Overview: Saving Lives (Focus on Africa) | GiveWell

You are here

Overview: Saving Lives (Focus on Africa)

A note on this page's publication date

The content on this page has not been recently updated. This content is likely to be no longer fully accurate, both with respect to the research it presents and with respect to what it implies about our views and positions.


We examined 59 organizations' evidence that they can reliably and cost-effectively save human lives (or prevent extreme debilitation such as blindness and severe deformities) in Africa; below we recommend the 3 that best demonstrated this quality. A total of 107 organizations were asked to apply, and 59 submitted applications, including many household-name charities including the American Red Cross, UNICEF, and World Vision. We start with our recommendations, followed by a brief summary of our reasoning, followed by a more detailed description of our reasoning.

Recommendations

1. Population Services International (PSI) markets and distributes life-saving materials (mostly condoms and insecticide treated bednets). Across the entire organization, we believe it can generally be expected to save lives for around $650-1000 each, while also conferring other benefits such as reducing unwanted pregnancies and slowing the spread of infectious diseases.

The connection between PSI's activities and lives saved is not as tangible and direct as for some of our other applicants, but its activities nonetheless appear to be extremely useful, cost-effective approaches to saving lives. PSI simultaneously runs programs that appear as cost-effective and well-monitored as anyone else's, and impresses us most as an organization - in terms of its rigorous self-documentation, commitment to transparency, and coherent overall strategy.


2. Partners in Health (PIH) creates comprehensive health programs (hospital, health center, and community health workers) in disadvantaged, generally rural, areas in the developing world. We have relatively little ability to quantify its impact on life outcomes, but our best estimates suggest it may be saving lives for somewhere around $3500 each, not including many general-health benefits it provides beyond saving lives.

We believe that some of our other finalists can make a donation go farther strictly in terms of extreme life change, but PIH provides significant benefits beyond its effect on mortality, and we also see it as the "lowest-risk" charity available: its model is extremely logical and tangible, and we have high confidence in it.


3. Interplast is devoted to correcting deformities requiring surgery (such as cleft lip and palate deformities), both through direct treatment and through support of local doctors. Across the organization, it spends around $500-1500 per corrective surgery.

We don't have a strong sense of how many of these surgeries are truly life-changing, as opposed to minor or partial corrections; we also hesitate somewhat to recommend an organization so narrow in its scope, working with people with such a variety of problems. However, Interplast's work is relatively cost-effective and extremely tangible; it's a good option for a donor uncomfortable with the uncertainty and range of outcomes associated with more complex organizations.


Reasoning in brief

We invited 107 organizations to apply, including nearly all household-name charities; 59 submitted applications. We chose our 12 finalists using a Round 1 application that asked them to focus on a single project, and give a concrete sense of (a) whether it has successfully changed lives; (b) how cost-effectively it has done so. (Details of our Round 1 process are here.) We then examined these 12 in far more depth. The following summarizes the strongest programs we've seen (an organization typically runs several programs).

Cost per significant life change: strongest programs
Program Organization Reliability/tangibility Cost per significant life change Type of life change Other benefits
Child survival (CSHGP) programs Multiple Moderate ~$1000 Life saved from infant/maternal mortality Improves nutrition; reduces diarrhea
Bednet marketing/distribution PSI Moderate $600-2400 Life saved from malaria Reduces malaria
Condom marketing/distribution PSI Moderate $150-2000 HIV infection prevented Slows spread of HIV; reduces other diseases; reduces unwanted pregnancies
Comprehensive health program PIH High ~$3500 Life saved Improves general health in many ways
Corrective surgery Interplast High $500-1500 Corrected deformity (may not always be major)
Vitamin A supplementation program HKI Moderate ? Reduced overall under-5 mortality Improves nutrition

Note that all dollar estimates are extremely rough; see individual organizations' pages for more details. Also note that "Reliability/tangibility" refers to how straightforward, direct, and generally "reliable" a program is: PIH and Interplast treat patients directly, while other programs rely more on mass distribution/education (or we have large questions about them that lower our confidence).

On balance, we find the first three options listed to be the strongest: all save lives for something around $1000 each, all have other major benefits to the population served, and all have some track record. However, we strongly believe in funding organizations, not programs, and that tips the balance away from organizations running CSHGP programs (we have little information on, or confidence in, these organizations' other activities) and toward PSI, PIH, and Interplast, organizations we can understand and have confidence in as a whole.

PSI has the most strategic, systematic, well-documented approach of any of our finalists, and its programs are among the most cost-effective. We recommend it above the others. For a donor seeking a more straightforward, tangible program, we recommend PIH or Interplast, both of which are working directly with patients rather than doing mass marketing and promotion of life-saving materials. Deciding between PIH and Interplast is a particularly tough judgment call: Interplast likely causes more "significant life changes" for the same funds, strictly speaking, but PIH likely has much larger non-mortality-related benefits.


More details

Goal of the cause

The official goal of this cause is to help people in Africa avoid death and extreme debilitation. There are many ways to do this, because there is a great variety of health problems in Africa that significantly damage people's lives, yet can be addressed fairly cheaply and simply.

Our problems and solutions overview gives a summary of all the problems we know of that fit this description. Our life expectancy page examines the difference between high-income countries and sub-Saharan Africa, in terms of when and why people die.

We focus, conceptually, on finding how to deeply change as many lives as possible per dollar; in this context, a "life changed" includes:

  • Prevention of a death (from malaria, tuberculosis, diarrhea, maternal complications, or other preventable cause of death - see our problems and solutions overview for details).
  • Prevention or correction of permanent blindness.
  • Correction of a permanent, extremely debilitating deformity, such as a cleft lip or palate.
  • Provision of antiretroviral therapy (ART), allowing an HIV-infected person to have a normal quality of life. Note that in this case, we think of a "life changed" as a person put on ART for at least 5-10 years (not just a single year).

There are many reasons these different outcomes can't be strictly compared, including:

  • Blindness, deformity and death are all very different things.
  • An AIDS infection prevented has the added benefit of slowing the spread of the disease as a whole; this is less true of a malaria case treated or a deformity corrected.
  • There will often be large differences in organizations' effects on less deep life outcomes - for example, Interplast has no direct effect on anyone except those who have deformities corrected, whereas Partners in Health helps with a large number of non-fatal health issues (so comparing them directly in terms of "lives saved" or even our expanded "lives changed" will overstate Interplast's cost-effectiveness relative to Partners in Health's).
  • Permanently protecting someone from a disease (either through a vaccine or by changing their behavior for the long term) is more valuable, in itself, than treating/preventing one instance of a disease.
  • As our life expectancy page shows, protecting an infant from a single cause of death leaves them at risk from many other causes of death; by contrast, an older person saved from death has a strong chance of living to age 60 or so.

For these reasons among others, we feel it would be a mistake to put too much stock in any "lives changed per dollar" number, but we still estimate this number where we can, specifically to see if there are cost-effectiveness differences that are large enough to overcome our other principles (outlined below).

Note that we have made an explicit decision not to make heavy use of the World Health Organization's Disability Adjusted Life-Year (DALY) metric. Our writeup of this decision is forthcoming; an informal exchange on the topic, on our blog, is available here.

Principles

Understand the entire organization.

The organizations in this cause tend to be extremely large, international organizations that execute many programs in many countries. Because of this, it is both difficult and necessary to get as complete a picture as possible of what they do. For an organization working at a single site in New York City (as some of our applicants in other causes do), we can use the quality of one program as a rough proxy for the quality of the organization as a whole; by contrast, for Africa-related charities, each project can represent a different strategy, staff, and country - in which case understanding one project in one area, without understanding the overall organizational strategy behind it, does relatively little for a donor.

In evaluating our Round 1 applicants, we looked for the organizations that seemed most likely to fit this criterion. That meant selecting organizations that either (a) follow relatively unified organization-wide strategies (such as Interplast, the HealthStore Foundation, and GNNTDC); or (b) demonstrated extremely thorough self-documentation, -monitoring, and -evaluation, to the point where we hoped they would be able to present a systematic view of their overall activities. See our writeup on Round 1 applicants for more detail.

Simple organizations

Partners in Health, Interplast, GNNTDC, and The HealthStore Foundation are all organizations focusing (with at least 50% of their budget) on a particular strategy.

Complex organizations

Our other finalists are extremely large organizations with varied activities. For each of these, we sent a Round 2 application that aimed to capture everything the organization does in every region (see the individual organizations' pages for this application, as it varies slightly by organization). We then spoke on the phone with each of these finalists, explaining that we seek an organization-wide view of activities, and specifying that this view need not match the format of our application - we asked each simply to send us the materials that could best give us a picture of the organization.

Food for the Hungry filled out our Round 2 application as requested, but we were still unable to use their submission to understand both the strategies and impacts of the organization. The American Red Cross and The International Eye Foundation chose not to continue with our process after our Round 2 conversations, citing time and budget constraints on assembling information.

Helen Keller International, UNICEF, the Aga Khan Foundation, and Project HOPE all declined to fill out our Round 2 Application, but agreed to send us materials to give us some sense of their overall organization's activities. In these four cases, we ultimately determined that we could not use these submissions to understand both the strategies and impacts of the organizations. See individual charities' pages for details.

Population Services International filled out our Round 2 application and sent us a "dashboard" that gave us an excellent overview of its activities across the organization (see the PSI page for details) - and it sent this document the same day as our phone call.

It appears that PSI's "dashboard" was already maintained and required no special effort to put together for us. That is what we seek: though others may disagree, we believe that an extremely large and complex organization should constantly maintain its own bird's-eye view of activities, and that showing a potential funder what it does - like showing its Executive Director or Board - should not require significant special effort.

Monitoring and evaluation are essential.

Many of our applicants - particularly those that did not advance past Round 1 - described projects in which materials (medical supplies, protective bednets, etc.) were distributed, or education classes were conducted, but did not provide evidence on (or appear to evaluate) the impact of these campaigns on the people served. (The former - measuring whether activities were conducted as intended - is often referred to as "monitoring," while the second is often referred to as "evaluation.")

Reporting how many nets were distributed is not the same as reporting how many nets were used; reporting how many people took a class is not the same as reporting how many changed their behavior; reporting that a well was constructed is not the same as reporting on the change in water quality and disease prevalence. Though the second type of outcome is much harder to measure, we believe that it is essential that an organization do so, at least for representative projects (if not for all projects). Otherwise, we suspect that an organization trying to help people thousands of miles away from its headquarters (and in a different culture) runs too great a risk of missing a key factor and providing no benefit at all (for an accessible book full of examples of this phenomenon, we recommend White Man's Burden by William Easterly).

As such, we use a relatively high burden of proof in assessing whether an organization has demonstrated its impact on people's lives. Our burden of proof is particularly high for programs that rely on changing behavior, such as programs that distribute materials people must use themselves (bednets, condoms) or programs focusing on local education.

Assessing an entire organization on this principle requires a bird's-eye view of its activities. The only organizations that we are confident have consistent monitoring and evaluation of their activities are PSI and Interplast, although PIH's model is straightforward enough that we don't feel as much need for strong documentation of outcomes.

Consider cost-effectiveness, but don't expect precision.

Considering how simple and inexpensive many methods of saving lives are, we feel it would be misguided simply to fund the most tangible, reliable approach. As our cost-effectiveness table above shows, we believe that approaches that are more targeted to those in need, and more aimed at the most cost-effective interventions, can help several times as many people for the same amount of money.

This is why we calculate the "lives changed per dollar" metric wherever it is feasible to do so. In all cases, our estimates of this figure involve many assumptions and extremely high degrees of uncertainty; also, since the figure looks only at lives saved or significantly changed, it leaves out other benefits such as better nutrition or fewer unwanted pregnancies. For this reason, we do not use our cost-effectiveness figures in any precise way; we consider them only when they point to large differences between strategies.

Aim to give where further funding is needed and can be productively used.

On one hand, we are against the practice of restricting grants; we see this as micromanagement, and believe that a good charity should know more about how to use its funds than we do. But when choosing between charities, we have tried to consider where our money is needed and can be productively used, a difficult endeavor when dealing with large and complex organizations.

Our three recommended organizations all seem to us to be capable of scaling up their operations with more funding. PIH has recently replicated its rural health care model in three different countries in Africa; it explicitly examines what is necessary to scale further, and we see no clear bottleneck to using further funds for further expansion. PSI's approach is extremely systematic, and its expenses have grown extremely quickly over the last few years; we believe that more funding would result in its doing more of its core activity (identifying needs and then marketing the appropriate products). Interplast has a Burn Outreach Center model that it has replicated recently in many regions, and we would guess that further funds would lead to further replication.

It is impossible for us to have any sense of an organization's ability to scale, without having a bird's-eye view of its activities; this means that we have far less confidence in what further funds would accomplish with any of the other organizations we discuss. Because most of them are extremely large and varied - and because the needs in Africa remain severe - we assume that more funding would result in more of some activity; we just don't know what.

The bottom line

We seek to fund a strong organization with strong programs; here we summarize what we know about both programs and organizations.

Programs
Program Organization Reliability/tangibility Cost per significant life change Type of life change Other benefits
Child survival (CSHGP) programs Multiple Moderate ~$1000 Life saved from infant/maternal mortality Improves nutrition; reduces diarrhea
Bednet marketing/distribution PSI Moderate $600-2400 Life saved from malaria Reduces malaria
Condom marketing/distribution PSI Moderate $150-2000 HIV infection prevented Slows spread of HIV; reduces other diseases; reduces unwanted pregnancies
Comprehensive health program PIH High ~$3500 Life saved Improves general health in many ways
Corrective surgery Interplast High $500-1500 Corrected deformity (may not always be major)
Vitamin A supplementation program HKI Moderate ? Reduced overall under-5 mortality Improves nutrition

As we explain above, we find the first three to be the strongest, though Interplast's and PIH's programs are notable for their tangibility and PIH's likely has more impact on quality of life in non-extreme ways (i.e., general health).

Organizations
Organization Organization-wide cost per significant life change % of activities we have good information on Organization-wide commitment to monitoring
Population Services International (PSI) $650-1000 per life saved 84% Strong
Partners in Health (PIH) ~$3500 per life saved ~50% Moderate
Interplast $500-1500 per surgery 75% Strong
Helen Keller International (HKI) ? Not apparent
Project HOPE ? Not apparent
Food for the Hungry ? Not apparent
GNNTDC ? 100% Not apparent
HealthStore foundation ? Very low Not apparent
American Red Cross ? Very low Not apparent
UNICEF ? Very low Not apparent
International Eye Foundation ? Very low Not apparent
Aga Khan Foundation ? Very low Not apparent

In some cases, having little information about an organization may simply reflect the time demands on its development staff. However, the ease with which PSI was able to send us a "bird's-eye view" shows that it has a commitment to strong documentation and transparency that is already in place, and we consider this a strong asset in an organization seeking funds from the public at large.

PSI simultaneously demonstrates the strongest commitment to systematic monitoring and evaluation - making it possible for an outsider to understand its varied activities - and carries out programs that we find highly cost-effective. A donation to PSI is the best way we know of to save lives in Africa.

Deciding between PIH and Interplast is a particularly tough judgment call, and depends partly on how often Interplast's surgeries represent significant life change vs. more minor corrections. Having little information on this, we prefer PIH, whose comprehensive and community-based model gives us higher confidence that people are experiencing significant improvements in their quality of life.

The full details

The links on the left give our full writeups on finalists, as well as our criteria for choosing finalists, and should make the basis for all of our above claims clear.