Yes, indeed, what we call ‘confidence interval’ in our report is better described by the term ‘credible interval’.
We chose to use with the term ‘confidence interval’ because my impression is that this is the more commonly used and understood terminology within EA specifically, but also global health in general—even though it is not technically entirely accurate.
Thanks, David! Nice post, and interesting to see a range of options pointed out by different people.
Some suggestions touch upon topics we’ve done research on at Rethink Priorities. For example, we have a report on charter cities and one on improving weather forecasting for agriculture for anyone who’s interested in more detail. We’re also planning to publish something on improving scientific research capacity in sub-Saharan Africa soon.
Hi Oscar, Thanks for your comment. I’ve actually read your post and thought your points are valid! The reason why it is not mentioned in our report is that we agreed with GiveWell that this aspect of discount rates would be out of scope for this particular report (which does not mean it is not important).
I’m neither an expert on aging, nor a biologist (only a silent consumer of the aging literature and stuff that Prof. David Sinclair says about the topic). Just wanted to say that I’d love to read a post on this!
Thanks a lot for your elaborate and thoughtful comment! A quick reaction to your thoughts:
Unfortunately, the literature we reviewed did not seem to be very clear-cut on the question of when exactly to use prizes vs. grants (or other incentives). Intuitively, I’d agree that a prize makes sense (vs. a grant) when identifying a suitable candidate is difficult. To me, this point is already broadly covered by “when the goal is clear, but the path to achieving it is not”, as when you don’t know how to solve something, you may also not know who could solve it. Could you give an example of how moral hazard can come into play?
Thanks for pointing out more design issues! Our report is definitely not exhaustive with regard to how best to design a prize. I don’t fully remember why we did not include the specific design issues you mention, but it is likely because we didn’t find good (quasi-) experimental literature on them. Case studies might be useful here.
I agree that recognition prizes are likely less useful than inducement prizes when you have a very specific problem to be solved. I think recognition prizes are useful when you generally want to increase research and attention to a specific topic, which can help reveal new problems to be solved that you didn’t even think of in the first place.
I think I share your intuition here. I can definitely imagine that financial incentives might potentially be more important in smaller, unglamorous prizes. We focused on large innovation prizes in our report, so I am not sure what’s the most effective incentive structure for small prizes.
Yes, you’re right, that could definitely be the case. We have not looked into that.
Good examples with auction theory and the deferred acceptance algorithm! I’ve been frustrated for a while that my municipality doesn’t want to try out the deferred acceptance mechanism for school/kindergarten choice :)
Thank you so much for sharing additional literature! I really appreciate the effort.
As far as I can tell from a skimming these articles, they seem to be mostly theoretical or modeling studies.
In our report, we mainly focused on the empirical (and especially the (quasi-)experimental) literature in our report because (1) we wanted to understand how well prizes work in practice, and (2) our impression was that the theoretical literature on prizes seemed somewhat too removed from the way that typical large prizes are implemented in real life.
Do you have any advice for individuals who are interested in starting a charity, but who cannot or do not want to go through the CE incubation program (e.g. because their application was rejected, or because they didn’t find the time to participate in the program, or because they do not fully agree with CE’s approach)?
Nice post! It could have been written by me (my daughter was also born in January 2021 and I also returned to work part-time after 9 months) :)
Had to smile many times while reading your post, as I can definitely relate to many points (e.g. finding it harder to find childcare than a job).
Thank you for this article, very interesting!
I would be careful with one statement here though “The official definition of burnout is “physical or mental collapse caused by overwork or stress”. That kind of implies that a person can’t work when burnt out, but that’s not my experience- ceasing work when you’re burnt out is a privilege.”
I think when reading this, it’s easy to come away with the impression that a burnout is actually not so bad. While it may not be so bad in some cases, it can be very severe in others, so severe that you’re literally physically unable to work—sometimes for years.
There are recent studies showing that a burnout can lead to changes in the anatomy and functioning of the brain, e.g. reduced connectivity between different areas of the brain, an enlarged amygdala, and a thinned prefrontal cortex, which is the part responsible for cognitive functioning, meaning that even if you try really hard, you just cannot function as well as you used to (or at all).
I had a burnout myself several years ago and I felt like I wasn’t able to access some parts of my brain and couldn’t do some things that I can normally do (e.g. write a text, do calculations, or concentrate for more than a minute). It took me more than 6 months to feel that I restored my full cognitive capacity.
I agree this is most likely a lower bound—we tried to emphasize this in the report.
I was not aware of the theory that fungal infections are the primary cause of cancer—many thanks for sharing!