# Allocation of discretionary funds from Q2 2019

In the second quarter of 2019, donors gave a combined $2.3 million to GiveWell for granting to recommended charities at our discretion. We greatly appreciate this support, which enables us to direct funding where we believe it can be used most effectively. We grant this funding to one or more of our top charities each quarter. We decided to allocate all$2.3 million to the Against Malaria Foundation (AMF). AMF is a GiveWell top charity that provides support for the distribution of long-lasting insecticide-treated nets to prevent malaria. AMF has been named a GiveWell top charity seven times. We chose to allocate the second-quarter funding to AMF because we believe AMF has a highly cost-effective and time-sensitive opportunity to spend it.

We continue to recommend that donors giving to GiveWell choose the option on our donation form for “grants to recommended charities at GiveWell’s discretion” so that we can direct the funding to the top charity or charities with the most pressing funding needs. For donors who prefer to give to a specific charity, we note that if we had additional funds to allocate at this time, we would very likely allocate them to AMF, which we believe could use additional funding for highly cost-effective work, even after receiving the $2.3 million in funding mentioned above. ## Summary In this post, we discuss: • what AMF will do with additional funding. (More) • other possibilities we considered. (More) • our process for deciding where to allocate funds. (More) ## What will AMF do with additional funding? AMF told us that it will use additional funding to support a distribution of nets scheduled for 2020 in the Democratic Republic of the Congo (DRC). Distributions are often delayed by a few months. Our best guess is that these nets will be delivered in late 2020 or in 2021. DRC has a higher malaria burden than most of the other countries where AMF supports distributions. We model AMF’s work in DRC to be more than 1.5 times as cost-effective as AMF’s past work, on average—we estimate that a donation of roughly$2,000 to support work in DRC will avert a death, compared to $3,600 for AMF’s work overall.1You can see our calculations by making a copy of our 2019 version 4 cost-effectiveness model; this will enable you to edit the sheet and change the values in the drop-down menu as described below: 1. Our estimate of the cost-effectiveness of AMF’s work in general: Go to the “Nets” tab. In cell B125, you’ll see the “Median cost per death averted (after accounting for leverage and funging)” for AMF. The value is$3,554.
2. Go to the “Country selection” tab, change the value for the Against Malaria Foundation in cell B7 from “Overall” (which includes all countries AMF works in) to “DRC” on the drop-down menu.
3. Our estimate of the cost-effectiveness of AMF’s work in DRC: Go back to the “Nets” tab. The value in cell B125 is now $2,072. 4.$3,554 divided by $2,072 = ~1.7. We consider this to be the most promising funding need among our top charities, in terms of timeliness and cost-effectiveness. Our process for comparing top charities’ needs each quarter is described in greater detail below. Open questions and uncertainties Although we see this as a very promising opportunity, we are somewhat unsure how AMF will actually allocate the funding it receives. AMF’s role in net distributions is to: 1. identify countries with funding gaps (funding needs that aren’t otherwise expected to be met) for nets; 2. find distribution partners (in-country non-profit organizations or government agencies) to carry out the distributions; 3. purchase nets; and 4. work with distribution partners to monitor the distribution and use of nets. While AMF has told us that it will allocate additional funding to DRC, it is possible that AMF will deviate from its funding plans in the face of changing circumstances, primarily changes in the status of discussions with governments and changes in the amount of funding it has available to allocate. The most common changes in AMF’s plans in recent years have been (a) delays in distributions, often due to governments taking longer to sign agreements than AMF had originally estimated, and (b) changes in the quantity of nets purchased by AMF due to larger population numbers being found during registration than the government had estimated at earlier stages of planning. According to AMF, the total funding gap in DRC over the next two years (2020-2021) is$55 million. In addition to its plans to fund work in DRC, AMF currently holds $39 million to fund distributions in three other countries. Although AMF is in discussions about funding these distributions, it has not yet signed formal agreements to do so. If any of the discussions fall through, we expect AMF to reallocate the funding it has set aside. In addition, AMF raised$38 million in 2018,2More specifically, this is AMF’s total revenue between February 1, 2018 and January 31, 2019, which is GiveWell’s 2018 “metrics year.” and we estimate that AMF will continue to raise at least half of that amount annually, independent of whether GiveWell allocates additional discretionary funding to AMF. This suggests that AMF will raise enough funding in the next year to substantially reduce the size of the funding gap in DRC, though the timing of when funding is received may affect the timing of distributions. If AMF fully fills the DRC funding gap, it seems intuitively likely that there would be other bottlenecks that might impede its progress, such as ability to find partner organizations with the capacity to implement the distributions and fulfill AMF’s reporting and monitoring requirements. We do not know where or when AMF would choose to fund nets if it had more funding than it could allocate to DRC in 2019 to 2021.

We incorporate our uncertainty about where AMF will use additional funding into our cost-effectiveness estimate of its work. When we made our first-quarter discretionary funding allocation, which also went to AMF, we modeled an 87 percent chance of AMF’s additional funding supporting nets in DRC.3This model has not been published. As we considered where to grant second-quarter discretionary funding, we made a minor downward adjustment to 75 percent due to AMF’s continued lack of signed agreements with other countries and thus our greater uncertainty over how funds will be spent. Even with this uncertainty incorporated, we model AMF’s funding gap in DRC as a highly cost-effective opportunity.

## Other possibilities we considered

Malaria Consortium’s seasonal malaria chemoprevention (SMC) program

When we granted discretionary funding we received in the first quarter, we focused on AMF and another top charity, Malaria Consortium’s SMC program, as the most promising recipients. Our decision centered on our comparison of the two organizations:

• Against Malaria Foundation
• We modeled additional funding to AMF as more cost-effective than additional funding to Malaria Consortium’s SMC program. Our best guess, which we did not subject to our formal internal review process, was that AMF was 38 percent more cost-effective than Malaria Consortium’s SMC program.
• We viewed AMF’s funding gap in DRC as time-sensitive because our expectation is that AMF receiving funding now will allow it to distribute nets sooner than if it receives the same amount of funding later this year.
• Malaria Consortium’s SMC program
• We viewed Malaria Consortium’s SMC program as likely to have more overall impact per dollar based on unmodeled qualitative factors described in “Principle 2” here.
• We did not expect that directing additional funding to Malaria Consortium would influence its spending on 2019 and 2020 programs—in other words, we didn’t see providing funding to Malaria Consortium as being particularly time-sensitive.

Weighing these factors, we ultimately chose AMF over Malaria Consortium based on its somewhat higher modeled cost-effectiveness and more time-sensitive funding need.

We now model additional funding to AMF as roughly 33 percent more cost-effective than additional funding to Malaria Consortium’s SMC program, as a result of adjusting the chance of additional funding supporting nets in DRC from 87 percent to 75 percent. We have not received any new information to update us on the time sensitivity of Malaria Consortium’s funding needs, and we continue to view Malaria Consortium as stronger than AMF on unmodeled qualitative factors.

We don’t view the comparison of the two organizations as meaningfully different than in the previous quarter, and we thus chose to prioritize AMF over Malaria Consortium again.

Other top charities

As far as we know, our six other top charities have not had any major changes in their funding needs or cost-effectiveness since March. We did not update our cost-effectiveness model since making our last quarterly allocation decision, nor did we receive any updates on our top charities’ room for more funding, beyond the $4.7 million in first-quarter discretionary funds that we allocated to AMF. ## Process for deciding where to allocate funds We follow the principles described in this blog post when deciding between funding opportunities. We ask our top charities to alert us throughout the year if they learn of any new funding opportunities that we should consider in our discretionary funding decisions. None of our top charities informed us of such an opportunity for second-quarter funding. With no new funding opportunities presented to us, we returned to our first-quarter funding recipient, AMF. When we granted first-quarter funding to AMF, we noted that AMF had a time-sensitive and cost-effective funding opportunity in DRC and a funding gap that was much larger than we were able to fill. As we considered where to allocate second-quarter funding, we asked AMF for information to help us assess whether that continued to be true. We asked AMF about its progress in signing net-distribution agreements, its ability to absorb additional funding for work in DRC, and whether additional funding sent to AMF in the next few months would contribute to filling the funding gap in DRC. Notes  ↑1 You can see our calculations by making a copy of our 2019 version 4 cost-effectiveness model; this will enable you to edit the sheet and change the values in the drop-down menu as described below: 1. Our estimate of the cost-effectiveness of AMF’s work in general: Go to the “Nets” tab. In cell B125, you’ll see the “Median cost per death averted (after accounting for leverage and funging)” for AMF. The value is$3,554. 2. Go to the “Country selection” tab, change the value for the Against Malaria Foundation in cell B7 from “Overall” (which includes all countries AMF works in) to “DRC” on the drop-down menu. 3. Our estimate of the cost-effectiveness of AMF’s work in DRC: Go back to the “Nets” tab. The value in cell B125 is now $2,072. 4.$3,554 divided by \$2,072 = ~1.7. ↑2 More specifically, this is AMF’s total revenue between February 1, 2018 and January 31, 2019, which is GiveWell’s 2018 “metrics year.” ↑3 This model has not been published.

• Thanks for this. I was reviewing the July impact calculator recently and was wondering why the impact per dollar was so much greater for AMF than it had been previously. Targeting net distributions in the DRC explains it adequately.

• Pallavi on August 27, 2019 at 6:33 am said:

Beautiful Article, Raise the fund for charity, NGOs is the best job. We always try to raise the fund for who needs the most.

• “ We continue to recommend that donors giving to GiveWell choose the option on our donation form for “grants to recommended charities at GiveWell’s discretion””

Would an unrestricted gift to GiveWell be even more effective than this recommendation since it allows for all possible uses (e.g. communication) including grants? If so, how much more effective?

I am preparing a donation to GiveWell so this has immediate applicability. My thinking on this subject is largely based on Dan Pallota’s TED Talk: “The way we think about charity is dead wrong“. I would be very interested in reading a GiveWell deep dive on this.

• Catherine (GiveWell) on October 24, 2019 at 12:16 pm said:

Hi Ben,