[Added August 27, 2014: GiveWell Labs is now known as the Open Philanthropy Project.]

While we haven’t spent as much time as hoped on GiveWell Labs, we have made some progress. This post summarizes how we’ve spent our time, what we’ve learned, and what we’re planning next.

We’ve put substantial time into each of the following:

• Partnering with Good Ventures on “co-funding” with major foundations. Good Ventures has spoken with multiple major foundations (some conversations off the record, some on the record), asking for recommendations for particularly strong projects that Good Ventures might join in funding. It has committed $1 million to a project with the Gates Foundation and is in the process of considering other possible projects. We have sat in on most of the relevant conversations, advised Good Ventures, and published notes when possible to our website. We had hoped this experience would give us an opportunity to “learn from the pros” – to understand how established foundations go about sourcing and evaluating giving opportunities. To some extent, it has; for example, we’ve learned about the importance and pervasiveness of active funding among major foundations, and we’ve gotten a sense for what program officers do to find giving opportunities (it seems that the most common core activity is networking intensively with people in the field, while emphasizing their own strategic priorities). At the same time, we’ve found the giving opportunities themselves difficult to evaluate, both before and after execution. As discussed previously, the information that is easily available is often not in-depth enough for us to gain high confidence regarding the merit of the project without substantial further investigation (the case for these gifts relies largely on the fact that the partner foundation finds them worthwhile, as well as the learning opportunities they present to us).
• Top-level investigation of the history and current state of philanthropy. We wrote about our understanding of philanthropy’s success stories, and collected data on philanthropy’s current funding allocations. Since then, we’ve continued to look for more information about the history of philanthropy – we think it’s important to understand what the major success stories have been and how they came about – but we have found little. We are now exploring the possibility of recommending funding to produce more such history.
• Meta-research. Last year we expressed an interest in meta-research, or “improving the incentives in the academic world, to bring them more in line with producing work of maximal benefit to society.” We ended up spending much of our GiveWell-Labs-related time on this cause because we not only found it promising (and had significant starting context on it due to our direct experiences as consumers of research), but also found it to be a fairly nascent field. Thus, we had opportunities to explore opportunities and participate in meetings that would have been difficult with our level of experience in other areas.
  • We were originally interested in meta-research for the field of development economics, the field most directly relevant to our work. We joined a major funder in having preliminary exploratory conversations with relevant researchers (some were off the record, but some have notes posted to the “social sciences” section of this link) and I attended an April 2012 meeting at the William and Flora Hewlett Foundation regarding preregistration.
  • We recommended a quick grant to the U.S. Cochrane Center, and in the process of investigating it we turned up some evidence that medical meta-research is a promising (largely because neglected) area. We are now finishing a more thorough investigation of that cause.
  • We also put a small amount of time into exploring the idea of meta-research for the “harder” sciences (biology, computer science, etc.) We advised a volunteer, Paul Christiano, who has had multiple conversations (notes forthcoming) in the field of computer science, and I attended the first meeting (notes) of Vannevar, a group started by Dario Amodei to explore ways to make scientific research more efficient, collaborative and productive. Dario and I also had a preliminary conversation about the world of biomedical research.
• We advised Good Ventures on its investigation of the cause of drug policy reform, seeing it as an opportunity to start learning about politics without a major commitment. Two consultants were retained to look into this area, and we aim to publish a public version of their report in the next month. I’ve also had many informal conversations about the world of political advocacy, and am starting to form a plan for understanding it better.

Our major takeaways to date from these investigations:

• Due partly to the potential importance of active funding, it appears helpful to think of a “cause” or “sector” – rather than a “project” – as the most relevant unit of inquiry. We now plan on investigating a large number of potentially promising causes at very low depth, and investigating a smaller number at a higher level of depth.
• Most of philanthropy seems to use some combination of (a) direct service delivery; (b) funding of research; (c) political advocacy. We feel that we have done substantial research on, and attained substantial understanding of, (a), while we understand very little about how (b) and (c) work in general. It is a priority for us to learn more, generally, about how the world of scientific research works (which involves understanding the incentives and evaluation mechanisms for academics) and about how the world of political advocacy works (which involves understanding the basic tools that are used in advocacy and how one might expect their effectiveness to vary with different issues).
• It is very difficult to get a sense of what has worked, and failed, in philanthropy’s past. Our value of transparency is hopefully a step in the right direction for the future, but for our own learning we may find it necessary to do substantially more investigation of history.

Accordingly, for the next few months we expect to prioritize the activities of

• Investigating a relatively large number of causes at relatively low depth, and investigating a smaller number at high depth. (We are currently finishing a higher-depth investigation of medical meta-research.)
• Forming plans for improving our general understanding of scientific research and political advocacy.
• Exploring the possibility of funding journalists and/or historians to produce better data on what has worked and failed in philanthropy in the past. (More)