Longlist of Causes
CEARCH keeps a running longlist of causes (link) that may merit further research to see if they are highly impactful causes worth supporting. The list, which covers the broad areas of global health & development, longtermism, as well as EA meta, is currently around 400 causes long.
In compiling this longlist, we have used a variety of methods, as detailed in this search methodology (link); core ones include:
- Using Nuno’s excellent list as a starting point.
- Conducting consultations and surveys (e.g. of both EA and non-EA organizations and individuals).
- Performing outcome tracing (i.e. looking at good/bad outcomes and identifying the underlying causes): The Global Burden of Diseases database and the World Database of Happiness are especially useful in this regard.
Our hope is that this list is useful to the community, and not just our own research team.
- Classification of causes is fairly arbitrary, and each organization has their own approach. CEARCH find it useful to think of causes in three distinct levels, from broadest to narrowest:
- (1) High-level cause domain, which are problems defined in the broadest way possible: (a) global well-being, which concerns human welfare in the near-term; (b) animal welfare, which is self-explanatory; (c) longtermism, which concerns human welfare in the long-term; and (d) EA meta, which involves doing good through improving or expanding effective altruism itself.
- (2) Cause areas, which are significantly narrowed down from high-level cause domains, but are still fairly broad themselves. For example, within global well-being, we might have global health, economic & development, political reform etc
- (3) Causes, which are problems defined in a fairly narrow way (e.g. malaria, vitamin A deficiency, childhood vaccination, hypertension, diabetes etc).
- Of course, causes can always be broken down further (e.g. malaria in sub-Saharan Africa, or childhood vaccination for diphtheria), and going through our list, you can also see that causes may overlap (e.g. air pollution in a general sense, vs ambient/outdoor particulate matter pollution, vs indoor air quality, vs specifically indoor household air pollution from soot). The reason for such overlap is partly a lack of time on CEARCH's part to rationalize the whole list; but partly it also reflects our view that it can be valuable to look at problems at different levels of granularity (e.g. at higher levels, a single intervention may be able to solve multiple problems at the same time, such that a broader definition of a cause areas helps find more cost-effective solutions; conversely, at lower levels, you can focus on very targeted interventions that may be very cost-effective but not generally applicable).
- Note that animal welfare causes are not in this longlist, as CEARCH has so far not focused on them, for want of good moral weights to do evaluations with. This should not be taken to imply that animal causes are unimportant, or that research into cost-effective animal causes is not valuable.
Cause Exploration Contest
Open Philanthropy had its excellent Cause Exploration Prize; here, we’ll like to do something similar but make the bar significantly lower.
- We invite people to suggest potential cause areas, providing a short justification if you feel it useful (e.g. briefly covering why the issue is important/tractable/neglected), or not, if otherwise (e.g. the idea simply appears novel or interesting to you). All ideas are welcome, and even causes which do not appear intuitively impactful can be fairly cost-effective upon deeper research.
- People are also welcome to suggest potential search methodologies for finding causes (e.g. consulting weird philosophy, or looking up death certificates).
Prizes will be awarded in the following way:
- USD 300 for what the CEARCH team judges to be the most plausibly cost-effective and/or novel cause idea (and that is not already on our public longlist of causes).
- USD 700 for what the CEARCH team judges to be the most useful and/or novel search methodology idea (and that is not already listed in our public search methodology document).
Entries may be made here. The contest will run for a month, until 31st July 2023 (23:59, GMT-12). Multiple entries are allowed (n.b. do make separate individual submissions). The detailed rules, for those who are interested, are available here.