Cause area classification

Cause area classification

Possible ways to classify cause areas

Name Notes
Sub-cause digraph The idea here is to take inspiration from existing similar taxonomies of e.g. academic fields (examples: 1, 2, 3, 4, 5). So similar tree things, but instead of academic fields, you have cause areas in the tree. See this Gist or this graph (graph with clickable text boxes) for a prototype. For this graph, A→B means “A has B as a sub-cause” or more specifically “if I am claiming that I work on B, then I can also claim that I am working on A”.
Associated academic fields For each cause area, use the existing academic outlines to figure out which fields are associated with the cause area, and categorize cause areas by these “associated fields”. The hard part is that the fields are already a tree, and you’re trying to fit cause areas into this tree, but the cause areas themselves have a graph-like structure.
Cause area as academic work For each cause area, think about which academic field research for that cause would be conducted in, and categorize based on that. Same difficulty as “Associated academic fields” idea.
Loop through fields Reverse the loop of the previous idea and ask for each field what cause areas could come out of it. So that could be a way to find new cause areas.
Positive vs negative framing Some cause areas have names like “reducing X” or “anti-X” or “X reform”. Is there something “deep” or fundamental about this, or can all causes be framed in positive and negative terms? Like “reducing suffering” might become “increasing lack-of-suffering” and “content creation” might become “lack-of-content reform”. If there is something fundamental about this positive vs negative framing, then dividing cause areas based on this might be interesting.
Values-based classification Some cause areas seem to be bound to certain terminal values. Like “animal welfare” is tied up with the terminal goal of reducing animal suffering. On the other hand, a cause area like “advancing liberty” might value freedom as a terminal value, but the idea is also that freedom allows you to get to your other terminal values. Something like “content creation” might be a terminal value for someone who values knowledge itself, but for many people it’s just an instrumental value.
People clusters The idea is to look at how people tend to cluster into groups in reality, and to use that as a way to group cause areas. So a “descriptive” classification rather than the more “prescriptive” or “ontological” ones mentioned above. This might be hard for really small areas without many people though.
Reinvestable causes In one of his essays, Eliezer Yudkowsky talks about “reinvestability” and that might be an interesting way to split up causes.
Incentive structure Some sort of classification based on incentives of the people involved in solving problems in each cause area.
Form of altruism See this post by Jiwoon Hwang. I’ve made a prototype that crosses this with a beneficiary group classification.
Beneficiary group I’ve made a prototype that crosses this with a “form of altruism” classification (see row above).
Modify some existing classification I have made a list of existing classifications at list of classifications of philanthropy.
Methods used in the cause area Some causes are mainly advocacy, some require original research, some involve building new technology, some are about building a movement, some require policy or lobbying work, and so on.
Look for cause schemas This is similar (or potentially identical) to the previous idea. Some causes like content creation act like cause schemas than like a single cause area, because there can be a “content creation for X” for each cause area X.
Prerequisite knowledge The idea here is that in the case of direct work (so ignoring earning-to-give and other grantmaking activity), when people are trying to find which causes to prioritize, they want to prioritize based on which causes they personally can contribute to. One of the things that would be useful to know of prerequisite knowledge to work in the area.
Causal analysis Some cause areas are problems that have causes for them. And those causes might themselves be cause areas. You can continue this to find “root causes” and also go the other direction to find “downstream causes”. (This description is somewhat confusing because “cause” means both “cause” as in causality and “cause” as in cause area.) One backchains from this problem/“root cause” to specific actions to take. I’ve seen this kind of backchaining in e.g. MIRI’s 2017 fundraiser post.
Social issues Similar to the ideas about academic fields above, but instead of academic fields, use social issues. There are various classifications of social problems, e.g. The Encyclopedia of World Problems & Human Potential has pages for numerous problems (example that has a graph connecting it to other problems). There is also an Encyclopedia of Social Problems as well as Social Issues in America: An Encyclopedia (these encyclopedia seem to only list issues rather than finding some structure for organizing them). Some papers that might be related are “A classification system of social problems: Concepts and influence on gps’ registration of problems” and “The classification and recording of ‘Social problems’ ”.

See also

External links