Here's a blog post about how GiveWell decided to fund a new initiative in India. I think it's interesting to see what they take into account and how little they trust their quantitative estimates:...
Here's a blog post about how GiveWell decided to fund a new initiative in India. I think it's interesting to see what they take into account and how little they trust their quantitative estimates:
We believe this grant may not only increase access to chlorinated water in the states it directly supports, but also inspire other states to adopt similar practices. A core part of the program’s theory of change is that governments in locations outside the grant area may take up a program they might not otherwise adopt. This is the first very large grant we’ve made where that’s been an important consideration. We think the upside is unusually high—if successful, this grant could eventually lead to tens or even hundreds of millions of additional people receiving safe water—but it’s also riskier than most of our grants, as there are a number of ways the program could fail to have the desired impact. Our hope is that this grant will reduce mortality and improve health at a very large scale.
...
We estimate that this program will reduce overall mortality by about 4% among people reached with in-line chlorination. Given that we anticipate the program may reach tens of millions of people over the course of around a decade, the impact could be huge—our very uncertain best guess is 75,000 deaths averted.
...
This kind of partnership between a nongovernmental organization and a government body creates the opportunity to leverage government resources for higher impact. This grant is large—$38.8 million—but is a small fraction of the overall $44 billion committed by the government to Jal Jeevan Mission. Government investment of this scale means philanthropic support can be focused on providing technical assistance at the national and state level rather than on providing in-line chlorination itself.
...
Our best estimate is that this grant is 22x cash—i.e., we estimate that each dollar spent on this grant will create as much value as roughly $22 in unconditional cash transfers to people living in poverty. The expected value of this grant is an additional 26 million people receiving chlorination for ten years; that figure includes subjective adjustments for both the upside discussed above and downsides like the risk of the program being discontinued. Some of the most uncertain factors in our analysis are discussed in the next section.
...
We’re quite uncertain about our estimates. For example, we think there’s a 50% chance that chlorination reduces mortality for all ages by somewhere between 1% and 7% (and a 50% chance that it reduces mortality by more or less than that range). If mortality were reduced by 1%, the program would have a cost-effectiveness of 5x cash, and if mortality were reduced by 7% the program would have a cost-effectiveness of 37x cash—so our assumptions here have a big impact on the bottom line!
Here's a blog post about how GiveWell decided to fund a new initiative in India. I think it's interesting to see what they take into account and how little they trust their quantitative estimates:
...
...
...
...