I’m announcing that the Centre for Effective Altruism’s (CEA) EA Grants project has now been reopened for public applications. If you have a project that you think is worth funding, you can fill out our short application here. Applications close on Sunday, October 14. We expect to run regular rounds of public EA Grant applications like this. We expect the next round after this one to be early next year but we want to review lessons from this round before committing to a date.
CEA’s mission is to build the EA community because we believe that a community of capable, dedicated, and altruistic people can make real progress on some of the world’s most significant problems. As part of this work, we look for ways to help coordinate the distribution of resources to the places in the community where they will have the highest impact. In many cases, people in this community have ideas for promising projects but lack sufficient access to funding to pursue their ideas. The goal of EA Grants is to find and support these kinds of projects.
Below I explain the current status of EA Grants, the application process, our evaluation criteria and what kinds of projects should apply.
Current status of EA Grants
We ran our first public round of EA Grants in the summer of 2017 and distributed around £500,000. In total this consumed ~800 hours of staff time. At its peak, EA Grants required three staff members working on it on a roughly full time basis. After the initial EA Grants round we concluded that we would likely want to run a similar project again in the future, preferably at a lower staff time cost than the previous round. You can read a review of that round here.
We had planned to open a public round of Grants in early 2018, but realized we did not have the staff capacity to process the number of applications we expected. We elected to devote staff time to launching EA Community Building Grants and to launching our individual outreach retreats (such as our Operations Forum) instead of devoting that time to reopening EA Grants for public applications. When we began our hiring round in the summer we then advertised for someone to take on our EA Grants Evaluator role full time. As a stop-gap measure we decided to evaluate EA Grants on a referral basis, and we distributed links to the application form to some members of the EA community whose judgement we trust. This had the benefit of meaning we had to review fewer applications and so could run the project with less staff capacity.
Through the referral process, we’ve agreed to fund 22 grants for a total of ~$850,000, using substantially less staff time than the 2017 EA Grants round. However, the referral system has the significant downside of making it less likely that we encounter projects from people outside of our networks. It also meant that some potentially promising applicants may have failed to develop projects that would have been good candidates for EA Grants funding, because they didn’t know that EA Grants funding was still available.
Our ultimate goal is to keep EA Grants open for anyone to apply on an ongoing basis. We’re currently looking to hire an EA Grants Evaluator to oversee the evaluation process and to give us enough capacity to reliably keep EA Grants open for applications. However, even with a full time EA Grants Evaluator on staff, when we reopen EA Grants to the public, we expect to get an initial surge of applications that might prove too much for us to evaluate in a timely manner.
To deal with this concern and to accelerate the timeline for a public EA Grants application, we’ve decided to open EA Grants to the public, but to restrict the number of applications that move from the first stage of the evaluation process (the short application) to the second stage of the process (the long application) to 100. This should ensure that we don’t end up with more applications than we can evaluate in a timely manner. We plan to notify applicants that we suspect might have advanced without the restriction and ask them to reapply at a later date.
What happens when I apply?
The application process for EA Grants now has five basic stages:
1. Short application
2. Long application
3. Reference check
4. Interview
5. Final decision
Applications are open now and will close on Sunday, October 14. We expect to begin evaluating applications before the application period ends, so it’s advantageous to submit your application sooner rather than later.
Between October 19 and 26, we will notify the top 100 applicants and ask them to complete our long application. As mentioned above, we’re restricting the number of applications that we evaluate further to ensure that we have sufficient capacity to perform the evaluations. This means that we expect to reject some applicants who we think might have been plausible candidates for funding after further evaluation. After the long application we will select the top applicants for interviews, reference checks and, potentially, funding. Given the stages required to assess grant applications, we expect to make final decisions in early December. We’ll keep applicants up to date on the progress of their application by email.
Changes from the previous EA Grants round
The biggest change over the 2017 round is that we broke up the application form into two stages: the short application and the long application. In the 2017 round, we noticed that many of the applications could be rejected after a brief review because they weren’t focused on a plausibly high priority area or because the applicant lacked any signs of alignment with EA values. We also suspect that some plausible applicants failed to apply because the application form was particularly daunting to complete.
The short application is designed to give us enough information to determine if an application is plausible. For applications that aren’t plausible, the short application reduces the time they spend on applying for an EA Grant. For applications that are plausible, the short application allows us to send a positive signal about the application before asking the applicant to spend more time on the application process. For the current application round, the number of applicants that can be moved to the long application will be restricted to 100, but in the future we plan to accept all applications that seem to be above our funding bar.
We’ve also added a request for three references to the long application. We’ll email each reference once the application is complete and ask them to fill out a brief survey about their relationship to the applicant. In particularly close funding decisions we may also call the references.
Evaluation criteria
EA Grants applications are challenging to evaluate. We accept applications in any cause area, and we accept requests for many different types of projects and funding amounts ranging from requests for a few thousand dollars to travel to a conference to requests for a hundred thousand dollars to start a new organization. In addition, because the average amount of funding requested is small, it often doesn’t make sense to spend as much time as we’d like on evaluating each request.
The wide range of possible projects means that we expect to use somewhat different evaluation criteria for different types of projects, and that we expect to evaluate larger funding requests much more thoroughly than smaller funding requests. However, we expect the basic evaluation framework to remain similar across projects.
Roughly speaking, we evaluate projects on the basis of three key questions.
How valuable is this project?
First, we look at how valuable the project is likely to be if it goes well. Roughly speaking there are three ways a project can go well: (1) the project itself produces some positive result; (2) the person running the project gains some valuable skills or experience; (3) some useful knowledge is created as a result of the project.
Ideally projects produce value along multiple dimensions, but we’re open to funding projects that produce value only on one dimension. For example, we may sometimes choose to fund projects where we are unsure of the object-level value of the project, if we think the project will produce useful knowledge for the community. While we didn’t fund this particular project, an example of a past project we consider worthwhile for the information it produced was Charity Science’s early work on grantwriting for effective charities, which did not achieve its goal of raising funds but did provide useful information to other EAs about whether grantwriting on behalf of other organizations was a viable project.
Is this person well-suited to this project?
We also look at whether the applicant is a good person to run the project. This breaks down into whether this is a good person to run an EA Grants project in general, and whether they seem suited to this particular project.
The general characteristics we look for in EA Grants applications are understanding of, and commitment to, the principles of effective altruism, demonstrated ability and drive, demonstrated ability to run projects autonomously, and overall good judgement. For evaluating fit between the applicant and a particular project, we look for evidence that the applicant has the skills and knowledge likely to be necessary to complete it.
What happens otherwise?
Finally, we attempt to roughly figure out the counterfactuals to us funding the project. In particular, we look at what the person would do if the project isn’t funded and whether they’d be able to access funding without our assistance. If we think that a person’s alternative plans are better than the project, then we may discuss that with the applicant. If we think the applicant can raise money easily from other sources, we may encourage them to do that instead.
Asking what happens otherwise should also predispose us in favor of funding people that are less well known in the EA community or funding projects that fall outside of established cause areas, as these projects are significantly less likely to be funded otherwise.
What kinds of projects should apply?
EA Grants has an intentionally broad mandate. We’re open to funding any project in any cause area. We expect to mostly fund less-established projects without easy access to other funding sources and we generally expect to commit less than $130,000 per project, although we may make exceptions in some cases.
One area we’re not currently funding is directly covering tuition expenses for people who want to go to grad school, although we are willing to fund related expenses, like living expenses to take a high-impact internship.
The projects that we funded in 2017 are available here. Some example projects that we’ve funded more recently include:
- Funding for a project working on productivity coaching for EAs
- Funding for some recent graduates to spend 6 months working out what cause area they should work on by working on some short independent projects.
- Funding to help a researcher transition from traditional political science into AI politics and strategy.
- Funding to support Charity Entrepreneurship in creating new high-impact charities.
- Funding for research on the psychology of effective and ineffective altruism.
If you have a project that you think is worth funding, you can fill out our short application here.
I have two questions:
Is there a full list of grantees and respective grant amounts for the referral-based round?
Is there some sort of evaluation process for funded projects that have concluded? I am curious especially about the outcomes of the projects that were funded in the 2017 round (that have now had the money for about a year). This question was asked about a year ago, but the details seemed uncertain at the time so I am re-asking the question.
1) We don't have a full list at this time. We're still in the process of figuring out what our approach to communication for EA Grant should be. My best guess is that we'll want to share details of some, but not all of the grants that we make.
2) Not at the moment. I'm planning to have the person we hire be in charge of this process since I think the feedback loops from past grants will be important for helping them make good grant decisions in the future.