Published: August 2017
Grant investigator: Nick Beckstead
This page was reviewed but not written by the grant investigator. CEA staff also reviewed this page prior to publication.
The Open Philanthropy Project recommended a grant of $2.5 million to the Centre for Effective Altruism (CEA) for general support. CEA is a central organization within the effective altruism (EA) community that we believe has a good track record of helping the community grow. This grant will allow CEA to expand several of its key activities.
This grant falls under our work aimed at growing and empowering the effective altruism (EA) community.
About the grant
Case for the grant
We believe that CEA has a good track record of helping the effective altruism community grow, and its leadership appears to be fairly value-aligned with us in terms of this goal. Impacts of CEA that we see as particularly valuable include:
- $1.4 billion worth of pledges made to Giving What We Can (GWWC), a project of CEA, as of April 2017.1 (While we expect that actual donations resulting from these pledges will be only a fraction of that total, this would still be substantial in absolute terms.)
- Introducing effective altruism to people who have become valuable members of the EA community.
We believe CEA has achieved these outcomes at a relatively low cost per unit and that the influence is likely to be substantially positive (see our discussion of CEA’s cost-effectiveness below).
CEA currently has a funding gap, and CEA’s proposed uses of additional funds seem valuable to us. Most of CEA’s other funders are individual donors who we expect are fairly value-aligned with Open Philanthropy (and so we are less concerned than usual about the chance that our funding will be fungible with theirs).
Budget and room for more funding
If we renew our grant for $2.5 million next year (see below), our funding will in total increase CEA’s 2017 budget by $1.25 million and its 2018 budget by $1.875 million, with the remaining $1.875 million partly offsetting reduced fundraising from other donors, and partly increasing CEA’s reserves for 2019.
Our funding will be used primarily to allow CEA to hire new staff; increase staff salaries (from what we see as previously low levels); provide additional support to local EA groups; increase its budget for EA Global and EAGx events (conferences about EA); and partially fund EA Grants2.
We consider CEA’s past activities to have been likely quite cost-effective, though we do not think all of CEA’s impact is easy to quantify. Some of the more quantifiable aspects of CEA’s impact include:
- Influencing nearly $23 million in charitable donations.3
- Securing roughly $1.4 billion in GWWC pledges.4 We roughly estimate these will eventually result in at least $150 million (as a relatively low-end estimate) in counterfactual donations to effective charities (i.e. donations that would not have been made in the absence of CEA).
- Attracting valuable talent to the EA community.
CEA has achieved the above outcomes while spending less than $5 million in total over its history. Nick Beckstead, who leads our grantmaking in this area (“Nick” throughout this page), believes that, so far, a large majority of these funds and this talent has been focused on substantially more impactful projects than they otherwise would be, and that this is likely to be the case in the future. Nick’s opinion is based on a combination of (i) survey data from GWWC pledge-takers on what they otherwise would have spent money on, (ii) conversations with people who have joined the EA community and started working on relevant projects, and (iii) a priors-based assumption that if thoughtful and well-intentioned people change their choices as a consequence of interaction with reasonable arguments – which Nick takes these changes to be – that’s likely to be good in expectation.
We expect additional funding for CEA to be fairly cost-effective as well (though we think it is likely that additional marginal CEA funding will be somewhat less cost-effective than its past spending).
Risks and reservations
- We have had some reservations about CEA’s management in the past; for instance, we think past EA Global events were not as well organized and executed as they might have been, and have sometimes emphasized prestige and popularity rather than optimizing more strongly for intellectual content. This opinion is primarily based on the experience of some Open Philanthropy Project staff participating in and observing EA Global in 2015 and 2016. Nick’s impression - based on conversations with others who attended - is that EA Global Boston 2017 improved along these dimensions.
- We believe CEA has made some missteps in terms of public relations and communications, and has used overly aggressive marketing in the past.
- We have some reservations about the value of the research component of CEA’s work.
Plans for learning and follow-up
Key questions for follow-up
The main questions we plan to follow up on are:
- How many “pledge equivalents”5 has CEA caused (i.e. how many people has CEA helped to become part of the EA community and/or to achieve more)?
- How many new GWWC pledges have there been?
- Has there been a noticeable improvement in how well run EA Global, EAGx, and/or local EA groups are?
- Is there compelling evidence linking CEA’s activities to the above outcomes? Is there evidence that the marginal work funded by this grant played a significant role?
- How has CEA’s fundraising gone?
Other questions we may follow up on include:
- Does CEA’s research and web content seem useful? For instance, do we find it useful to link to in order to explain core EA concepts to our readers, or do we learn important things from reading it?
- Is CEA supporting student groups more effectively?
- Is CEA presenting compelling policy arguments to key decision-makers, and if so, have these resulted (or nearly resulted) in any concrete policy outcomes?
Questions about EA Global that we may follow up on include:
- How many people attended EA Global events?
- Did long-time members of the EA community find EA Global valuable? Did newer EA community members?
- Did EA Global help forge any new and especially valuable connections from outside the current EA community?
- Does CEA have evidence of specific positive outcomes of EAGx events? Did attendees of EAGx events consider them beneficial?
- Did we think CEA’s marketing and communications around EA Global were professional and non-aggressive?
We plan to renew this grant for between $1.25 million and $2.5 million next year depending on the outcomes of the various projects CEA plans to try out this year, and at a level consistent with our funding being less than 50% of CEA’s total budget. Nick plans to make a recommendation on this point around mid-November 2017. For the duration of our support of CEA, we plan to produce a significant update and choose whether and at what level to renew our funding once per year.
Predictions relevant to the grant
Given that we think GWWC pledges account for a significant portion of CEA’s impact, we have tried to predict the likely number of pledges that will be made over the course of our grant. Nick’s probability distribution for the total number of GWWC members on January 1, 2018, is:6
- Less than 3,500 (i.e., growth to 1.4x the current total or less): 7%
- Between 3,500 and 5,000 (1.4x–2x growth): 71%
- Between 5,000 and 7,500 (2x–3x growth): 17%
- More than 7,500 (more than 3x growth): 5%
- By November 2017, CEA provides compelling evidence that each of the following projects results on average in a new GWWC pledge for less than $2,000:7
- Doing Good Better book giveaway: 60%
- EAGx events: 30%8
- Standard marketing and outreach (e.g. online ads): 65%
- Nick considers EA Global 2017 better executed than EA Global 2016: 70%
- At the start of 2018, Nick believes that cause prioritization working groups at universities are a cost-effective way of identifying new talent in the EA community (conditional on there being at least three such working groups, which he thinks is likely): 70%
- At the start of 2018, Open Philanthropy Research Analyst Ajeya Cotra thinks CEA’s support of EA student groups has improved since September 2016 (conditional on Ajeya feeling she has enough evidence to make such a judgment): 75%
|CEA Budget Scenarios||Source|
|Effective Altruism Grants [archive only]||Source|
|Giving What We Can Dashboard, Monthly totals||Source|
|Giving What We Can Homepage [archive only]||Source|
- 1. Giving What We Can Homepage [archive only]
- 2. Effective Altruism Grants [archive only]
- 3. Giving What We Can Homepage [archive only], as of April 2017.
- 4. Giving What We Can Homepage [archive only], as of April 2017.
- 5. CEA is working on a metric that is somewhat similar to the impact-adjusted significant plan change used by 80,000 Hours (an organization that is part of CEA but operates fairly autonomously), but encompasses plan changes and other ways of becoming more involved in effective altruism in addition to donations.
- 6. Nick bases these predictions on past growth in GWWC membership, summarized here:
Year Total pledges on January 1 Year-over-year growth Numerical growth 2013 268 2014 383 ~1.43x 115 2015 ~800 ~2.09x ~417 2016 ~1500 ~1.88x ~700 2017 ~2500 ~1.67x ~1000
Linear growth during 2017 would predict roughly 3,500 total pledges by 1/1/2018; a “same % growth” model would predict roughly 4,166 total pledges by 1/1/2018.
Over the first two months of 2017, there were about 260 new GWWC pledges. (Giving What We Can Dashboard, Monthly totals)
Two factors that we think may potentially boost this year’s increase in pledges over previous years are 1) CEA’s recent participation in Y Combinator, and 2) Tara Mac Aulay joining as COO. However, in making his predictions, Nick put the most weight on the base rate (which itself incorporates various improvements over time).
- 7. CEA describes how it evaluates the success of various projects partially in terms of resulting GWWC pledges here.
- 8. Nick predicts that EAGx events will be better executed in 2017 than in the past, but expects that he will find the evidence as to whether extra funding for these events resulted in more pledges inconclusive.