Hi Luis, thank you for writing this up! I think it’s a well-written forum post that clarifies important distinctions and makes valuable points, and on a topic that AFAIK is being considered and discussed by many people in the community currently.
I agree with most of your points. In fact, I’ve come around on one that we previously disagreed on, namely on defining “local priorities research” more restrictively than how many people are currently using it: I now see more of the value of having “local” in LPR clearly refer to the altruistic scope rather than the resources being prioritised, consistent and contrasting with the “global” in GPR referring to the global altruistic scope.
There are two further comments/suggestions I’d like to make:
Consider using a different name than “contextualization research”
I think adding new terminology that isn’t immediately clear should be avoided whenever possible, and CR isn’t self-explanatory nor does defining this new term look unavoidable to me. I think using something like “resource-specific GPR” or “targeted GPR”—though a mouthful—is much clearer than CR, because (1) the terms themselves are more self-explanatory (including by using the word “priorities”) and (2) it’s immediately clear that they refer to a subset of GPR.
There may be cases in which “pure” LPR is worth doing for local groups
I expect there are cases in which it will be worth doing LPR (according to your new restricted definition) as a local group from the perspective of doing the most good impartially even when it doesn’t overlap with CR/resource-specific GPR. In other words, I expect there to be cases where doing “pure” LPR would be a recommendation coming out of doing resource-specific GPR for a local group.
Take your Latin American example: my guess is it will often be valuable to do the type of research you link to there even if you expect not to find anything close to competitive to GW top charities, and even if there isn’t a large pot of restricted funds that may be influenced by it.
A few reasons for doing LPR in such a case could be:
showing one cares and is knowledgeable about local problems before making the claim one should focus on other geographies or cause areas
substantiating and better being able to communicate the claim that donations go further in those other countries or cause areas
using this as a training opportunity for charity researchers who could later move on to charity evaluation that would qualify as GPR
As we’ve discussed elsewhere there are clear trade-offs and risks for engaging in LPR at a local group level (e.g. opportunity costs, risk of motivated reasoning or value drift) - so I share your caution in recommending this as an activity. However, I don’t think a blanket recommendation against doing (well-considered and careful) LPR at a local group level is justified either, for the above reasons.
Thanks again for your comments, Sjir! Both of your points are great, and the second one which has led me to think LPR is more important than I thought before.
I still stand by the approach to doing LPR that I propose in the post. Given that there are cases where LPR is highly likely to be effective, I believe that starting with these cases, learning from them, and subsequently determining the best strategy for other situations is a great compromise between the risks and benefits involved. That said, I do think that LPR has the potential to be really successful and get a lot of people involved.
Addressing the specific advantages of LPR you outlined:
On the first two points, my intuition is that local groups could learn enough about that without getting anywhere near the work required for charity recommendations. Alejandro’s analysis is an example of the type of research that I believe moves in this direction, though a more comprehensive exploration is likely warranted.
On the last point, this could indeed be one of the main benefits of LPR. However, prioritization research based on geographic location is not the only way to train people for GPR charity evaluation. Some examples, which I consider to be GPR, include replicating GiveWell’s work, or identifying the best donation opportunities from a non-welfarist perspective (such as those that promote justice).
On the “contextualization research” term, I think I’m a bit more satisfied with it than you are, but I also recognize that it isn’t the ideal name. Suggestions for a better alternative are welcome!
Hi Luis, thank you for writing this up! I think it’s a well-written forum post that clarifies important distinctions and makes valuable points, and on a topic that AFAIK is being considered and discussed by many people in the community currently.
I agree with most of your points. In fact, I’ve come around on one that we previously disagreed on, namely on defining “local priorities research” more restrictively than how many people are currently using it: I now see more of the value of having “local” in LPR clearly refer to the altruistic scope rather than the resources being prioritised, consistent and contrasting with the “global” in GPR referring to the global altruistic scope.
There are two further comments/suggestions I’d like to make:
Consider using a different name than “contextualization research”
I think adding new terminology that isn’t immediately clear should be avoided whenever possible, and CR isn’t self-explanatory nor does defining this new term look unavoidable to me. I think using something like “resource-specific GPR” or “targeted GPR”—though a mouthful—is much clearer than CR, because (1) the terms themselves are more self-explanatory (including by using the word “priorities”) and (2) it’s immediately clear that they refer to a subset of GPR.
There may be cases in which “pure” LPR is worth doing for local groups
I expect there are cases in which it will be worth doing LPR (according to your new restricted definition) as a local group from the perspective of doing the most good impartially even when it doesn’t overlap with CR/resource-specific GPR. In other words, I expect there to be cases where doing “pure” LPR would be a recommendation coming out of doing resource-specific GPR for a local group.
Take your Latin American example: my guess is it will often be valuable to do the type of research you link to there even if you expect not to find anything close to competitive to GW top charities, and even if there isn’t a large pot of restricted funds that may be influenced by it.
A few reasons for doing LPR in such a case could be:
showing one cares and is knowledgeable about local problems before making the claim one should focus on other geographies or cause areas
substantiating and better being able to communicate the claim that donations go further in those other countries or cause areas
using this as a training opportunity for charity researchers who could later move on to charity evaluation that would qualify as GPR
As we’ve discussed elsewhere there are clear trade-offs and risks for engaging in LPR at a local group level (e.g. opportunity costs, risk of motivated reasoning or value drift) - so I share your caution in recommending this as an activity. However, I don’t think a blanket recommendation against doing (well-considered and careful) LPR at a local group level is justified either, for the above reasons.
Names: How about context-specific GPR? (or some variation on this). It takes into account resources / location etc.
I like that one!
I like this too!
I also like it!
Thanks again for your comments, Sjir! Both of your points are great, and the second one which has led me to think LPR is more important than I thought before.
I still stand by the approach to doing LPR that I propose in the post. Given that there are cases where LPR is highly likely to be effective, I believe that starting with these cases, learning from them, and subsequently determining the best strategy for other situations is a great compromise between the risks and benefits involved. That said, I do think that LPR has the potential to be really successful and get a lot of people involved.
Addressing the specific advantages of LPR you outlined:
On the first two points, my intuition is that local groups could learn enough about that without getting anywhere near the work required for charity recommendations. Alejandro’s analysis is an example of the type of research that I believe moves in this direction, though a more comprehensive exploration is likely warranted.
On the last point, this could indeed be one of the main benefits of LPR. However, prioritization research based on geographic location is not the only way to train people for GPR charity evaluation. Some examples, which I consider to be GPR, include replicating GiveWell’s work, or identifying the best donation opportunities from a non-welfarist perspective (such as those that promote justice).
On the “contextualization research” term, I think I’m a bit more satisfied with it than you are, but I also recognize that it isn’t the ideal name. Suggestions for a better alternative are welcome!