/today-s-armed-conflicts), it seems reasonable to assume the per annum risk of conflict or other severe geopolitical event is higher than 0.5%; this was then somewhat arbitrarily adjusted down to ensure that it focuses on events severe enough to interrupt/undo the intervention being modelled. | | Pandemics | 1% |

This study from Duke University found the probability of a pandemic with similar impact to COVID-19 is about 2% in any year. I did not have time to review this study carefully.

The relevance probability was set as 50%, fairly arbitrarily.

| | AI risk | 0.1% |

This was derived by taking the estimate of 10% chance of an AI-related extinction event in the century (source: The Precipice, Toby Ord) and converting it to an annual figure.

  • Arguably this is too low, since it only covers extinction events, and an AI-related catastrophe which doesn’t lead to extinction could still be relevant.
  • Conversely, in some ways it’s too high, since it spreads the risk evenly across the coming century, however the probability of transformational AI in the next (say) 5 years is much lower than in (say) 20 years.
| | Other | 0.2% | Figure chosen slightly arbitrarily | | Total | 2.3% | Because of time constraints, several components of this estimate were done quickly, and would likely be improved if assessed with more time available. | --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://research.ambitiousimpact.com/explainers-and-house-views/discount-rates.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.