A conversation with charity: water on April 13, 2012


GiveWell: You've described examples of people you've helped who had extremely poor access to clean water before you arrived. Would you say these examples are representative of the places you work? 

charity: water: Yes, the water supply situation where we work is extremely dire.  The water sources are filthy, or miles away, or miles away and filthy.  We’ve been to places where you can smell the water before you see it.

GiveWell: Can you elaborate on the specificity of your staff site visits? Would a staff member visit each village?

charity: water: Staff from our Water Programs team visit every country program before we make a grant. For example, they travel to a prospective country, meet with the country program staff, visit completed water projects as well as communities that will be supported under the grant. This gives us an understanding of the type of organization we’re supporting, the work they have implemented, and the areas where our funding will support. Within a grant, we’ll fund many water points, from 20 to over 100.

Our Water Programs team has regional responsibilities — W. Africa, E. Africa, Asia and Latin America — and travel extensively throughout the year. However, our staff simply cannot see every village. Our grants have funded over 6,000 water points and it is not feasible to visit every one. Our largest investment is in Tigray, Ethiopia, which we visit frequently and have seen more than 300 sites. But even this is only a sample of the given territory.

Because our staff have regional responsibilities, and charity: water has a global presence, we can take the opportunity to share best practices across countries for our partners. For example, “we have this challenge with fluoride in the water and country X, and our partner in country Y can share their experience in addressing the same issue.” The communication and interest in sharing information helps us all do our work better in the focused region.

We currently share knowledge in an ad-hoc manner, and that’s an opportunity that we could continue to improve in this space across all of our implementing partners. We talk about doing more information sharing because if we can put our partner in Orissa, India who’s focused on 100% sanitation in contact with another partner in Malawi who’s struggling with sanitation … something magical would happen.

GiveWell: So one thing to know about GiveWell is that we are very skeptical when it comes to charities we evaluate and appreciate physical materials. Are there documents that when charity: water comes back from a trip that say we check for a specific criteria A, B, C, and D in specific number of water points? Do you identify that X numbers of water points were good and X number of water points failed? Or is it more informal?

charity: water: We have trip reports from every field visit, which reviews the water programs on a macro level — not a proper audit of implementation. Then, for programs that are over a certain size — over a million dollars — we do a formal technical and financial audit. This involves evaluating the quality of work and the final cost of the program through an independent audit that we pay for out of our operations budget.

We have trip reports and financial reports that we are happy to share with GiveWell and Good Ventures.

GiveWell: Another thing we wonder about is what was the access to clean water before charity: water came in. For example, how clean is clean? What are the different levels of access? So if that’s a question that GiveWell wanted to answer about charity: water, would that be an answerable question?

charity: water: In general, the quality of previous sources vary widely. Communities will use open sources such as rivers, lakes, and streams, and the water is filthy. We often find animals living in the water sources, or cattle grazing near by and defecating into the source.

We also see broken wells — so at one point in time the community did have access to clean water.  But for various reasons, the water point fell into disrepair.  A fair amount of our funding goes into fixing wells from prior programs installed by national governments and other organizations. We see this a lot in post-conflict countries, but significantly in stable contexts, too.

The number of people per water point is important, and we try for the standard of 250 people per water point. We’re not always able to achieve that. Occasionally, we’ll go to Sphere standards (500 people per water point) or the national government standards which falls somewhere between those. However, we can’t force any of our grantees to go against national standards or policies.

One of the challenges we face is in the choice between making sure that everyone in a given community or district gets at least one water point (even if that is more than 250 people per water point)? Or do you make sure that a community has enough water points to have 250 people per water point at the expense of other communities getting water in a timely manner? We honestly don’t know what the right answer is, but we  do both, following the local guidance from our implementing partners.

In Tigray, Ethiopia, we work strategically with our partner. They operate in clusters to minimize mobilization costs, but they also have the flexibility to reach out to other areas outside the clusters with great need.

The important thing is that there’s always a plan — whether if it’s following the government standard, or reaching out to communities of greatest need.

GiveWell: Going back to the prior question, how do we figure out the before and after states prior to charity: water’s intervention?

charity: water: Some of our partners give us prior water source data for the communities. You see the same thing over and over again — people that don’t have water and walk miles to get it and people that have nearby sources, but they are heavily polluted.

You go to places like Liberia where it’s water rich and the distance isn’t very far but the quality of water is terrible. Also, distances don’t necessarily reflect the amount of time it takes — nearby sources may be insufficient for the community and so the line is long. Or the terrain is very difficult, taking more time to travel.

We have some of this data, but we’ve found that it varies considerably.  The challenge is to find a way to draw conclusions despite the variability.

GiveWell: On your website you list the GPS coordinates of different sites. Could we pick out some of the coordinates and then ask you share as much information as is available about particular sites?

charity: water: Yes, happy to. However, the level of detail will vary.  We always collect a minimum of data — e.g., GPS, photo, population size, community name, and cost.  Some partners give us more information than that, while others do not.

GiveWell: The academic evidence about the impact of water programs on health is mixed and doesn't consistently show large impacts. What's your take on why this is?

charity: water: It’s so difficult to determine causality between water, sanitation, hygiene and health impacts. The other side of the coin is time saved, which is not necessarily health impact but the measure of productivity impacts. We are not a research organization so we would not be able to demonstrate with any degree of robustness the poverty or health impact of work. We rely on information coming from researchers such as Esther Duflo, or organizations that do research such as the World Bank.

We think you do find more compelling evidence in the developed world in terms of the impact of water supply. Say we look at NY, London, Paris — cities that have gone through the safest in drinking, sanitation, etc. You would be able to find better information on demonstrated impacts. But for the areas where we work in developing countries, there are too many variables to demonstrate that causality.

GiveWell: Is it right that when you say it’s hard to demonstrate causality, you believe that these programs have health impacts, but you just think that it hasn’t been demonstrated in academic research because of limitations of the research?

charity: water: We think when you dig into the research, you see limitations within the research, typically small sample sizes. The research usually shows how well that specific intervention works but not necessarily transferable to other interventions.

You also know that a lot of young kids die from diarrhea. However, once a child survives 5 years, the impacts from diarrhea lessen and it is rare that an adult dies of diarrheal diseases. But the damage has been done already in the first years of their life — chronic diarrhea manifests itself as malnutrition, which can impact cognitive development and other health problems.

Water is the baseline in regards to health, education, and economics. In the projects we fund, we don’t have a lot of data just yet because the projects are so young and some of the communities are so new. Simple before and after data — a woman had to walk this far compared to before is data that we can find, but health and economic impacts is relatively challenging.

GiveWell: Are you worried that you might be funding areas that already have reasonable access to water?

charity: water: That's not a worry of ours. We invest quite a lot in vetting programs, and we know the need is massive.  What keeps us up at night is the long-term sustainability. To that end, we’ve started funding programs that are a little bit out of the box such as setting up well mechanic businesses in India to fix wells. We want to be able to promise a donor that their water point would have a minimum usage of 10 years. We’re pushing for more innovation in this space and trying to share what our partners are already doing. For example, our partner in Central Africa Republic visits charity: water sites at least three times a year to do preventative maintenance and collect performance data, which is a dream for us.

This is our primary concern: are water projects sustained over time? We would say that that the industry practice is to work with communities, set up water committees and help create a structure so that if a water point fails, they know what to do. At the moment we’re not strong in collecting data, but we’re working towards becoming a data-driven organization. We know that, for example, 160 water projects were built, we know they’re functioning, but what we really want to know if they’re functioning 5-10 years from now.

More recently, we’ve begun planning to start a new fund that will support more capacity building for our partners and help fund training. Additionally, we’ve had discussions to figure out ways on how we could leverage folks who are interested in this issue and go out to collect data.

GiveWell: So the big question that keeps coming up for me is how to understand the two parts of charity: water’s work. The first part is: getting people excited and funneling funds. The second is to find the best projects and pick the right partners to implement water. We can see how charity: water does a great job with the first, but would like to further understand the latter.

charity: water: We look for excellent partners who have a strong track record of implementing water projects in the countries where they work. We find these partners through various mechanisms — perhaps we’ve worked with them in another country, or colleagues in the sector have recommended them to us. But no matter what, we vet all new programs through extensive criteria, both financially and technically.

We want to demonstrate integrity in the system, and do that by showing where all the money goes. What we hope to do is to raise the bar on transparency and sustainability. We are continuously improving in our work.

GiveWell: Are you turning away a lot of partners that you don’t think are good?

charity: water: Yes. But we wouldn’t want to give names to protect partners in the space. But it’s like an investment portfolio that we develop to identify the level of risks of various partners doing this work. There are many organizations that submit proposals but don’t meet our criteria. We don’t act like a foundation and we don’t think of ourselves as grant makers. We spend an incredible amount of time vetting our projects. We believe we’re really good at this and understand the skeptical approach is important. We want to give our donors the highest returns on their investments. 

GiveWell: Why should a donor give to charity: water instead of, for example, Water for People?

charity: water: The value we add is that we show the donor where his money goes — every dollar — and they can be certain that 100% of their funds supported a water program in the field.  We’re also really good at telling stories, and utilize those skills to raise awareness, mobilize resources, and bring people closer to the impact they’re making around the world.

GiveWell: An important question to us is something we call 'room for more funding' — where would the marginal dollar donated go? What is going to be different in 2013, if charity: water raises a significant amount of funds relative to what you currently project?

charity: water:  There’s a few ways we approach this: first of all, we focus a great deal on planning, including weekly forecasts that facilitate strategic discussions and alignment between our fund raising and water programs teams. This allows us to manage capacity on an ongoing and systematic basis.  Secondly, our water programs team builds in a certain amount of flexibility for surprises. For additional, unexpected funds we have an “accordion” plan, which are proposals we can grant to at the last minute should additional resources come in. Additionally, we manage our grants constantly, potentially shifting grants from one quarter to the next. This involves constant communications with our implementing partners to properly manage expectations.  Finally, it’s our policy to leave a certain amount of funding on the books at the end of each fiscal year, based on factors such as current economic conditions, giving trends, and first quarter forecasts of the subsequent fiscal year.  Our objective is to provide as much capital as possible in any given fiscal year to bring people clean water, but also to remain an ongoing, consistent, and reliable source of funding for our implementing partners. Marginal dollars often help bridge this gap.