I've had a couple of discussions about discount rates in energy systems models and I find it quite a fascinating topic. It's got a bit of a Cat's Cradle - Kurt Vonnegut vibe about it. Anyway, I decided to write the short article below on the subject, which proved surprisingly difficult. Apparently I didn't fully understood the subtleties of why we discount the future, since this actually depends on the context (e.g. are you a government or an investor, what's the project, what's the time frame, etc). Subtlety is not my strong point! I still don't think I fully understand it if I'm honest, but I'm a smidge less ignorant than I was before.

Enjoy!

The ethical implications of discount rates in energy transition models

Money is great. Not just when you have it, but also because it allows us to compare and trade almost anything. You can use the cash in your pocket to buy yourself an apple, a theatre ticket or a trip to Mars because we agree on these things having a monetary value. In the field that I work in one of the fundamental questions we ask is “How can we best decarbonise our energy systems?”, and our definition of “best” usually boils down to what cost less. We do this precisely because we can compare different technologies such as gas power plants and wind turbines on the basis of cost. We also have to compare these across time scales: what is the cost of installing a wind turbine now against doing it in 50 years time? To do this, we need to use discount rates.

For a variety of reasons, money that you have now is not as valuable to you as the same amount of money in the future. 100 Euros is not worth much to you a century from now, since you'll be dead, and it might not be worth as much to you in 20 years time when you'll probably be richer (I hope that I will be at least). The former is an example of the “time preference of money”, while the latter is true in general since supposedly we're all getting richer as time moves inexorably forward. This can be conveniently described using discount rates. Concretely, a discount rate of 1% per year means that 100 USD in 100 years time is worth 371 USD to you now [1].

Why you’re discounting determines your discount rate. If you’re a private individual deciding whether to invest in a project or not you may use the interest rate of your bank as your discount rate, since this will allow you to compare your investment with the alternative of leaving your money in the bank. If you’re a government deciding what would be the best path towards a carbon neutral energy system, you will use what’s charmingly called a social discount rate, which allows us to weigh up the benefits future generations would receive against the costs we would incur now.

This cost-benefit analysis can be done with integrated assessment models, and these are very sensitive to your choice of discount rate as a paper by Emmerling et al. elegantly demonstrates. A low discount rate will mean it’s cheaper to reach net zero carbon emissions earlier than a higher one. A low discount rate will also also lead to less reliance on negative emissions technologies, e.g. planting trees, which remove carbon from the atmosphere. What’s happening is that with a high discount rate you will put off any climate action since it will be cheaper to do so in the future. On top of that, the cost of environmental damage will be reduced simply because it’s happening to future generations who are worth less to you.

This sensitivity to the discount rate cuts to the heart of the rationale behind climate action. Climate change is only a problem if you care about future generations - if you don’t, well who cares, we’ll probably be dead before the worst happens, and we might as well enjoy ourselves in the meantime. Environmentalism can be seen then as an extension of conservatism, in that it upholds the conservation of the natural world of today for the society of tomorrow.

Engineers and economists can find it quite embarrassing that one number so wildly influences the best course of action in the face of climate change, and so they hide this dilemma by reference standard textbook recommendations. This is a shame, since it’s not often that we can translate an ethical and political decision so elegantly into mathematical form, and we should embrace this complication rather than avoid it.

[1]: Obtained by raising 0.99 to the power of 100 (years) and multiplying it by 100 USD.