The case against “EA cause areas”

post by nadavb · 2021-07-17T06:37:52.114Z · EA · GW · 18 comments

Contents

  Even if a cause is not neglected, one might still have a comparative advantage working on it
  Spreading the attention of EA across more causes may be a better exploration strategy for finding the best opportunities
  Being supportive of most causes can help the growth and influence of the movement
  Helping people become more effective in what they already do might be more impactful than trying to convince them to change cause area
  More independent thinking could be healthy for EA
  Cause divergence is important to preserving the essence of EA
  What should we do about it?
    you have more arguments against the prominence that a small number of cause areas take within EA? Do you have counterarguments? Do you have concrete suggestions or more open questions? I’d really like to hear your thoughts!
  Acknowledgements
None
18 comments

Everyone reasonably familiar with EA knows that AI safety, pandemic preparedness, animal welfare and global poverty are considered EA cause areas, whereas feminism, LGBT rights, wildlife conservation and dental hygiene aren't.

The state of some very specific cause areas being held in high regard by the EA community is the result of long deliberations by many thoughtful people who have reasoned that work in these areas could be highly effective. This collective cause prioritization is often the outcome of weighing and comparing the scale, tractability and neglectedness of different causes. Neglectedness in particular seems to play a crucial role in swaying the attention of many EAs and concluding, for example, that working on pandemic preparedness is likely more promising than working on climate change, due to the different levels of attention that these causes currently receive worldwide. Some cause areas, such as AI safety and global poverty, have gained so much attention within EA (both in absolute terms and, more so, compared to the general public) that the EA movement has become somewhat identified with them.

Prioritizing and comparing cause areas is at the very core of EA. Nevertheless, I would like to argue that while cause prioritization is extremely important and should continue, having the EA movement identified with specific cause areas has negative consequences. I would like to highlight the negative aspects of having such a large fraction of the attention and resources of EA going into such a small number of causes, and present the case for more cause diversification and pluralism within EA.

The identification of EA with a small set of cause areas has many manifestations, but the one I’m mostly worried about is the feeling shared by many in the community that if they work on a cause that is not particularly prioritized by the movement (like feminism) then what they do "is not really EA", even if they use evidence and reason to find the most impactful avenues to tackle the problem they try to solve. I want to stress that I’m not attempting to criticize any specific organizations or individuals, nor the method or philosophy of the movement. Rather, I’d like to address what I see as a general (largely unintentional) feature of our community.

In this post I try to focus on the downsides of our community being too narrowly focused. I recognize there are also benefits to this approach and counterarguments to my claims, and look forward to them being introduced to the discussion.

I will elaborate on the main arguments I see for being more supportive of most causes:

  1. Even if a cause is not neglected, one might still have a comparative advantage working on it.
  2. Spreading the attention of EA across more causes may be a better exploration strategy for finding the best opportunities.
  3. Being supportive of most causes can help the growth and influence of the movement.
  4. Helping people become more effective in what they already do might be more impactful than trying to convince them to change cause area.
  5. More independent thinking could be healthy for EA.
  6. Cause divergence is important to preserving the essence of EA.

Even if a cause is not neglected, one might still have a comparative advantage working on it

Many cause areas are pushed off EA’s radar due to the criterion of neglectedness. For example, there is no doubt that cancer research has enormous potential in terms of the scale of the problem, and it also seems quite promising with respect to its tractability. But because “curing cancer” is such a mainstream cause, with tons of talent and funding [EA · GW] already pouring into it, EAs tend to not see the point of dedicating yet another individual’s career to solving the problem. It is commonly believed that an EA-aligned individual with a background in biology would do more good working on clean meat, pandemic preparedness or consciousness research, all far more neglected than cancer.

However, this calculus can be somewhat incomplete, as it doesn’t take into account the personal circumstances [? · GW] of the particular biologist debating her career. What if she’s a very promising cancer researcher (as a result of her existing track record, reputation or professional inclinations) but it’s not entirely clear how she’d do in the space of clean meat? What if she feels an intense inner drive working on cancer (since her mother died of melanoma)? These considerations should factor in when she tries to estimate her expected career-long impact.

Such considerations are acknowledged in EA, and personal fit plays a central role in EA career advice. But, while personal circumstances are explicitly discussed in the context of choosing a career, they are not part of the general considerations raising specific causes to prominence within EA (as personal conditions cannot be factored into general-purpose research). As a result, while the cancer researcher knows in the abstract that her personal fit should be taken into account, she still feels like she’s doing something un-EA-like by pursuing this career.

Furthermore, pandemic preparedness might be very neglected when considering the resources allocated to the cause by humanity as a whole, but if we only consider the resources allocated by the EA movement, then it is very likely that cancer research is actually more neglected than pandemic preparedness within EA. Scarcity of EAs working in an otherwise crowded area could matter if we think that EAs have the capacity to contribute in unique ways. For example, impact evaluation and impact-oriented decision making, which are standard tools in the EA toolkit, could be highly valuable towards most causes I could think of. I suspect that very few LGBT activists think in those terms, and even fewer have the relevant tools. The competence of EAs in doing specific types of work (that involves thinking about impact explicitly) is a generic type of comparative advantage that most of us have. I believe that in many cases this comparative advantage could have a massive effect, big enough to counteract even sizable diminishing returns in otherwise crowded areas.

From the perspective of the entire EA movement, it might be a better strategy to allocate the few individuals who possess the rare “EA mindset” across a diverse set of causes, rather than stick everyone in the same 3-4 cause areas. Work done by EAs (who explicitly think in terms of impact) could have a multiplying effect on the work and resources that are already allocated to causes. Pioneer EAs who choose such “EA-neglected” causes can make a significant difference, just because an EA-like perspective is rare and needed in those areas, even in causes that are well-established outside of EA (like human rights or nature conservation). For example, they could carry out valuable intra-cause prioritization (as opposed to inter-cause prioritization).

Rather than considering how neglected a cause is in general, I think it is often more helpful to ask yourself how much comparative advantage you might have in it compared to everyone else currently working in this area. If very few people work on it (in other words, if it’s neglected) then your comparative advantage is simply your willingness to work on it. In this particular case the criteria of neglectedness and comparative advantage align. But neglectedness is only a special case of comparative advantage, which in general could include many other unique advantages (other criticisms of the neglectedness criterion have been expressed here [EA · GW] and here [EA · GW]).

Generally, I feel there is a lot of focus in EA on cause areas, but perhaps not enough emphasis on specific opportunities [EA · GW] to improve the world. Even if a cause area is truly not very impactful in general, in the sense that most work done there is not very high-impact, it doesn’t necessarily mean that every single path pursued in this area is destined to be low-impact. Moreover, when considering the space of unique opportunities available to a specific individual (as a result of their specific background, or just sheer luck), it’s possible they would have exceptionally good options not available to other EAs (such as a highly influential role). In that case, having encountered a unique opportunity to do good can be considered another special case of comparative advantage. 

Spreading the attention of EA across more causes may be a better exploration strategy for finding the best opportunities

It’s widely acknowledged in EA that we don’t want to miss the very best opportunities to do the most good. To tackle the risk that we might, our general countermeasure is to think long and hard about potential missed opportunities and candidates [? · GW] for cause-X [? · GW], trying to study and map ever greater territories. In other words, we generally take a top-down approach: we first try to make the case for working on cause X, and only then, if sufficient evidence and reason point to it having a high potential, we actually start to work in the area and explore it in greater detail. But is it the most efficient strategy to explore the space of opportunities? I suspect that in many cases you can’t really see very clearly the full utility of acting in some space until you actually try to do it. Moreover, exploring the space of opportunities to improve the world through the lens of cause areas is a rather low-resolution mapping strategy, which may lead us to miss some of the highest (but narrow) peaks on the impact landscape. Operating at the resolution of opportunities rather than cause areas could therefore be useful also at the community level. 

If EAs felt more comfortable to pursue diverse causes and report back to the community about their conclusions and insights working in those spaces, then as a community we might do better in mapping our options. By ignoring most of the cause areas that exist out there we might be under-exploring the space of possible ideas to improve the world. Allowing more ideas to receive more attention by EAs, we might find out that they are more promising than appeared at first sight. Encouraging cause diversification, where more EAs feel comfortable working on causes that they feel personally attracted to, might prove a more effective exploration strategy than collective deliberation. This can be seen as a sort of a hits-based approach: if you operate in an area not recognized as generally impactful, the probability of having an incredibly impactful career is lower, but if you identify a huge opportunity the EA community has missed, you could make an enormous impact.

Being supportive of most causes can help the growth and influence of the movement

When people first encounter EA, they often get the impression that becoming seriously involved with the movement would require them to make radical life changes and give up on what they currently work on and care about. As a result, they may prefer to carry on with their lives without EA. I feel we might be losing a lot of promising members and followers as a result of being identified with such a narrow set of causes (an intuition also supported by some empirical evidence [EA · GW]). I know many talented and capable individuals who could do high-impact work, but feel like they don’t really fit in any of the classic EA causes, due to lack of relevant background or emotional connection. Many people also can’t find career opportunities in those areas (e.g. due to the low number of job openings in EA organizations and their limited geographic distribution). In the end, most people can’t be AI researchers or start their own organization. 

Most EAs I know personally are very open minded and some of the least judgemental people I know. That’s one of the reasons I really enjoy hanging out with EAs. Yet, strangely, it seems that as a collective group we somehow often make each other feel judged. From my experience, a biologist choosing to spend her career doing cancer research would often feel inferior to other EAs choosing a more EA-stereotypic career such as pandemic preparedness or clean meat. When introducing herself in front of other EAs, she may start with an apology like “What I’m working on isn’t really related to EA”.

Scott Alexander described (humorously, but probably with a grain of truth) his experience from EAG 2017:

I had been avoiding the 80,000 Hours people out of embarrassment after their career analyses discovered that being a doctor was low-impact, but by bad luck I ended up sharing a ride home with one of them. I sheepishly introduced myself as a doctor, and he said “Oh, so am I!” I felt relieved until he added that he had stopped practicing medicine after he learned how low-impact it was, and gone to work for 80,000 Hours instead.

What if we tried more actively to let people feel that whatever they want to work on is really fine, and simply tried to support and help them do it better through evidence and reason? I believe this could really boost the growth and influence of the movement, and attract people with more diverse backgrounds and skills (which is certainly a problem in EA [? · GW]). Moreover, maybe after engaging with EA for a while, some would eventually come to terms with the harder-to-digest aspects of EA. Learning how to do things one cares about more effectively could serve as a “gateway drug” to eventually changing cause area after all. By focusing on a very narrow set of causes we make ourselves invisible to most of the world.

Helping people become more effective in what they already do might be more impactful than trying to convince them to change cause area

What is the purpose of cause prioritization? The obvious answer is that by knowing that cause A is more effective than cause B, we could choose A over B. But are we always able to make this choice? What if it’s up to someone else to decide? What if that someone else is not receptive to making big changes?

If we encounter a nonprofit promoting dental hygiene in US schools, chances are we won’t be able to make it pivot into an AI think tank, or even just into operating in developing countries. At the point of encountering us, the nonprofit may already be too constrained by pre-existing commitments (e.g. to its funders), the preferences of its employees and volunteers and their emotional connection to the cause area, and inertia. On the other hand, the nonprofit team might be open to start doing impact evaluation and follow evidence-based decision making.

I’m sure there are far more organizations and individuals who are open to advice on how to be more effective in what they already do than there are folks who are open to changing cause areas. As a result, even if it’s more impactful to engage with someone in the latter group, the overall impact of engaging with the former group might still be dramatically greater overall. Of course, there are more conditions that have to be met for that to be true, and I’m not trying to say it is necessarily so, only raise the possibility.

The kind of cause prioritization comparing the effectiveness of distributing bed nets in malaria-struck countries vs. producing US-school-targeted videos portraying the horrors of tooth decay is only relevant in certain contexts. It used to be very relevant in the early days of EA when the primary goal of the movement was to find proven effective charities to donate to, and the audience was a small group of people highly committed to impartiality and doing the most good. If we are in the business of giving career advice to wider publics, these comparisons are not always as relevant.

The attitude of making things better without attempting to replace them by something else entirely could also be relevant to individual career choices, for example if presented with an opportunity to influence a program already funded by government or philanthropic money. While such programs tend to have a defined scope (meaning they are unlikely to turn into top GiveWell charities), there might still be a great deal of flexibility in how they operate. If led by someone who is truly impact-oriented, the counterfactual impact could be quite substantial.  

More independent thinking could be healthy for EA

I think we have a lot of trust within the EA community, and that’s generally a good thing. If a prominent EA organization or individual puts significant efforts into investigating the pros and cons of operating in certain areas and then makes an honest attempt to thoroughly and transparently present their conclusions, then we tend to take them seriously. However, our mutual trust might have downsides as well. I think that having more critical thinking within EA, putting each other’s work into more scrutiny and doubt, could actually be healthy for our movement.

For example, for a long period of time many EAs used to downplay climate change as a cause area and believed that it wasn’t a very effective cause to work on (among other reasons, for not being extremely neglected outside of EA). Only quite recently this view started to get some pushback [EA · GW]. Mild groupthink could have played a role in this dynamic, as prominent EA figures underappreciated climate change early on, and other EAs just went along without thinking about it too much. Maybe our community is more susceptible to groupthink than we would like to think. I sometimes get the impression that many EAs reiterate what other EAs are saying, just because it’s been said in EA so it’s probably true [EA · GW] (I often catch myself in this state of mind, saying things with more confidence than I probably should, only because I trust my fellow EAs to have gotten it right). Likewise, just as we shouldn’t automatically accept everything said within EA without questions, it would also be a bad idea to overlook ideas and beliefs held by people outside of EA just because they are not presented to us in our preferred style.

Cause prioritization is notoriously difficult, because there are so many crucial considerations and high-order effects to contemplate. For example, there is still an ongoing debate within EA on how we should think about systemic change (which is relevant to many cause areas). I think there’s a non-negligible chance that we got a wide range of causes wrong. And, similarly, we might be mistaken in putting too much emphasis on certain cause areas just because they are popular and trendy in EA. By spreading our collective attention into more areas we minimize the risk of groupthink and getting things horribly wrong.

This reminds me of the “replication crisis” recently coming to the attention of many scientific fields. Psychologists used to be too eager to believe claims made by their peers, even those backed by single studies, until they realized that a shockingly large number of these studies just didn’t replicate. What if we are overlooking the possibility of a similar replication crisis in EA? I think it would be beneficial to dedicate more resources into revisiting long-held EA beliefs.

Cause divergence is important to preserving the essence of EA

Cause neutrality [EA · GW] is a central value in EA. If EA collapsed into 3-4 specific causes, even if circumstances justified that, it would later be difficult for EA to recover and remember that it once was a general, cause-agnostic movement. I can imagine a scenario where EA became so identified with specific causes, that many EAs would feel pressured to move into these areas, while those less enthusiastic about the causes would become frustrated and eventually leave the movement. At the same time, many non-EAs may start identifying themselves with EA just because they work on causes so identified with the movement, even if they are not really impact-oriented. At this point, “being part of EA” may just become synonymous to working on a specific cause, whether or not one really cares about impact.

What should we do about it?

You may worry that I advocate for turning EA into something it is not, a cheery bouncy feel-good everyone-is-welcome society like so many other communities out there, thereby taking the edge off EA and irreversibly converting it into a distasteful “EA lite” for the masses. That’s really not what I want. I think we should continue doing cause prioritization (maybe even more of it [EA · GW]), and we shouldn’t be afraid to say out loud that we think cause A is generally more impactful than cause B. I am, however, worried about the movement becoming identified with a very small set of specific causes. I would like members of EA to feel more legitimacy to pursue mainstream, non-EA-stereotypic causes, and feel comfortable to openly talk about it with the community without feeling like second-class citizens. I would like to see more EAs using evidence, reason and the rest of the powerful EA toolkit to improve humanity’s impact in medicine, science, education, social justice and pretty much every mainstream cause area out there. 

To bring the discussion more down to earth, I list a few concrete suggestions for things we can do as a community to address some of the concerns mentioned in this post, without compromising our movement’s core values and integrity (notice that this is far from being an exhaustive list, and I’d be happy to see more suggestions):

Additionally, I think we should have a discussion on the following questions (again, this is not an exhaustive list):

Do you have more arguments against the prominence that a small number of cause areas take within EA? Do you have counterarguments? Do you have concrete suggestions or more open questions? I’d really like to hear your thoughts!

Acknowledgements

I am grateful to Edo Arad [EA · GW], Sella Nevo [EA · GW] and Shay Ben Moshe [EA · GW] for incredibly valuable discussion and feedback on this post.

18 comments

Comments sorted by top scores.

comment by technicalities · 2021-07-18T10:09:19.994Z · EA(p) · GW(p)

Agree with the spirit - there is too much herding, and I would love for Schubert's distinctions [EA · GW] to be core concepts. However, I think the problem you describe appears in the gap between the core orgs and the community, and might be pretty hard to fix as a result.

What material implies that EA is only about ~4 things?

  • the Funds
  • semi-official intro talks and Fellowship syllabi [EA · GW]
  • the landing page [? · GW] has 3 main causes and mentions 6 more
  • the revealed preferences of what people say they're working on, the distribution of object-level post tags

What emphasises cause divergence and personal fit?

  • 80k have their top 7 of course, but the full list of recommended ones has 23
  • Personal fit is the second thing they raise, after importance
  • New causes, independent thinking, outreach, cause X, and 'question > ideology' is a major theme at every EAG and (by eye) in about a fifth [? · GW] of the top-voted Forum posts.

So maybe limited room for improvements to communication? Since it's already pretty clear. 

Intro material has to mention some examples, and only a couple in any depth. How should we pick examples? Impact has to come first. Could be better to not always use the same 4 examples, but instead pick the top 3 by your own lights and then draw randomly from the top 20.

Also, I've always thought of cause neutrality as conditional - "if you're able to pivot, and if you want to do the most good, what should you do?" and this is emphasised in plenty of places. (i.e. Personal fit and meeting people where they are by default.) But if people are taking it as an unconditional imperative then that needs attention.

comment by Aaron Gertler (aarongertler) · 2021-07-22T00:51:39.248Z · EA(p) · GW(p)

As technicalities [EA(p) · GW(p)] noted, it's easy to see the merits of these arguments in general, but harder to see who should actually do things, and what they should do.

To summarize the below:

  • EA orgs already look at a wide range of causes, and the org with most of the money looks at perhaps the widest range of causes
  • Our community is small and well-connected; new causes can get attention and support pretty easily if someone presents a good argument, and there's a strong historical precedent for this
  • People should be welcoming and curious to people from many different backgrounds, and attempts to do more impactful work should be celebrated for many kinds of work
    • If this isn't the case now, people should be better about this

If you have suggestions for what specific orgs or funders should do, I'm interested in hearing them!

*****

To quote myself [EA(p) · GW(p)]:

My comments often look like: "When you say that 'EA should do X', which people and organizations in EA are you referring to?"

Open Philanthropy does more funding and research than anyone, and they work in a broad range of areas. Maybe the concrete argument here is that they should develop more shallow investigations into medium-depth investigations?

Rethink Priorities probably does the second-most research among EA orgs, and they also look at a lot of different topics.

Founders Pledge is probably top-five among orgs, and... again, lots of variety.

Past those two organizations, most research orgs in EA have pretty specific areas of focus. Animal Charity Evaluators looks at animal charities. GiveWell looks at global health and development interventions with strong RCT support. If you point ACE at a promising new animal charity to fund, or GiveWell at a new paper showing a cool approach to improving health in the developing world, they'd probably be interested! But they're not likely to move into causes outside their focus areas, which seems reasonable.

After all of this, which organizations are left that actually have "too narrow" a focus? 80,000 Hours? The Future of Humanity Institute?

A possible argument here is that some new org should exist to look for totally new causes; on the other hand, Open Philanthropy already does a lot of this, and if they were willing to fund other people to do more of it, I assume they'd rather hire those people — and they have, in fact, been rapidly expanding their research team.

*****

On your example of cancer: Open Philanthropy gave a $6.5 million grant to cancer research in 2017, lists cancer as one of the areas they support on their "Human Health and Wellbeing" page, and notes it as a plausible focus area in a 2014 report. I'm guessing they've looked at other cancer research projects and found them somewhat less promising than their funding bar. 

Aside from Open Phil, I don't know which people or entities in EA are well-positioned to focus on cancer. It seems like someone would have to encourage existing bio-interested people to focus on cancer instead of biosecurity or neglected tropical diseases, which doesn't seem obviously good.

In the case of a cancer researcher looking for funding from an EA organization, there just aren't many people who have the necessary qualifications to judge their work, because EA is a tiny movement with a lot of young people and few experienced biologists. 

The best way for someone who isn't a very wealthy donor to change this would probably be to write a compelling case for cancer research on the Forum; lots of people read this website, including people with money to spend. Same goes for other causes someone thinks are neglected. 

This path has helped organizations like ALLFED and the Happier Lives Institute get more attention for their novel research agendas, and posts with the "less-discussed causes [? · GW]" tag do pretty well here.

As far as I can tell, we're bottlenecked on convincing arguments that other areas and interventions are worth funding, rather than willingness to consider or fund new areas and interventions for which convincing arguments exist.

*****

Fortunately, there's good historical precedent here: EA is roughly 12 years old, and has a track record of integrating new ideas at a rapid pace. Here's my rough timeline (I'd welcome corrections on this):

  • 2007: GiveWell is founded
  • 2009: Giving What We Can is founded, launching the "EA movement" (though the term "effective altruism" didn't exist yet). The initial focus was overwhelmingly on global development.
  • 2011: The Open Philanthropy Project is founded (as GiveWell Labs). Initial shallow investigations include climate change, in-country migration, and asteroid detection (conducted between 2011 and 2013).
  • 2012: Animal Charity Evaluators is founded.
  • 2013: The Singularity Institute for Artificial Intelligence becomes MIRI
  • 2014: The first EA Survey is run. The most popular orgs people mention as donation targets are (in order) AMF, SCI, GiveDirectly, MIRI, GiveWell, CFAR, Deworm the World, Vegan Outreach,  the Humane League, and 80,000 Hours.
    • To be fair, the numbers look pretty similar for the 2019 survey [EA · GW], though they are dwarfed by donations from Open Phil and other large funders.

Depending on where you count the "starting point", it took between 5 and 7 years to get from "effective giving should exist" to something resembling our present distribution of causes.

In the seven years since, we've seen:

  • The launch of multiple climate-focused charity recommenders (I'd argue that the Clean Air Task Force is now as well-established an "EA charity" as most of the charities GiveWell recommends)
  • The rise of wild animal suffering and AI governance/policy as areas of concern (adding a ton of depth and variety to existing cause areas — it hasn't been that long since "AI" meant MIRI's technical research and "animal advocacy" meant lobbying against factory farming when those things came up in EA)
  • The founding of the Good Food Institute (2016) and alternative protein becoming "a thing"
  • The founding of Charity Entrepreneurship and resultant founding of orgs focused on tobacco taxation, lead abatement, fish welfare, family planning, and other "unusual" causes
  • Open Philanthropy going from a few million dollars in annual grants to in the neighborhood of ~$200 million. Alongside "standard cause area" grants, 2021 grants include $7 million for the Centre for Pesticide Suicide Prevention, $1.5 million for Fair and Just Prosecution, and $0.6 million for Abundant Housing Massachusetts (over two years — but given that the org has a staff of one person right now, I imagine that's a good chunk of their total funding)
  • Three of the ten highest-karma Forum posts of all time (1 [EA · GW], 2 [EA · GW], 3 [EA · GW]) discuss cause areas with little existing financial support within EA

I'd hope that all this would also generate a better social environment for people to talk about different types of work — if not, individuals need better habits.

*****

Everyone reasonably familiar with EA knows that AI safety, pandemic preparedness, animal welfare and global poverty are considered EA cause areas, whereas feminism, LGBT rights, wildlife conservation and dental hygiene aren't.

I think that any of these causes could easily get a bunch of interest and support if someone published a single compelling Forum post arguing that putting some amount of funding into an existing organization or intervention would lead to a major increase in welfare.  (Maybe not wildlife conservation, because it seems insanely hard for that to be competitive with farmed animal welfare, but I'm open to having my mind blown.)

Until that post exists (or some other resource written with EA principles in mind), there's not much for a given person in the community to do. Though I do think that individuals should generally try to read more research outside of the EA-sphere, to get a better sense for what's out there.

If someone is reading this and wants to try writing a compelling post about a new area, I'd be psyched to hear about it [EA · GW]!

Or, if you aren't sure what area to focus on, but want to embrace the challenge of opening a new conversation, I've got plenty of suggestions for you (starting here).

*****

However, this calculus can be somewhat incomplete, as it doesn’t take into account the personal circumstances [? · GW] of the particular biologist debating her career. What if she’s a very promising cancer researcher (as a result of her existing track record, reputation or professional inclinations) but it’s not entirely clear how she’d do in the space of clean meat? What if she feels an intense inner drive working on cancer (since her mother died of melanoma)? These considerations should factor in when she tries to estimate her expected career-long impact.

I think that very few people in this community would disagree, at least in the example you've put forth.

*****

From my experience, a biologist choosing to spend her career doing cancer research would often feel inferior to other EAs choosing a more EA-stereotypic career such as pandemic preparedness or clean meat. When introducing herself in front of other EAs, she may start with an apology like “What I’m working on isn’t really related to EA”.

What if we tried more actively to let people feel that whatever they want to work on is really fine, and simply tried to support and help them do it better through evidence and reason?

This is where I agree with you, in that I strongly support "letting people feel that what they want to work on is fine" and "not making people feel apologetic about what they do".

But I'm not sure how many people actually feel this way, or whether the way people respond to them actually generates this kind of feeling. My experience is that when people tell me they work on something unusual, I try to say things like "Cool!" and "What's that like?" and "What do you hope to accomplish with that?" and "Have you thought about writing this up on the Forum?" (I don't always succeed, because small talk is an imperfect art, but that's the mindset.)

I'd strongly advocate for other people in social settings also saying things like this. Maybe the most concrete suggestion from here is for EA groups, and orgs that build resources for them, to encourage this more loudly than they do now? I try to be loud, here and in the EA Newsletter, but I'm one person :-(

*****

I think that the EA community should be a big tent for people who want to do a better job of measuring and increasing their impact, no matter what they work on.

I think that EA research should generally examine a wide range of options in a shallow way, before going deeper on more promising options (Open Phil's approach). But EA researchers should look at whatever seems interesting or promising to them, as long as they understand that getting funded to pursue research will probably require presenting strong evidence of impact/promise to a funder.

I think that EA funding should generally be allocated based on the best analysis we can do on the likely impact of different work. But EA funders should fund whatever seems interesting or promising to them, as long as they understand that they'll probably get less impact if they fund something that few other people in the community think is a good funding target. (Value of learning is real, and props to small funders who make grants with a goal of learning more about some area.)

I think that EA advice should try to work out what the person being advised actually wants — is it "have an impactful career in dental hygiene promotion", or "have an impactful career, full stop"? Is it "save kids from cancer", or "save kids, full stop"? 

And I think we should gently nudge people to consider the "full stop" options, because the "follow your passions wherever they go" argument seems more common in the rest of society than it ought to be. Too many people choose a cause or career based on a few random inputs ("I saw a movie about it", "I got into this lab and not that lab", "I needed to pay off my student loans ASAP") without thinking about a wide range of options first.

But in the end, there's nothing wrong with wanting to do a particular thing, and trying to have the most impact you can with the thing you do. This should be encouraged and celebrated, whether or not someone chooses to donate to it.

Replies from: nadavb
comment by nadavb · 2021-07-22T21:02:36.382Z · EA(p) · GW(p)

Thank you Aaron for taking the time to write this detailed and thoughtful comment to my post!

I'll start with saying that I pretty much agree with everything you say, especially in your final remarks - that we should be really receptive to what people actually want and advise them accordingly, and maybe try to gently nudge them into taking a more open-minded general-impact-oriented approach (but not try to force it on them if they don't want to).

I also totally agree that most EA orgs are doing a fantastic job at exploring diverse causes and ways to improve the world, and that the EA movement is very open-minded to accepting new causes in the presence of good evidence.

To be clear, I don't criticize specific EA orgs. The thing I do criticize is pretty subtle, and refers more to the EA community itself - sometimes to individuals in the community, but mostly to our collective attitude and the atmospheres we create as groups.
 

When I say "I think we need to be more open to diverse causes", it seems that your main answer is "present me with good evidence that a new cause is promising and I'll support it", which is totally fair. I think this is the right attitude for an EA to have, but it doesn't exactly address what I allude to. I don't ask EAs to start contributing to new unproven causes themselves, but rather that they be open to others contributing to them.

I agree with you that most EAs would not confront a cancer researcher and blame her of doing something un-EA-like (and I presume many would even be kind and approach her with curiosity about the motives for her choice). But in the end, I think it is still very likely she would nonetheless feel somewhat judged. Because even if every person she meets at EA Global tries to nudge her only very gently ("Oh, that's interesting! So why did you decide to work on cancer? Have you considered pandemic preparedness? Do you think cancer is more impactful?"), those repeating comments can accumulate into a strong feeling of unease. To be clear, I'm not blaming any of the imaginary people who met the imaginary cancer researcher at the imaginary EAG conference for having done anything wrong, because each one of them tried to be kind and welcoming. It's only their collective action that made her feel off.

I think the EA community should be more welcoming to people who want to operate in areas we don't consider particularly promising, even if they don't present convincing arguments for their decisions.  

comment by GidonKadosh · 2021-07-17T16:57:59.217Z · EA(p) · GW(p)

Thank you for putting this together, I strongly agree with many of these points, especially the point of independent thinking. 

I think the strength of this post's argument varies when taking into account different "services" that the EA movement can provide individuals. For instance, for someone in their mid-career who is interested in EA in order to rethink their career path, there would be much more value in a more divergent EA movement that is focused on the  EA toolkit. 
Yet, that wouldn't be the same for someone who looks for donation advice, for which we'd rather put much more focus on very few cause areas and donation opportunities.

That might also be true for someone in their early career who looks for career advice, but that would depend on how much independent thinking they're willing to do, because I strongly agree that this is missing.  I'll add a quote from a Q&A with Will MacAskill ( Aug 2020) [EA · GW] supporting that:

Habiba: ... "What mistakes do you think most EAs, or people in the effective altruism community, are making?"

Will:  ..."more people should just be of the mind that they want to figure out the basics, not take anything for granted, and not defer to [others’ opinions].
At the moment, I think there's a very small number of people doing that, even though I regard it as the core of what EA is about."

I'm personally quite worried that the EA movement would end up filled with people who are fans of a certain cause area without being neutral about their cause.  EA shouldn't be about cheering for certain cause areas, it should be about prioritizing opportunities to do good, and not communicating this good enough internally and externally could be very dangerous for the movement in the long term and would make us miss a lot of our potential.

I've noticed that there are quite a few downvotes for this post, and not enough criticizing comments.  I'd be happy to hear others' opinions on this subject!

Replies from: GidonKadosh
comment by GidonKadosh · 2021-07-17T17:08:56.033Z · EA(p) · GW(p)

I'm adding another suggestion to the list: I think that instead of removing emphasis from the movement's top causes, we might want to put an equal emphasis on the EA toolkit. 

I believe that if you would ask all highly active EAs "what tools do EAs use in order to prioritize opportunities to do good?" you'd get very different answers, while I would hope that everyone could easily be able to recall a closed set of actionable guiding principles.

comment by G Gordon Worley III (gworley3) · 2021-07-17T15:13:19.744Z · EA(p) · GW(p)

I think the obvious challenge here is how to be more inclusive in the ways you suggest without destroying the thing that makes EA valuable. The trouble as I see it is that you only have 4-5 words to explain an idea [EA · GW] to most people, and I'm not sure you can cram the level of nuance you're advocating for into that for EA.

Replies from: nadavb, edoarad
comment by nadavb · 2021-07-17T16:59:07.124Z · EA(p) · GW(p)

I agree that when you first present EA to someone, there is a clear limitation on how much nuance you can squeeze in. For the sake of being concrete and down to earth, I don't see harm in giving examples from classic EA cause areas (giving the example of distributing bed nets to prevent malaria as a very cost-effective intervention is a great way to get people to start appreciating EA's attitude).

The problem I see is more in later stages of engagement with EA, when people already have a sense of what EA is but still get the impression (often unconsciously) that "if you really want to be part of EA then you need to work on one of the very specific EA cause areas".  

comment by IanDavidMoss · 2021-07-21T20:49:40.226Z · EA(p) · GW(p)

Great piece! FYI, I wrote an essay with a similar focus and some of the same arguments about five years ago called All Causes are EA Causes [EA · GW]. This article adds some helpful arguments, though, in particular the point about the risk of being over-identified with particular cause areas undermining the principle of cause neutrality itself. I continue to be an advocate for applying EA-style critical thinking within cause areas, not just across them!

Replies from: nadavb
comment by nadavb · 2021-07-22T21:32:35.024Z · EA(p) · GW(p)

Thank you for bringing this post to my attention, I really like it! We appear to make similar arguments, but frame them quite differently, so I think our two posts are very complementary.

I really like your framing of domain-specific vs. cause-neutral EA. I think you also do a better job than me in presenting the case for why helping people become more effective in what they already do might be more impactful than trying to convince them to change cause area.

comment by Charles He · 2021-07-19T21:37:02.800Z · EA(p) · GW(p)

Hi,

The identification of EA with a small set of cause areas has many manifestations, but the one I’m mostly worried about is the feeling shared by many in the community that if they work on a cause that is not particularly prioritized by the movement (like feminism) then what they do "is not really EA", even if they use evidence and reason to find the most impactful avenues to tackle the problem they try to solve….However, this calculus can be somewhat incomplete, as it doesn’t take into account the personal circumstances [? · GW] of the particular biologist debating her career.
 

I think I strongly agree with this and I expect most EA do too. 

My interpretation is that EA as a normative, prescriptive guide for life doesn’t seem right. Indeed, if anything, there’s evidence that EA doesn’t really do a good job, or maybe even substantively neglects this while appearing to do so, in a pernicious way [EA · GW]. From a “do no harm” perspective, addressing this is important. This seems like a “communication problem” (which seems historically undervalued in EA and other communities).

From the perspective of the entire EA movement, it might be a better strategy to allocate the few individuals who possess the rare “EA mindset” across a diverse set of causes, rather than stick everyone in the same 3-4 cause areas. Work done by EAs (who explicitly think in terms of impact) could have a multiplying effect on the work and resources that are already allocated to causes. Pioneer EAs who choose such “EA-neglected” causes can make a significant difference, just because an EA-like perspective is rare and needed in those areas, even in causes that are well-established outside of EA (like human rights or nature conservation). For example, they could carry out valuable intra-cause prioritization (as opposed to inter-cause prioritization).

This is a really different thought than your other above and I want to comment more to make sure I understand. 

While agreeing with the essence, I think I differ and I want to get at the crux of the difference:

Overall, I think “using data”, cost effective analysis, measurement and valuation, aren’t far from mainstream in major charities. To get a sense of this, I have spoken (worked with?) to leaders in say, environmental movements and they specifically “talk the talk”, e.g. there’s specific grants for “data science” like infrastructure, for example. However, while nominally trying, many of these charities don’t succeed—the reason is an immense topic beyond the scope of this comment or post. 

But the point is that  it seems hard to make these methodological or leadership changes that motivates you dissemination.

Note that it seems very likely we would agree and trust any EA who reported that any particular movement or cause area would benefit from better methods. 

However, actually effecting change is really difficult. 

To be tangible, imagine trying to get the Extinction Rebellion to use measurement and surveys to regularly interrogate their theory of change. 

For another example, the leadership and cohesion of many movements can be far lower than they appear. Together with the fact that applying reasoning might foreclose large sections of activity or initiatives, this would make implementation impractical.

While rational, data driven and reasoned approaches are valuable, it’s unclear if EA is the path to improving this, and this is a headwind to your point that EAs should disseminate widely. I guess the counterpoint would be that focus is valuable and this supports focus on cause areas closer to the normal sense that you argue against.

Replies from: nadavb
comment by nadavb · 2021-07-20T06:43:49.586Z · EA(p) · GW(p)

Thank you for sharing your thoughts!

About your second point, I totally agree with the spirit of what you say, specifically that:

1. Contrary to what might be implied from my post, EAs are clearly not the only ones who think that impact, measurement and evidence are important, and these concepts are also gaining popularity outside of EA.

2. Even in an area where most current actors lack the motivation or skills to act in an impact-oriented way, there are more conditions that have to be met before I would deem it high-impact to work in this area. In particular, there need to be some indications that the other people acting in this area would be interested or persuaded to change their priorities once evidence is presented to them.

My experience working with non-EA charities is similar to yours: while they also talk about evidence and impact, it seems that in most cases they don't really think about these topics straightly. I've found that in most cases it's not very helpful to have this conversation with them, because, in the end, they are not really open to change their behavior based on evidence (I think it's more a lip service for charities to say they want to do impact evaluation, because it's becoming cool and popular these days). But in some cases (probably a minority of non-EA charities), there is genuine interest to learn how to be more impactful through impact evaluation. In these cases I think that having EAs around might be helpful.

Replies from: Charles He
comment by Charles He · 2021-07-20T18:15:08.821Z · EA(p) · GW(p)

Thanks for the thoughtful reply.

I think we are probably agreed that we should be cautious against prescribing EAs to go to charities or cause areas where the culture doesn't seem welcoming. Especially given the younger age of many EAs, and lower income and career capital produced by some charities, this could be a very difficult experience or even a trap for some people.

I think I have updated based on your comment. It seems that having not just acceptance but also active discussion or awareness of "non-canonical" cause areas seems useful.

I wonder, to what degree is your post or concerns addressed if new cause areas were substantively explored by EAs to add to the "EA roster"? (even if few cause areas were ultimately "added" as a result, e.g. because they aren't feasible).

Replies from: nadavb
comment by nadavb · 2021-07-21T05:20:33.078Z · EA(p) · GW(p)

I totally agree with you that many charities and causes can be a trap for young EAs and put their long-term career in danger. In some cases I think it's also true of classic EA cause areas, if people end up doing work that doesn't really fit their skill set or doesn't develop their career capital. I think this is pretty well acknowledged and discussed in EA circles, so I'm not too worried about it (with the exception, maybe, that I think one of the possible traps is to lock someone with career capital that only fits EA-like work, thereby blocking them from working outside of EA).

As to your question, if new cause areas were substantively explored by EAs, that would mitigate some of my concerns, but not all of them. In particular, besides having community members theoretically exploring diverse causes and writing posts on the forum summarizing their thinking process (which is beneficial), I'd also like to see some EAs actively trying to work in more diverse areas (what I called the bottom-up approach), and I'd like the greater EA community to be supportive of that. 

comment by awhitney · 2021-07-20T02:52:07.322Z · EA(p) · GW(p)

Thank you for the write up!

comment by robertskmiles · 2021-07-23T11:22:52.341Z · EA(p) · GW(p)

A minor point, but I think this overestimates the extent to which a small number of people with an EA mindset can help in crowded cause areas that lack such people. Like, I don't think PETA's problem is that there's nobody there talking about impact and effectiveness. Or rather, that is their problem, but adding a few people to do that wouldn't help much, because they wouldn't be listened to. The existing internal political structures and discourse norms of these spaces aren't going to let these ideas gain traction, so while EAs in these areas might be able to be more individually effective than non-EAs, I think it's mostly going to be limited to projects they can do more or less on their own, without much support from the wider community of people working in the area.

Replies from: nadavb
comment by nadavb · 2021-07-23T14:37:55.837Z · EA(p) · GW(p)

I totally agree. In order for an impact-oriented individual to contribute significantly in an area, there has to be some degree of openness to good ideas in that area, and if it is likely that no one will listen to evidence and reason then I'd tend to advise EAs to stay away from there. I think there are such areas where EAs could contribute and be heard. And I think the more mainstream the EA mindset will be, the more such places will exist. That's one of the reasons why we really should want EA to become more mainstream, and why we shouldn't hide ourselves from the rest of the world by operating in such a narrow set of domains.

comment by Arepo · 2021-07-22T07:47:51.380Z · EA(p) · GW(p)

One downside of this is that without some cause curation, you could end up with zero sum games when charities or causes were in opposition - encouraging more effectiveness from feminists and MRAs, nuclear power advocates vs Greenpeace, gun control charities and the NRA, etc.

That's not to say a greater emphasis away from central cause areas would be bad, just not an uncritical one.