Effective altruism groups are an integral part of the effective altruism community, and have played a central role in helping the community achieve many of its aims:
They have played a significant role in directing funds to effective causes. Many groups have raised significant amounts of money through fundraising campaigns to effective charities. The annual Giving What We Can Pledge Campaign was first conceived of and orchestrated by a university effective altruism group, receiving national news coverage and generating a record number of commitments to donate to effective charities. Groups have also played a role in giving philanthropic advice to organizations and donors.
They have helped to influence people’s career trajectories and place people in high-impact organizations. Many employees at EA organizations became members of the community via participation in an effective altruism group, and effective altruism groups have helped 80,000 Hours by hosting their workshops and referring people for their coaching.
They have also contributed to the community's intellectual progress, both indirectly by organizing events such as EAGx conferences and directly by conducting independent research.
As these various successes illustrate, there are many possible routes to impact that effective altruism groups can pursue. Because of this, it is valuable to ask which of these methods EA groups should prioritize, given what we know about effective altruism and the current state of the effective altruism community. This writeup aims to give a tentative answer to the question: ‘How can EA groups have the highest impact?’
This writeup will be most valuable for people who already have a good grasp of the basic concepts surrounding effective altruism. Owen Cotton-Barratt’s talk ‘Prospecting for Gold’ provides a brief introduction to some of the most relevant concepts (and is available both as a video and transcript).
The first section introduces three features of effective altruism: (i) the value of opportunities to do good likely follows a heavy-tailed distribution, (ii) discovering these opportunities is difficult, and (iii) we should expect our understanding of the best opportunities to change over time.
On the basis of these features, the second section identifies three things that the effective altruism community needs to be able to do in order to successfully grasp the opportunities out there: It needs to (i) attract great people, (ii) coordinate, and (iii) communicate with nuance.
The third and final section relates these features of effective altruism and the needs of the community to the goals of effective altruism groups and provides further reading. It concludes that the comparative advantage of effective altruism groups is to find and foster the development of people who are highly dedicated to, have a sophisticated understanding of, and have skills relevant to ‘doing the most good’, and to integrate these people into the broader effective altruism community.
The value of opportunities to do good (cause areas, interventions, donations, careers etc.) seems to be best modelled as following a heavy-tailed distribution. This can be expressed by the 80/20 principle, where roughly 80% of the contribution is made by 20% of the things in consideration. If opportunities to do good have a heavy-tailed distribution, this means that the most valuable opportunities will be far more valuable than the average. A classic effective altruism example of a heavy-tailed distribution is the cost-effectiveness of different interventions in terms of averting DALYs (Disability Adjusted Life Years): the most effective interventions are over an order of magnitude more effective than the average intervention. 
The disproportionately large impact of the best opportunities means identifying and taking them is particularly important in effective altruism.
Human civilization has made moral progress as it has aged, but it seems that there is plenty more progress to be made. In Classical Athens, it seemed obvious and normal that only men should have political rights, that fathers should be allowed to discard unwanted infants, and that slaves should serve their masters. Just as previous generations have often been oblivious to important moral issues and considerations, we might suspect that there are major moral issues and considerations that we remain unaware of, the discovery of which could radically improve the opportunities that we consider in the pursuit of doing good. For example, making progress on the question ‘Which beings have moral worth?’ seems of paramount importance to our ability to do good.
Effective altruism also requires grappling with complex empirical questions. We may have a good grasp of the short-term effects of some specific interventions, like distributing anti-malarial bednets, but the indirect long term effects are much more difficult to predict, and are also likely to be much more important in determining the value of interventions, and even whether they are positive or negative overall. For example, it is possible that some typical areas of EA activity could have a negative impact, due to increased CO2 emissions increasing the chances of catastrophic climate change.
There are many crucial considerations, both theoretical and empirical, that dramatically influence our ability to do good, and missing any of these could significantly limit the impact we have. For further discussion of some of these ideas, see Michael Page’s EA Global 2016 talk ‘Embracing the Intellectual Challenge of Effective Altruism’ and Nick Bostrom’s - ‘Crucial Considerations and Wise Philanthropy’.
Over the years since the beginning of the effective altruism community, its understanding of the best methods of doing good has developed, the external world has changed, and the resources available to the community have expanded. Because of this, the community looks very different today than it did even just a few years ago. One example of this is the increased proportion of the community’s resources allocated to the risks of artificial intelligence as a cause area: this is in part because our understanding of the topic has developed, and because significant progress in machine learning has been made. Another is earning to give, which has been de-emphasized as a result of a changing understanding and the EA community having more financial resources at its disposal. Moreover, we should expect EA’s priorities and methods to continue to change in the future, and the EA community will likely look significantly different again in another few years.
One of the main limiting factors of the EA community’s ability to succeed is finding and engaging people who are willing and able to make valuable contributions to solving the world’s biggest problems. There are three attributes that are particularly important here.
The EA community needs people who are particularly dedicated to doing good. As impactful opportunities appear heavy-tailed, we should expect there to be a small number of career opportunities that will be exceptionally valuable. Because of this, people that identify and pursue these opportunities can have a very large impact. Similarly, the impact that someone can have within a particular career may differ significantly depending on their focus and the choices they make, and people can amplify their impact by making choices based on their understanding of how to do good.
The EA community needs people who have a sophisticated understanding of how to do good. Doing good is complicated, and the answers are not obvious. We need people who deeply understand the ideas and methodology contained within effective altruism, and who can expand the frontier of our understanding. But the importance of having a deep understanding is not restricted to those who aim to discover new cause areas and crucial considerations. We also need people who can weigh up different options and select the one with the highest expected impact, who can identify promising opportunities within their area of work, and so on. According to a recent survey by 80,000 Hours, ‘good calibration, wide knowledge, and ability to work out what’s important’ are arguably the most needed skills in the community.
The EA community needs people who are skilled in areas relevant to doing good. As the effective altruism community grows, it has been able to specialize and professionalize. The community now has many more domain-experts with skills relevant to their particular avenue of impact, and as this occurs it becomes increasingly important to have people with experience in particular domains, whether management, policy, machine learning or organizational operations.
Importantly, the EA community doesn’t just need people who are dedicated to, have a sophisticated understanding of, and have skills relevant to doing good, but it needs people who have all of these attributes.
Having able and committed people is important to the success of the community. Intuitively, the more people we have in the effective altruism community, the more people we have with these qualities, and the more capacity we have to find and take the best opportunities for doing good.
However, the community’s capacity isn’t just determined by the number of individuals within it, but also by our ability to coordinate. The most valuable opportunities, those which have an impact across the world, or for many generations to come, will only be identified and grasped if people with shared goals and understanding coordinate in the endeavour. Acting as a community means that we can benefit from the understanding of others, and in cases where we cannot assimilate the knowledge ourselves, we can defer to trusted experts within the community. Acting as a community also means that we can achieve ‘gains through trade’ by making decisions based on our comparative advantages for doing good relative to the rest of the community rather than as individuals.
A community can coordinate best when it has shared goals and a shared understanding. The effective altruism community’s goals are somewhat unusual, and our shared understanding is somewhat complex. This can make coordination within a growing community challenging.
In general, coordination within a community is more difficult with more people, and so there is a trade off between expanding and coordinating. It is difficult to say what the ideal size or target growth rate would be for the effective altruism community, but given the benefits conferred by coordination, it seems a smaller, well-coordinated community is likely to be more successful than a larger, diffuse one.
Doing good is complicated, and because of this the effective altruism community’s understanding of it is complex, nuanced, and difficult to summarize. In conveying any complex set of ideas, some form of simplification will be necessary, but there are reasons to be cautious of oversimplified messaging, which may have a detrimental effect on the community.
Oversimplified messaging can propagate misunderstandings of effective altruism, for example that it is just about saving lives, is focused on earning to give, or is only for utilitarians. This has the potential to put off people who would have been valuable to the community, but disagree with or are not interested in the oversimplified version of effective altruism. It is particularly unfortunate when people are put off because they think effective altruism is itself an over-simplified approach to doing good, as it is important to attract people who appreciate the complexity involved in doing good.
Over-simplified messaging may also attract people who don’t share the community’s goals, or who don’t appreciate the complexity of the endeavour, and this can reduce the community’s ability to coordinate.
There are two broad approaches or aims that an EA group might consider:
Community Building: Having an impact by increasing the ability of individuals to take high impact actions in the future, and increasing their likelihood of doing so.
Direct Work: Having an impact through the outcomes of particular projects, independent of the effect that these have on the future actions of the people involved, for example by holding fundraisers, or undertaking political lobbying.
Given the nature of opportunities for doing good and the needs of the effective altruism community, EA groups have a strong comparative advantage for EA community building.
Doing valuable direct work often requires specialized skills, knowledge, and unique opportunities, which are often difficult to find or develop when not working professionally within an area. On the other hand, EA groups (and student groups in particular) are ideal environments for attracting new individuals to the EA community, helping them learn about the ideas of effective altruism, and connecting them with future opportunities for doing good. In this way, effective altruism groups can contribute to generating people willing and able to take high-impact career opportunities, and help solve one of the community’s major bottlenecks.
We can break down the goal of EA community-building into three activities: finding people, fostering their development as effective altruists, and integrating them into the wider community.
Finding people involves introducing them to the ideas of effective altruism, and in doing this it is vital to present nuanced understandings. This will minimize the risk of spreading misunderstandings of effective altruism, and also ensures that you can find people who share the goal of using reason and evidence to do the most good.
People will provide the most value to the community once they have developed skills and understanding relevant to EA. EA groups contribute to this by creating an environment and culture which fosters learning, and by running related activities such as discussion groups, reading groups and research projects.
The impact of people with skills and understanding relevant to EA is significantly increased if they are integrated within the wider effective altruism community. This ensures that they can reap the benefits of coordination with other effective altruists, they are aware of the community’s best understanding of how to do good (as it develops), and they can find and take opportunities to do good, such as high impact careers related to their skills.
A Model of an EA Group expands on these aims, and gives concrete recommendations for how to implement them.
EA community builders can provide an incredibly valuable service to both the effective altruism community, and to the world. Doing good is complicated, and so is building communities of do-gooders. Because of this, it is especially important that EA community builders have a sophisticated understanding of EA community building, and of effective altruism itself. The following recommended reading aims to provide a starting point for this.
EA Community - Building
80,000 Hours - Promoting Effective Altruism Problem Profile
Ben Todd - The Value of Coordination
Kerry Vaughan - The Fidelity Model of Spreading Ideas
Owen Cotton-Barratt - How Valuable is Movement Growth?
Robert Wiblin - EAG 2015 Movement Development
Nick Beckstead - EA Community Building
Jess Whittlestone - Building an Effective Altruism Community
Stefan Schubert, Ben Garfinkel, and Owen Cotton-Barratt - Considering Considerateness: Why communities of do-gooders should be exceptionally considerate
For an example of independent research conducted by an effective altruism group, see [https://www.lesswrong.com/posts/h99f5AqSJRPaqsT8m/oxford-prioritisation-project-review) ↩︎
80,000 Hours - The world’s biggest problems and why they’re not what first comes to mind and How to compare different global problems in terms of impact (a discussion of power laws in career opportunities) ↩︎
Open Philanthropy Project - Technical and Philosophical Questions That Might Affect Our Grantmaking (for elaboration on moral patienthood and moral weight) ↩︎
80000 Hours - Why You Should Focus More on Talent Gaps Not Funding Gaps ↩︎
80000 Hours - Talent Gaps Survey and Why You Should Focus More on Talent Gaps Not Funding Gaps ↩︎
For an alternative perspective, see Kaj Sotala - An argument for broad and inclusive "mindset-focused EA" ↩︎