A Fermi estimate (or back-of-the-envelope calculation, often abbreviated BOTEC) is a rough calculation that aims to be right within about an order of magnitude, prioritizing getting an answer good enough to be useful without putting the large amounts of thought and research needed to attain greater accuracy. Fermi estimates typically approximate an answer by making various simplifying assumptions and decomposing the problem into smaller tractable units. Open Philanthropy and other organizations in the effective altruism community routinely use BOTECs for impact assessment and cause prioritization.
Further reading
Berger, Alexander (2014) Alexander’s back of the envelope “importance” calculations, GiveWell.
Muehlhauser, Luke (2013) Fermi estimates, LessWrong, April 11.
von Baeyer, Hans Christian (1988) How Fermi would have fixed it, The Sciences, vol. 28, pp. 2-4.
Related entries
cause prioritization | expected value | forecasting | impact assessment | model uncertainty | prediction markets | value of information
If you want to try your hand at making some Fermi estimates, you can play The Estimation Game. New games are released each month, and you can play the previous months' games in the archive.
This work is licensed under a Creative Commons Attribution 4.0 International License.
I think this video is a good, practical intro to Fermi Estimates! (If you can ignore the channel name...)
I have a slightly less terse blogpost on this topic here, which might be of interest.