On December 4, 2025, we held a panel discussion titled “Growing Needs, Shrinking Aid: Cost-Effective Action in a Year of Funding Cuts.” The discussion, which was moderated by GiveWell co-founder and CEO Elie Hassenfeld, addressed the effects of recent cuts, how GiveWell is responding, and what we’re learning along the way—including our predictions and uncertainties about the future. GiveWell researchers Alex Bowles, Dilhan Perera, Meika Ball, and Rosie Bettle answered questions and shared their latest insights. Watch the video or read the transcript.

Because of limited time, we were unable to answer all of the questions live. Below we’ve answered some of the questions that weren’t addressed during the webinar, along with some additional questions submitted by audience members, focusing on four themes:

• The challenges of an uncertain future
• Taking action in the face of uncertainty
• The role of cost-effectiveness
• Future GiveWell grantmaking

The challenges of an uncertain future

What is the biggest impact of the cuts you have seen in your research area?

Meika Ball: My work has focused on health systems strengthening, which are interventions that strengthen the foundational systems and processes needed to deliver health services, such as supply chain management, data systems, and health financing. Health systems strengthening is a new area for GiveWell. We prioritized it this year largely because a lot of this work had lost funding after previously being supported by the US government and other donors.

We have consistently heard about large cuts to this work, though the full magnitude has been challenging to understand. Especially during the stop-work order, we heard about countries losing access to their own data systems, health supplies being stuck at various points in the supply chain, and staff being laid off. This loss of staff capacity was especially evident during our team’s recent trip to Zambia. Officials in one province told us they had lost funding for around half of their staff. Community-based outreach work, where health worker stipends had been cut, was also impacted.

Because various components and processes within health systems underlie many other health programs, the cuts to health systems have been particularly challenging. I think strengthening health systems, including support for health workers, is likely to be a substantial need in the future as governments face more limited budgets.

Could you share more about the data challenges arising from funding cuts and provide an example of how it is impacting your work?

Dilhan Perera: Understanding the impact of changes to US government funding for HIV services has been very challenging. Many programs halted by the stop-work order were eventually terminated, but others were resumed—and sometimes revised or expanded to cover gaps left by the terminated programs for lifesaving services. In addition, some people who were accessing services through terminated programs may have found other ways to access similar services, such as through government-funded health facilities. On the other hand, barriers like stigma may have prevented some people from going to those facilities.

Data from President’s Emergency Plan for AIDS Relief (PEPFAR), the US government’s initiative to combat HIV, used to be released quarterly with information about service delivery across all of the programs and clinics PEPFAR supported. Data for 2025 has not yet been made available, and there’s no timeline for when the data might be released. Moreover, funding disruptions impacted the capacity of some programs to continue collecting accurate data about service provision. As a result, we do not have a clear picture of to what extent reductions and resumptions in US-funded HIV programs translated into a reduction in service delivery or of which geographies and populations have been most affected.
All of this has made it difficult to know which places and groups have lost access to health services and to evaluate how cost-effective it would be to provide new funding or reinstate funding for terminated programs.

Taking action in the face of uncertainty

What was the most impactful action you took this year to respond to cuts?

Rosie Bettle: Let me speak specifically to our actions in response to malaria treatment needs, rather than to the most impactful actions of GiveWell as a whole. As a result of the cuts, some countries have been running alarmingly low on essential malaria supplies. These include rapid dual tests, or RDTs, which are the test kits used to diagnose malaria; artemisinin-based combination therapies, or ACTs, which are the standard first-line treatment for malaria infection; and artesunate which is specifically for severe malaria cases.

Normally, organizations funded by USAID, like the President’s Malaria Initiative (PMI), order these supplies from manufacturers and ship them to countries in need. This process typically takes about six months from when an order is placed to when it arrives in the country. But when PMI was put on a stop-work order, many orders were severely delayed.

We worked closely with the Clinton Health Access Initiative (CHAI), who tracks how much stock countries have left and how quickly they’re using it up, and we started seeing evidence that countries were about to run out of these life-saving medicines. Our rapid response research team made five grants to tackle these shortages. We funded organizations like CHAI and PATH to procure supplies directly from manufacturers on an emergency basis and ship them very quickly to countries facing gaps caused by the delays.

We think that the work we funded through CHAI helped prevent a national stockout of ACTs in Malawi, although it’s always difficult to work out what would have happened in the counterfactual. We also think this work shortened a national-level stockout of RDTs in Nigeria. We also estimate that we reduced periods of extremely low stock—which we think leads to increased clinic-level stockouts during which patients show up at local health facilities and aren’t able to be tested or treated—for severe malaria treatment in Mozambique and for RDTs in Senegal.

What expertise proved most valuable this year, and where did you most need to adapt?

Alex Bowles: Uncertainty about what other funders are planning and doing has been a key challenge this year. There has been a substantial risk that we would either miss urgent cost-effective funding gaps or would fill funding gaps that ultimately would have been filled by another funder.

The effort we’ve put over the last couple of years into building our networks among others working on malaria prevention and treatment has paid dividends. Our partners—other funders like the Global Fund as well as grantees and other implementers like PATH, Clinton Health Access Initiative (CHAI), Against Malaria Foundation, and Malaria Consortium—were able to share information with us quickly. We want to be sure the work we’re supporting, whether that’s seasonal malaria chemoprevention campaigns or net durability surveys that lost funding, wouldn’t have funding replaced by another funder, and our networks have been an important source of information about this.

In terms of adaptation, we have had to look into areas where we have a less substantial history of grantmaking. For example, we’ve done a lot of work on insecticide-treated net campaigns to prevent malaria, but we have provided less funding in the past for malaria treatment supplies, where we found a number of urgent time-sensitive funding gaps, so we had to learn quickly there.

We have expertise—and our grantees have even more—that was really useful for allowing us to act quickly but carefully this year. We are very used to making hard choices between options when there are too many gaps to fill. We are now trying to build that understanding across a range of areas that may now need funding, due either to the funding cuts or to the emergence of new technologies, such as spatial emanators.

The role of cost-effectiveness

What are the biggest questions for cost-effectiveness caused by this rapidly changing environment?

Rosie Bettle: One key question is understanding how the funding landscape is changing. Before, it felt like we had a decent sense of what other funders were going to support and therefore where our funding could be most additive. Now, we don’t know exactly where the gaps are going to be. We’re trying to make educated estimates based on our experience in these sectors and on what we’re hearing from other stakeholders, but there’s still a lot of uncertainty. The changing funding landscape also gives us uncertainty around whether the “systems” that our grantees rely upon will still be functional. Global health programs are highly interconnected and rely on a range of health systems to carry out their work, and we’re uncertain about the impact of cuts on those systems. For example, we might be funding an implementer to distribute ready-to-use therapeutic food (RUTF) to treat malnutrition in Cameroon, but that distribution depends on a functional supply chain bringing RUTF into Cameroon. Some of those systems are USAID funded, so we have a lot of questions about whether they will continue functioning and how we ought to respond to keep the most vital platforms going.

Additionally, we’ve seen an increase in time-sensitive grants, so we’re working to estimate the cost-effectiveness of a given grant quickly while maintaining research rigor. For some grants to fill gaps in malaria treatment supplies, we needed to make a decision within one week, much faster than our usual months-long process, which has led us to explore new methods. For example, in some cases we are using multiple simple models to triangulate a program’s cost-effectiveness rather than one complex model. We’ve also built out our network so we can more quickly check in with experts, program implementers, and other stakeholders.

How is GiveWell balancing and addressing cost-effectiveness and immediate needs with long-term sustainability and systems change?

Meika Ball: While we have experienced some tensions in our work over the past year, they haven’t necessarily been between cost-effectiveness and systems change—for example, we believe there may be cost-effective opportunities to improve health systems. I think one important tension we have seen is around the level of uncertainty we are comfortable accepting in the grants we fund.

Historically, evaluating health systems strengthening work has been difficult for GiveWell because there is more limited evidence on health outcomes than for many other areas we support. It’s unlikely we’ll be as confident about the impact of health systems strengthening interventions as we are about, for example, insecticide-treated nets, just because the evidence base is less robust—it’s hard to run a randomized trial on national-level health systems interventions!

A key challenge for GiveWell when evaluating these grants is estimating the likelihood that a program’s impact will be sustained in the long term. However, if these programs can be sustained over the long term, or lead to enduring systems change (and the benefits therefore happen over a long period of time), we think they could plausibly be very cost-effective. This is one of the reasons we have focused on learning more about health systems strengthening this year.

Given our uncertainties, we are trying to think about how to embed learning into grants wherever possible, rather than treating research and programmatic work as separate. We’re looking for opportunities where we can fund implementation and then bolster program monitoring or add additional evaluations. We hope this will enable us to respond to time-sensitive needs while also generating the evidence that will help us and others make better decisions going forward. We are also thinking more about whether we can learn about the ongoing benefits of some programs after the grant period ends. One example of this would be funding data collection to assess coverage after a technical assistance program has concluded.

Overall, we expect to make health systems strengthening grants we think are highly cost-effective, but these will inherently be less certain than grants we make to other programs that have a stronger evidence base and are easier to measure. Our goal is to build in sufficient opportunities to learn more so we can update our understanding over time and have a more informed view of the long-term sustainability of some of the programs we support.

Future GiveWell grantmaking

How do you think GiveWell’s grantmaking pace might change in 2026?

Dilhan Perera: I’ll focus on grantmaking in one particular area: HIV prevention and treatment. When the US aid freeze hit earlier this year, there were a number of previously well-funded areas, including HIV, that looked like they might have large funding gaps. Because they had previously been well-funded, GiveWell hadn’t spent much time evaluating them. So in addition to the challenge of learning what funding had been cut and how those cuts were impacting health service delivery, we also had a lot to learn about these cause areas and how cost-effective interventions within them might be (including how cost-effectiveness has changed since we last briefly considered them). We devoted substantial time to this effort in 2025.

Because we’ve now done a lot more of this background work, we are better positioned to evaluate the cost-effectiveness of specific funding opportunities in 2026. For example, we’ve been updating our cost-effectiveness models for both HIV treatment and HIV prevention interventions, so we’ll be able to assess whether specific grants focused on these interventions look cost-effective enough for us to recommend funding. I expect this will also allow us to increase the pace of our grantmaking—or at least the pace of our decision making, which includes decisions to decline funding.

Aid cuts have created more uncertainty than before. What strategies are you using to navigate this, and how is this impacting your grantmaking process?

Alex Bowles: We’re taking a number of steps to deal with the substantial uncertainty that continues to affect health and development programs. First, we’re leaning into our networks. We’d like to ensure that where there’s good information, we’re aware of it. This means making sure we are connecting with the right people, attending the right meetings, and even subscribing to the right mailing lists!

Second, we’re becoming more tolerant of unavoidable uncertainty. In some cases, there will be uncertainties that aren’t currently resolvable—like how much the US government will spend on malaria programs in 2027. Sometimes these will affect our cost-effectiveness estimates, but we can’t let these wider ranges of uncertainty prevent us from making good grants.

Third, we need to be thinking hard about the biggest uncertainties in a grant investigation—they might be different than those we had before the cuts. In many cases we’re placing relatively more importance on addressing big-picture qualitative questions—for example, “Is this country likely to dramatically change its malaria strategy because of cuts?”—and less on the details of individual parameters in our cost-effectiveness analyses, ones that are unlikely to sway our decisions and where we generally already have solid research.

And finally, uncertainty is pushing us to continue our efforts to learn from our implementation grants rather than focusing all our learning before a grant decision. This could be as intensive as adding a randomized evaluation of a program component or as light-touch as making clear and verifiable predictions about what will happen so we can go back later to see if we were right.