EA - Changes to our top charity criteria, and a new giving option by GiveWell

Link to original articleWelcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Changes to our top charity criteria, and a new giving option, published by GiveWell on August 17, 2022 on The Effective Altruism Forum. Since 2007, GiveWell has maintained a list of top charities. The organizations and programs on that list have changed over time, but the goal has remained the same: to help donors decide where to give. In service of that goal, we’ve spent roughly the last year working on a plan to add to and update our criteria for top charities, so that they accurately reflect our prioritization of funding opportunities and are more helpful to donors. The revised criteria emphasize programs from which we expect high impact, with the additional requirement that we have a high degree of confidence in our expectation. We hope these changes will also draw a brighter line for donors between our top charities and other excellent funding opportunities we support, a distinction that we haven’t always made clear. Recognizing that some donors want to contribute to grants outside our top charities list, we’re also introducing a new giving option: the All Grants Fund. Providing this option as a complement to the Maximum Impact Fund is an important step as grants to programs outside our top charities become a bigger part of our work. In this post, we will: Explain why we are making these changes Share our updated criteria and our updated top charities list Explain how these changes affect donors’ giving options Why we’re revising our top charity criteria Our top charities list—and by association the Maximum Impact Fund—is often the first thing that people encounter when they’re introduced to GiveWell. We want it to be easy to understand why these programs are on our list. Additionally, our top charity programs should reflect the kinds of giving opportunities that people are seeking from such a list. After soliciting feedback, we believe that most donors are looking for opportunities that are both high-impact (i.e., GiveWell expects donations will do a lot of good per dollar) and high-confidence (i.e., GiveWell is relatively sure that that expectation will bear out).[1] By adding criteria that are more specific, we also hope to establish clearer guidance on when programs should be added or removed. Our revised criteria Previously, we listed four criteria for top charities: evidence of effectiveness, cost-effectiveness, room for more funding, and transparency. These continue to be the criteria we’ll use to evaluate funding opportunities, but they no longer fully describe the few programs that we consider top charities. Those programs will meet the following additional, more precise criteria: We’ve directed a significant amount of money to this program and seen it operate effectively at scale. A program meets this criterion if: We’ve directed a single grant to it of at least $10 million or multiple grants totaling at least $20 million, and We’ve supported the program as carried out by the implementing organization for at least one year. It’s a good sign of our confidence if a program fulfills these two requirements. If we’ve provided significant, sustained support, that means we’ve already analyzed it thoroughly and judged it likely to clear our bar for funding. And if we’ve supported it for a year or more, it’s less likely (though possible) that we’ll receive information that would dramatically update our cost-effectiveness analysis. We think there’s a high likelihood of significant impact from funding this program (as opposed to lower likelihood of enormous impact). The goal is to move grants with higher risk profiles into our other portfolio of grantmaking. For example, we fund public health advocacy because the potential benefit is so large that these grants have high expected value—even though there is a chance they do not have a significant impact. We feel th...