Get Your Weekly Digest

Receive the best podcast summaries and insights delivered to your inbox every week.

We respect your privacy. Unsubscribe anytime.

Planet Money

Saving lives with fewer dollars

with Madeleine Tronceau, Joel Kambale-Kamete, Taryn Maddox, Rosie Bettle, Susan Shepard, Alice Redfern

Published November 27, 2025
View Show Notes

About This Episode

The episode follows ALIMA's healthcare program in conflict-affected northern Cameroon after it abruptly loses $1.9 million in funding from USAID, forcing cuts to lifesaving services. It contrasts ALIMA's on-the-ground medical work with GiveWell's data-driven philanthropy, showing how GiveWell evaluates whether to fund the program despite limited and imperfect data. Ultimately, GiveWell decides to fully replace the lost USAID funding for one year, while the story highlights broader global cuts to aid and the resulting loss of both services and information about needs.

Topics Covered

Nonprofit organizations Healthcare system Effective altruism International aid cuts Philanthropic funding Child mortality Malnutrition Conflict zones Randomized controlled trials Data-driven decision-making

Disclaimer: We provide independent summaries of podcasts and are not affiliated with or endorsed in any way by any podcast or creator. All podcast names and content are the property of their respective owners. The views and opinions expressed within the podcasts belong solely to the original hosts and guests and do not reflect the views or positions of Summapod.

Quick Takeaways

  • ALIMA's medical program in northern Cameroon lost its $1.9 million USAID grant, forcing immediate cuts to clinics and services in a conflict zone with high child mortality and malnutrition.
  • GiveWell, a philanthropy focused on maximizing lives saved per dollar, created rapid-response models to decide whether funding ALIMA would be more cost-effective than alternative programs or even direct cash transfers.
  • Severe data limitations, security constraints, and differing measurement standards in humanitarian contexts made it difficult for GiveWell to apply its usual evidence thresholds.
  • A key turning point came when an independent hospital director in Cameroon described ALIMA's potential withdrawal as a "catastrophe," reinforcing the plausibility of ALIMA's impact data.
  • GiveWell ultimately approved a $1.9 million grant to fully fill the USAID gap for one year, using the opportunity to learn how to make high-impact decisions with more uncertainty than usual.
  • Global cuts to international aid by multiple governments are shrinking funding for programs like ALIMA's and also eroding the information flow about needs on the ground.
  • As aid programs shut down, communities lose access to care and donors lose the data needed to recognize and respond to crises.
  • Taryn from GiveWell describes their framework as a way to "discipline compassion," accepting that limited funds mean prioritizing interventions that save the most lives per dollar even when that feels emotionally difficult.

Podcast Notes

Introduction and overview of two life-saving approaches

Contrasting hands-on medical work and desk-based number crunching

Description of frontline healthcare in northern Cameroon[0:35]
• Doctors and nurses provide basic healthcare and medical supplies in the far north region of Cameroon
• They give vaccines, monitor pregnancies, and train patients to identify malnutrition
• A simple color-coded measuring tape (red, yellow, green) is used by mothers to check their children's mid-upper arm circumference for signs of malnutrition
ALIMA's role and scale of work[1:14]
• Madeleine Tronceau manages grants for ALIMA (Alliance for International Medical Action)
• ALIMA treated almost 400,000 people in Cameroon last year
• They have stayed operational during years of armed conflict by staying out of the fray, building trust, and managing difficult logistics

Funding crisis triggered by USAID cuts

Loss of USAID funding for ALIMA's Cameroon program[1:28]
• ALIMA's Cameroon program was supposed to receive $1.9 million from USAID this year
• The Trump administration announced it was gutting USAID, suddenly removing this funding
• Madeleine had just returned from visiting ALIMA's doctors and nurses in the mountains around Mokolo when she learned of the cuts
Immediate human implications of the funding loss[1:49]
• Madeleine recalls malnourished children in hospital beds whose treatment would stop if funding ended
• She feared having to discharge children who would not get treated elsewhere
• When asked who would take over, she answers that the local health system is not able to absorb those patients, meaning they would not get care

Global scale of the USAID funding hole

Magnitude of lost funding and estimated deaths[2:41]
• ALIMA's Cameroon project is one of thousands of programs worldwide that suddenly stopped having money
• Tens of billions of dollars in aid funding disappeared with the USAID cuts
• An estimated 620,000 people have already died for lack of care due to these funding losses

Introduction of GiveWell and their decision problem

Framing the parallel triage done by philanthropists

Philanthropic group working with numbers[3:02]
• A philanthropic group with money is trying to figure out if and how they can help fill the gap left by USAID
• They use a very particular, math-heavy way of making decisions under time pressure
Planet Money episode setup[3:20]
• Host Mary Childs introduces the episode as a behind-the-scenes look at one organization trying to fill a tiny part of the funding hole left by USAID
• She notes that their calculations about deaths averted are ruthless, citing a figure of about 650 deaths averted in total in one calculation
• The central decision they wrestle with is whether to give money to the ALIMA project in Cameroon

How GiveWell becomes involved after USAID is cut

Compilation of affected projects and GiveWell's mandate

List of defunded projects reaches GiveWell[5:58]
• Recently unemployed USAID workers compiled a list of all projects that had just lost funding, from water and sanitation in Yemen to research on bio-fortified maize in Guatemala
• That list landed on the desk of GiveWell, a philanthropic group
GiveWell's explicit goal and available budget[6:16]
• GiveWell donates hundreds of millions of dollars each year
• Their explicit goal is to make their money save or improve the most lives per dollar based on evidence and calculations
• GiveWell created a rapid response team and expected to spend about $50 million to fill a small part of the USAID gap

Rapid response team and focus on Cameroon

Rosie's role on the rapid response team[6:50]
• Researcher Rosie Bettle describes herself as hoovering up opportunities that fall between the cracks
Selection of Cameroon as a candidate project[7:35]
• Rosie and the team begin zeroing in on a few specific projects from the list, including ALIMA's work in the mountains of Cameroon
• GiveWell must determine if the Cameroon project is a better use of limited funds than other options they could fund right now
• Planet Money asked GiveWell to record their internal discussions about evaluating this project

GiveWell's methods, values, and initial information-gathering

First internal GiveWell meeting about Cameroon

Taryn frames the challenge[7:28]
• Director of research Taryn Maddox notes a steep learning curve because Cameroon's context is terrain they do not know well
• GiveWell is mapping out what they need to know to decide whether to fund the grant
Reasons for prioritizing Cameroon[7:54]
• Taryn asks Rosie why they are prioritizing this opportunity among others
• Rosie explains the region is heavily affected by conflict and the populations are at higher risk with higher mortality

GiveWell's usual niche and canonical example

Difference between USAID-style programs and GiveWell's traditional focus[8:08]
• USAID had historically funded big, multi-pronged programs in areas with conflict, poverty, malnutrition, and disease, as well as programs promoting global stability and trade
• By contrast, GiveWell has usually looked for specific, efficient, and relatively neglected projects they can evaluate rigorously
Mosquito nets as a canonical cost-effective intervention[8:35]
• Buying mosquito nets to prevent malaria is cited as the canonical example: the nets are super effective and super cheap
• GiveWell likes this type of intervention because impact can be strongly supported by data and randomized controlled trials

Concept of life saved equivalent and effective altruism context

Definition of life saved equivalent[10:03]
• GiveWell aggregates benefits such as averting disability, improving income, and improving cognitive outcomes into a single measure called an equivalent life saved
• They assign numerical values to different benefits so they can compare interventions on one scale
Philosophical and methodological background[10:17]
• This style of calculation is characteristic of a small group of nonprofits and philanthropic organizations associated with effective altruism
• Effective altruism is described as trying to do the most good based on evidence
• The approach emerged with increased data collection and randomized controlled trials that transformed development economics
• GiveWell was founded in 2007 in this context and promises donors that it will save or improve the most lives per dollar

Direct engagement between GiveWell and ALIMA's Cameroon team

First Zoom call: understanding current operations and demographics

Status of ALIMA's work after USAID cuts[11:37]
• GiveWell asks if everything is still fully operational in Cameroon
• Joel Kambale-Kamete, in charge of the program, replies that the program is not running at 100%
• He explains that in Makari, ALIMA went from operating in 14 health centers to fewer; in Mokolo, from eight facilities down to just one hospital
• Joel says if no other funding comes, they will have to shut down the project
Population and demographic data[12:35]
• Rosie asks Madeleine for clarification on the mix of internally displaced people and host communities
• Madeleine estimates the Mokolo health district population at around 350,000 people

GiveWell's internal mortality and impact calculations

Child mortality mechanisms and double-counting concerns[13:27]
• Rosie notes that when children die because of malnutrition, it's usually due to complications like pneumonia or diarrhea that their weakened bodies cannot handle
• She worries they might be double-counting deaths in their models if they are not careful with how mortality data is categorized
Challenges of missing and unreliable data in a conflict zone[13:55]
• Patients are often displaced people or refugees who are mobile and may not return for follow-up visits
• GiveWell needs information such as under-five mortality rates that effectively does not exist in robust form for this context
• They struggle to interpret a data point of 730 total deaths per year and whether 650 deaths averted is calculated with or without ALIMA's program

Balancing spreadsheets with on-the-ground understanding

Need for on-the-ground research and safety constraints

Importance of ground truthing alongside models[15:03]
• Rosie later reflects that relying only on spreadsheets and models without understanding what is happening on the ground can lead them astray
• GiveWell would typically hire a research firm to verify numbers like population and staffing, but they cannot do that here because the area is unstable and two researchers were killed there a few years ago
Debate over commissioning new studies[15:51]
• GiveWell and ALIMA discuss whether to commission a study and whether it would be safe to run surveys in the hardest-to-reach areas
• ALIMA notes the hardest-to-reach areas have the worst figures, but a comprehensive study would take about a year while needs are urgent now
• GiveWell worries that without time for proper research and modeling, they cannot be sure they are backing the right project

Taryn's example of counterintuitive cost-effectiveness

Comparing maternal mortality intervention vs syphilis testing add-on[16:33]
• Taryn recalls evaluating an intervention aimed at reducing maternal mortality and another that helped a government switch from an HIV-only test to one that also tests for syphilis in pregnant women
• When they did the math, maternal mortality was extremely rare while syphilis was less rare, and its effects were lifelong and severe
• The syphilis test intervention turned out to be about a thousand times more cost-effective than the maternal mortality program
Emotional tension and disciplined compassion[17:32]
• Taryn acknowledges that strong intuitions can be wrong, and that a framework for disciplined calculations helps guide compassion
• She admits that they do think about the mothers who are not saved when funding the higher-yield syphilis intervention
• She concludes that with a limited pot of money, they must use it in the best way possible by their metrics

Methodological clash: humanitarian metrics vs GiveWell metrics

Discussion of mortality measurement standards

Two children per 10,000 metric and differing conventions[18:27]
• GiveWell brings up a statistic of two children per 10,000 mortality and notes it is becoming a key point for them
• Susan Shepard from ALIMA asks if this is GiveWell's first project evaluation in a humanitarian context
• She explains humanitarian settings commonly measure deaths per 10,000 people per day, whereas demographic and health surveys measure child mortality by live births per year
• Susan says she has never found a way to join those two measurement systems and simply treats them separately

Humanitarian priorities: mortality reduction and dignity

Why focus on conflict areas instead of just high mortality regions[19:46]
• Taryn asks why service provision is targeted at places with conflict rather than simply areas with high mortality
• Susan explains that humanitarian aid is aimed at people forced to flee or move due to conflict, and includes maintaining some level of dignity through services, not just reducing mortality

GiveWell's modeling, comparison to cash transfers, and multi-pronged complexity

Back-of-the-envelope calculations (BOTEX) and cash comparison

Three models and benchmark against cash transfers[20:41]
• On April 30, Rosie presents three BOTEX models to estimate how ALIMA's work affects mortality
• GiveWell asks whether the Cameroon programs perform better than simply giving people cash
• Rosie reports that all three models indicate the Cameroon program does better than cash transfers by a fair amount
• Taryn notes this is somewhat surprising given the very different approaches
• However, GiveWell usually prefers programs that beat cash transfers by an even larger margin than Rosie's current estimates show

Difficulty of modeling a multi-component health program

Rosie's "problem child" model[21:30]
• Rosie describes one of her BOTEX models as a problem child because it is hard to calculate effectiveness across many different interventions combined
• The Cameroon program includes prenatal care, pediatric medicine, vaccines, water sanitation, and treatment for malnutrition
Indirect and hard-to-quantify benefits[21:45]
• Mothers often bring children for nutrient-dense peanut paste for malnutrition and, while there, receive vaccines, malaria treatment, and training on malnutrition detection
• Rosie highlights downstream effects such as ALIMA staff educating and training other medical staff and community health workers
• She notes uncertainty about what happens in three or four years if doctors capable of training others leave and do not pass on their knowledge
• Taryn acknowledges that these indirect effects plausibly contribute to child health and survival, even if they are difficult to quantify

Ongoing data requests, uncertainty, and probability estimates

Cameroon team's perspective on GiveWell's data demands

ALIMA's reaction to extensive questioning[22:50]
• Over several weeks into May, GiveWell continues asking for more data and clarifications, including additional mortality data
• Madeleine later says ALIMA does not mind the many questions; they see it as a sign of donor interest and understanding of urgency

May 21 Zoom: timelines, mortality data, and probability of funding

Challenges obtaining mortality data[23:30]
• Rosie again asks if there is more mortality data from Cameroon they can check
• Madeleine explains that updated data is difficult to obtain; the most recent survey is old, and their internal data is informal and not a proper assessment
Deadlines and planning needs[23:58]
• ALIMA asks about GiveWell's deadlines and whether they can get an answer or budget indication by the end of the month
• Madeleine says that the longer they wait, the worse it will be for patients
Rosie's probability estimate for the grant[24:13]
• Madeleine asks what the probability is that they will receive the money
• Rosie estimates, with caveats, that she is about 55% to 60% sure the grant will come through
• She clarifies they are still considering the grant but are not certain yet
• ALIMA explains they need to know which grants they can rely on for the coming months and next year

Worsening conditions in Cameroon as decisions drag on

Timing and onset of rainy season[25:19]
• By May 21, it has been more than three months since the USAID shutdown news
• The Cameroon project has already pulled staff from multiple health centers
• The rainy season has begun, bringing increased mosquitoes and malaria risk

Consequences of service cuts and ALIMA's persistence

Services cut due to funding loss

Programs already reduced or eliminated[27:38]
• With uncertainty about funding, ALIMA cut mental health programs and educational programming on nutrition, hygiene, and information about free project services
• Health facilities are open fewer hours, which discourages people from attempting to seek care
Impact of closures on patient behavior[27:53]
• Joel notes that if people travel to the hospital and find it closed, they give up
• He says a woman about to give birth might deliver in the bush without medical care if the hospital is closed, and such experiences make her less likely to try again

Madeleine's refusal to lose hope despite systemic trends

Maintaining hope under deteriorating conditions[28:21]
• Asked if she ever loses hope, Madeleine says they do not, because otherwise they would stop working
• She says the situation is so dire in the field that she cannot accept losing hope
Broader decline in global aid budgets[29:17]
• Madeleine notes that countries around the world-including the US, France, Germany, and the UK-are reducing the money they allocate to international aid
• She characterizes aid as less and less everywhere
• Reasons mentioned include tighter budgets, austerity politics, and increased spending on defense
• These trends make the Cameroon project one of many organizations now competing for limited philanthropic funding

GiveWell's remaining concerns and attempt at independent verification

Two main hang-ups: overestimation risk and lack of independent data

Risk of overestimating impact[29:45]
• GiveWell worries that, given imperfect data, they may be overestimating lives saved or improved by the Cameroon program
Need for independent on-the-ground confirmation[29:54]
• They are also concerned about how to be confident in ALIMA's data without independent on-the-ground confirmation
• Researcher Alice Redfern repeatedly pushes on the need for some source of information not directly from ALIMA
• Alice expresses discomfort with taking so much on face value when they have reservations and no external checks
• She commits to trying to source a reliable contact in the area

Phone call with hospital director as ground-truthing

Conversation with hospital director in French[32:08]
• On May 30, Alice speaks in French with the director of one of the hospitals staffed by ALIMA's Cameroon project
• Rosie describes this as the closest they can get at that moment to ground-truthing
• Alice cautions that it is just one person's opinion and they should not over-update, but notes he was blunter than expected
Hospital director's assessment of ALIMA's importance[32:08]
• The hospital director uses the word "catastrophe" to describe what would happen if ALIMA pulled out
• He says ALIMA is the only driving force bringing people into the hospital
• Without ALIMA, many patients would not get any care at all
• He explains that without ALIMA there is not much to offer at the hospital, and that ALIMA's nutrition services are a major draw for mothers
• He tells Alice that if the nutrition services are removed, mothers will not come for antenatal care

Effect of ground-truthing on GiveWell's confidence

Alice's updated view and remaining unease[33:18]
• Alice remains uncomfortable with the level of uncertainty in the calculations
• However, the hospital director's blunt assessment makes her more willing to believe some of the numbers she was skeptical about before
• She and Taryn reflect that they have learned a lot in recent weeks and hope to close the analysis

Taryn's mindset shift about uncertainty and impact

From "we may not fund this" to "this might be high impact despite uncertainty"[32:36]
• Taryn recalls a moment when she thought they might not be able to fund the project because they would not get good data
• She then reasons that ALIMA has already done significant work to gain access to these difficult places, which suggests high potential impact
• She shifts to accepting that they may not know everything this time but can use the experience to learn and save more lives in the future
• She observes that saving the most lives is much harder without local governments providing basic services and without international governments contributing to stability and safety

Final decision, grant approval, and remaining gaps

GiveWell's approval of the $1.9 million grant

Decision moment and grant details[33:40]
• Over a weekend, Taryn reviews all information, and on Tuesday, June 3rd, she approves the grant
• GiveWell grants $1.9 million to ALIMA's work in the far north of Cameroon, fully replacing the USAID funding for one year
• The two groups-number-based GiveWell and hands-on ALIMA-find a way to work together
Restarting paused services and ongoing worries[34:30]
• GiveWell's grant allows ALIMA workers in Cameroon to restart activities they had paused
• Madeleine says she is really happy and that they succeeded in proving the proposed intervention was cost-effective
• Joel is happy too but is already thinking about how to maintain the projects next year
• He says ALIMA will continue to seek other private actors to sustain the projects

Scale of unmet needs and information loss

Small share of defunded programs that received GiveWell money[34:40]
• GiveWell started with a list of 140 programs that lost funding
• As of the episode's reporting, GiveWell has given money to 23 of those programs
• They have disbursed $39 million to address a funding hole of tens of billions left by USAID
• Many other aid programs in numerous countries are shutting down or reducing services
Loss of data and visibility when aid is cut[35:19]
• Madeleine warns that as aid programs shut down, information from those areas is lost
• Without aid agencies gathering data, it can appear as though there are fewer needs simply because no one is documenting them
• She says the risk is that with less aid, it looks like there are fewer needs, which is not the case
• She states that around the world there is more need, but we are increasingly not going to know about it

Lessons Learned

Actionable insights and wisdom you can apply to your business, career, and personal life.

1

When resources are limited, using clear quantitative frameworks to compare interventions can reveal counterintuitive choices that save far more lives per dollar than our intuitions would suggest.

Reflection Questions:

  • • Where in your own work or life are you relying mostly on intuition instead of comparing options with concrete numbers or evidence?
  • • How might your current priorities change if you systematically estimated the impact or return on different choices before committing?
  • • What is one decision this month where you could explicitly model the potential outcomes and costs instead of going with your first instinct?
2

High-impact decisions often require acting under uncertainty, striking a balance between waiting for perfect data and moving fast enough to prevent real harm.

Reflection Questions:

  • • In what situations do you tend to delay action because you are waiting for more information that may never fully arrive?
  • • How could you define a threshold of "good enough" information that would allow you to move forward without being reckless?
  • • What important decision are you currently postponing where a partial but timely action could still meaningfully reduce risk or harm?
3

Ground-level perspectives and qualitative evidence can be essential complements to models and spreadsheets, especially in complex or hard-to-measure environments.

Reflection Questions:

  • • Which stakeholders or front-line voices are missing from your current understanding of a problem you are trying to solve?
  • • How might a direct conversation with someone on the ground alter the way you interpret your data or metrics?
  • • What is one concrete step you can take this week to gather first-hand input from people directly affected by a decision you control?
4

Building access and trust in challenging environments is itself a form of capital that can make certain organizations uniquely positioned to create impact, even when formal data is sparse.

Reflection Questions:

  • • Where have you or your organization already invested in relationships or local knowledge that others lack but you may be undervaluing?
  • • How could you factor trust and access into how you assess opportunities, partners, or projects-not just hard numbers?
  • • What is one relationship or context where deepening trust could significantly increase your ability to do meaningful work over the next year?
5

Systemic funding shifts can silently erode not only services but also the information needed to recognize and respond to emerging needs, making vigilance and transparency even more important.

Reflection Questions:

  • • How aware are you of the upstream budget or policy changes that could affect the data and signals you rely on in your field?
  • • In what ways could you help maintain or share information about needs or risks, even when formal structures are being cut back?
  • • What is one channel-reports, conversations, partnerships-you could strengthen to ensure that critical problems in your domain remain visible rather than disappearing from view?

Episode Summary - Notes by Kai

Saving lives with fewer dollars
0:00 0:00