We wish we had more top charities, and as we look to the future we expect (and hope) that there will need to be more recommended charities in order to productively use all the donations that GiveWell-influenced donors are making. One of our major activities is trying to expand our top charities list – both by investigating charities that already exist, and by supporting activities (from new nonprofits to studies) that could eventually result in a larger set of evidence-backed programs and a larger set of top charities.

This post discusses types of charities that we would be excited to learn more about if they existed. We would also consider providing support to individuals trying to create the types of organizations described below. In a similar spirit to a request for startups, we’re sharing this list in the hopes that it might help us find out about such charities – or might help us find and support people looking to create them.

In brief, we would be excited to see:

• Charities that implement GiveWell’s priority programs: vitamin A supplementation, immunizations, conditional cash transfers, micronutrient fortification, or even bednets and deworming (since our top charities that focus on the latter two have limited room for more funding). More
• Charities implementing potential priority programs that are particularly challenging, particularly those revolving around (a) treatment of treatable conditions in a hospital or clinic setting; (b) behavior change for improving health. We see several hurdles to successfully focusing on such programs, but would be excited to see charities that overcome such hurdles. More
• Charities that collect or generate information and data relevant to our recommendations. Currently, we recommend charities based partly on the data they themselves collect and share. But we could potentially recommend an organization that does not, itself, collect and share strong monitoring data, if we had independent data showing its activities’ effectiveness. More

Charities that implement GiveWell’s priority programsThe first step in our research process is considering the independent evidence of effectiveness and cost-effectiveness of a program to determine whether we should classify it as a priority program. The programs that our top charities implement — bednet distribution, cash transfers and mass deworming programs — are among our priority programs.

There are cases where we have classified a program as a priority, but have not found a promising charity that focuses on that program.

We would be excited to see new charities implementing our priority programs who plan to publicly share significant monitoring and evaluation data and generally expect to be extremely transparent about their work.

Some specific examples of organizations we would be interested in:

• Providing vitamin A supplements to areas with high rates of vitamin A deficiency and child mortality. One illustrative example: the Central African Republic has a high child mortality rate (139 deaths per 1,000 children under 5, data from Gapminder here) and low rates of vitamin A supplementation (40%, data from the World Bank here). More on vitamin A supplementation and the questions we would ask a charity to answer here.
• Providing immunizations in areas with low coverage rates. An illustrative example: the Central African Republic has low rates of measles (49%) and DTP (diphtheria, tetanus and pertussis) immunizations (47%). (Data from UNICEF here.) We have not recently completed an intervention report on routine immunization but our report on maternal and neonatal tetanus immunization campaigns lays out the questions we would use to evaluate charities.
• Providing conditional cash transfers to encourage school attendance, clinic visits, etc. while also transferring wealth to low-income people. (Note that we currently support a charity taking this approach, in the hopes that it may become a top charity in the future.)

There are also some cases where we have identified charities that run priority programs, but are not completely satisfied with the evidence we’ve seen for the charities’ track records, either because we don’t find the case for impact compelling or because the charities have been hesitant to share information to the degree necessary. In these cases, we would be excited about organizations that implemented these programs and shared a significant amount of information about their work.

Possibilities in this category include:

• Organizations working on micronutrient fortification, such as salt iodization or folic acid fortification.
• Organizations working on immunizations campaigns for tetanus or measles.
• Organizations working on mass drug administration for neglected tropical diseases other than deworming.

Note that in some of the above cases, we have not completed our intervention reports, so it’s possible that we might conclude that a program does not have sufficient evidence of effectiveness or is not cost-effective enough to be a priority program. In general, we allocate our time with the goal of finding the combination of a strong program and strong organization; there are promising programs we have not completed our investigations of because we have not found promising charities running them.

Finally, we would be excited about organizations working on priority programs where we currently have top charities that may have limited room for more funding. In particular, we would be excited to see a new charity focusing on bednet distribution or deworming programs, while having a high willingness to collect and share data.

Note that we have intentionally not included areas like surgery, education, and family planning in this post. We have not prioritized research in these areas recently, but we do hope to revisit them in the future. Because we have not recently looked into these areas, we want to put more time into determining whether existing organizations may be able to meet our criteria before calling for new ones.

Charities implementing potential priority programs that are particularly challengingTwo types of promising programs are those (a) that focus on treating specific diseases or (b) that aim to promote behavior change. We have not completed intervention reports for either because both seem particularly challenging for charities to implement successfully.

Treatment programs

Treatment programs would focus on treating individuals after they contract a disease (e.g., malaria, HIV/AIDS, pneumonia, diarrhea, and tuberculosis). Unlike the programs we currently recommend, which target all members of a population (e.g., all women of childbearing age, all children under age 5), treatment programs are significantly more complicated. To receive treatment, (a) an individual generally must go to a clinic when s/he requires treatment, (b) be accurately diagnosed as having the condition, and (c) the clinic must have the necessary drug in stock. In some cases (e.g., HIV/AIDS or tuberculosis), the individual would have to return to the clinic and replenish his/her supply of the medication and also adhere to a long (or perpetual, in the case of HIV/AIDS) treatment regimen.

Because of the costs associated with providing treatments (training skilled diagnosticians or keeping drugs in stock), it seems unlikely that a charity should focus on providing just one of the above treatments. Possible approaches a charity could take include:

• Setting up high-quality clinics that provide treatments and other medical care. We have previously reviewed and recommended two organizations that follow this model: Partners in Health and Possible. Our guess is that clinics are unlikely to be as cost-effective as our current top charities.
• Running a program that involves community health workers to provide a limited range of treatments. This is somewhat similar, though not identical, to Living Goods’ model. Our impression is that there is a relatively large literature related to programs implemented by community health workers and the quality of the services they provide vary widely. Were we evaluating a charity implementing this model, we would be particularly focused on its monitoring and evaluation data as well as its cost-effectiveness.
• Providing treatments or diagnostics to clinics that would otherwise not have them. Were we evaluating a charity implementing this model, we would seek compelling evidence that the charity is causing the clinics to have access to treatments it otherwise would not have had, that the treatments are ultimately provided to people who need them (i.e., are accurately diagnosed as needing them) and that the recipients follow the prescribed regimen. Note that that this model seems similar to the some of what the Global Fund does. We evaluated the Global Fund in 2009 and 2010 but were not able to obtain the data we needed to recommend it.

We have not completed intervention reports for treatment programs, but we would likely consider them priority programs were we to find a charity effectively implementing one of the models above.

Behavior change

We have also not completed intervention reports for behavior change programs. Our impression is that programs to promote handwashing, breastfeeding or other health behaviors could be highly cost-effective because they can reach many people at low cost. (Note that Development Media International, a standout charity of ours, implements a behavior change intervention using mass media.)

The key challenge facing behavior change charities is demonstrating that their intervention causes behavior change and that that behavior change improves health. DMI is running a randomized controlled trial of its program to address this question, but very few charities are in a position to run a study like this. In GiveWell’s early years, we recommended PSI, a behavior change organization, but we changed its recommendation status because we no longer felt the evidence for its effectiveness was sufficiently strong.

We would be excited to evaluate a behavior change organization that could make a compelling case for its impact and is ready to share significant information about its activities with us.

Charities that collect or generate information and data relevant to our recommendationsThe types of charities discussed above are ones that could meet our criteria and receive a recommendation. We would also be interested in supporting groups that generate information or data that could inform our recommendations.

This could include groups that:

• Collect data that directly informs our views on current and potential top charities or the programs they implement. For example, Good Ventures provided funding (based on our recommendation) to IDinsight to conduct additional monitoring on the Schisotosomias Control Initiative’s programs. We could also imagine IDinsight, or a group like them, collecting and sharing better data that would inform our view of large-scale bednet distrubtions implemented by groups other than the Against Malaria foundation, salt iodization programs, or tetanus immunization campaigns. In all of these cases, we could recommend an organization that does not, itself, collect and share strong monitoring data, if we had independent data showing its activities’ effectiveness.
• Run randomized controlled trials (or replications of RCTs) of interventions that could be at least as cost-effective as our current priority programs. We’d be particularly excited about interventions that could plausibly be significantly more cost-effective than our current top charities.
• Qualitative research or journalism that informs our views of top charities or the programs they implement. This could include (a) articles about current top charities or the programs they implement (e.g., these pieces by Jacob Kushner that we commissioned in 2014 and 2015), (b) surveys of people served by the programs we recommend or the aid community in general, (c) research that directly addresses unanswered questions in our research (e.g., to what extent do individuals served by Development Media International’s program have access to clinics that can diagnose their conditions and provide them with medicine?), and/or (d) provides additional context on the lives of people living in extreme poverty, among others.
• GiveDirectly has suggested the idea of creating a facility for funding and implementing cash transfers as a control group for randomized controlled trials of development interventions. GiveDirectly has told us that this is something it doesn’t plan to currently prioritize and that it would be excited to see another organization undertake this. (It has expanded on this idea, suggesting a broader mandate for supporting cash transfer work, in this recent article.)