Against Malaria Foundation – Comprehensive Summary

Last updated: November, 2023

Problem

Malaria is a potentially life-threatening disease caused by parasites that are transmitted by certain mosquitoes. In 2020, there were an estimated 627,000 deaths due to malaria (WHO, 2021).

Intervention

Long-lasting insecticidal nets (LLINs) protect against exposure to malaria-carrying mosquitoes at night, when they are most active, and reduce the mosquito population by killing those that come in contact with the nets. 

Organisation

The Against Malaria Foundation (AMF) funds, and helps coordinate, the distribution of LLINs to help prevent malaria in sub-Saharan Africa, South Asia, South and Central America, and Oceania.

Evaluation

Methods

LLINs can promote wellbeing in two ways:

1. They can save lives
2. They can improve the quality of lives

We assessed the total impact of AMF by estimating the subjective wellbeing benefits due to both. Assessing the impact of each requires distinct methods. 

Life-saving effects

The simplest method to estimate the life-saving effects of bednets is to multiply the additional years of life by the expected wellbeing experienced during those years. We calculate these figures using existing information about:

• The average age of death due to malaria
• The average levels of life satisfaction in the countries AMF intervenes in
• The average life expectancies in the countries AMF intervenes in

However, completing this analysis requires addressing several philosophical issues. 

First, it is conceivable that people may have positive or negative wellbeing, so we must set a point on the subjective wellbeing scale that represents the “neutral point” of wellbeing. This is the point on the scale where scores above represent positive wellbeing, while scores below represent negative wellbeing. For example, on a 0-10 life satisfaction scale, the neutral point could be set at 0 or 5. 

Second, we must decide how bad we think death is. We could assume that:

• the badness of death consists of the wellbeing you would have had if you’d lived longer, which we described as the “simplest method” above (this view is called “deprivationism” in philosophy).
• We could assume that the badness of death for the individual depends on how “connected” they are to their possible future self¹For example, infants may be only weakly psychologically connected to their later selves, so their ‘interest’ in living may be less than that of an adult. . Under this view, lives saved at different ages are assigned different weights (this is called the “time-relative interest account”).
• We could assume that death is not bad for those who die (this is called “Epicureanism”).

These philosophical factors – the choice of where to place the neutral point and the badness of death – can strongly influence the estimated impact of AMF, and reasonable people will hold different views. We estimate the impact of AMF across a subset of possible views, but we do not take a stance about which view is correct. 

For more detail, see our report (Plant et al., 2022). You can also explore the estimated impact of AMF under different philosophical views using our online app. 

Life-improving effects

To estimate the life-improving impact of AMF, we would ideally use direct evidence of the impact of malaria prevention on subjective wellbeing. To our knowledge, no such research exists. So as a proxy, we use GiveWell’s existing estimates of the impact of malaria prevention on income ²We use GiveWell’s estimate that a year of malaria prevention in childhood increases income by 1.2% over the person’s lifetime. We then make several adjustments that together lead to a 81% discount. The first adjustment is accounting for a key difference between the way that cash transfers and malaria prevention increase incomes. When it comes to increases in income, part of the subjective wellbeing benefit may be due to a comparison effect: comparing yourself with others and finding yourself better or worse off (Clark et al., 2018; Easterlin, 2021). Comparison effects apply to cash transfers, but they probably do not apply to malaria prevention: with mass distribution of bednets, the benefits accrue to everyone in an area, while cash transfers are only sent to the poorest. We estimate that these comparison/relative income effects comprise 26% of the total income effect. Including this adjustment discounts the life-improving effect of malaria prevention from 23 down to 17 WELLBYs. We proposed further discounts to AMF’s income effects in our critique of GiveWell’s cost-effectiveness analyses (McGuire et al., 2022c). We estimate that, cumulatively, the issues we find suggest a 75% reduction to GiveWell’s income effect (row 35, sheet “Inputs”). Thus, we estimate the effect is (1 – 0.75) x 17 = 4 WELLBYs. , and then convert these income effects into WELLBYs using figures from our previous work on the impact of cash transfers on subjective wellbeing ³GiveWell estimates that spending $1,000 on AMF’s malaria prevention programme increases economic benefits by 18 log-units (or about 13 doublings of income; see row 247, “Modified” tab). We use our cash transfers meta-analysis (Model 2 in Table 2 of McGuire et al., 2022d) to translate the economic benefits into WELLBYs. We estimated that a 1-unit increase in log-economic benefits leads to a gain of 1.3 WELLBYs. . 

AMF may provide additional life-improving benefits by averting the grief of bereaved family and friends. We conduct a shallow calculation of this effect using existing estimates from studies on the effect of the death of a child on life satisfaction and a study on the duration of grief. 

Impact

For each life saved through AMF, we estimate the life-saving benefit ranges from 0-247 WELLBYs across the philosophical views we explore. 

The estimated life-improving benefit from increased income is 12.51 WELLBYs, and the life-improving benefit from averting grief is 7.26 WELLBYs. 

Overall, we estimate the total effect of AMF for each life saved is 20-267 WELLBYs.

Cost

We use GiveWell’s estimates that it costs $3 to provide a bednet to one child, which provides effective coverage for two years. According to GiveWell’s calculations it costs $2,941 to save a life. In other words, a $1,000 donation will save 0.34 lives in expectation; or, it has a one-in-three chance of saving a life.

Cost-effectiveness

The cost-effectiveness of AMF depends heavily on the philosophical view. 

The view we examined that is most favourable to extending lives is deprivationism with a neutral point of 0. Under this view, the total cost-effectiveness of AMF is $2,941 / 267 WELLBYs = $11 per WELLBY. Or 90 WELLBYs per $1,000 donated to AMF.

The view we examined that is least favourable to extending lives is Epicureanism with any neutral point. Under this view, the total cost-effectiveness of AMF is $2,941 / 20 WELLBYs = $149 per WELLBY. Or 7 WELLBYs per $1,000 donated to AMF.

Under the time-relative interest account, the cost-effectiveness is somewhere between these two extremes, with the exact value depending on both the value of the neutral point and the age at which a person is considered to be fully connected to their future self. We plot the cost-effectiveness ranges for each view in Figure 1.  

Figure 1. Cost-effectiveness of AMF under different philosophical views. 

Note. The lines represent the upper and lower bound of cost-effectiveness for different philosophical views. Think of them as representing moral uncertainty, rather than empirical uncertainty. The upper bound represents the assumptions most generous to extending lives and the lower bound represents those most generous to improving lives. The assumptions depend on the neutral point and one’s philosophical view of the badness of death (see Plant et al., 2022, for more detail). These views are summarised as:

• Deprivationism: The badness of death consists of the wellbeing you would have had if you’d lived longer.
• Time-relative interest account (TRIA): The badness of death for the individual depends on how “connected” they are to their possible future self. Under this view, lives saved at different ages are assigned different weights.
• Epicureanism: Death is not bad for those who die.

You can see how the cost-effectiveness of AMF (relative to other charities) changes across a number of philosophical views in Figure 2. You can also explore the cost-effectiveness of more views in this online app.

Figure 2. Cost-effectiveness of AMF under different philosophical assumptions and compared to other charities

Note. The solid lines represent the cost-effectiveness of AMF depending on the view (differentiated by colours) and neutral point (the x-axis). AC stands for ‘age of connectivity’, the age at which lives are assigned full weight in the model. All the dashed lines represent the primarily life improving charities we’ve evaluated. The grey horizontal dashed line at the top of the figure represents our previous cost-effectiveness estimate for StrongMinds. Below that, the dashed horizontal light blue line represents the cost-effectiveness of Friendship Bench. Below that, the dashed blue horizontal dark blue line represents the new, lower cost-effectiveness of StrongMinds. At the bottom of the graph, the dashed green line represents GiveDirectly. 

Quality of evidence

We think that the quality of evidence supporting the effect of AMF is moderate. The primary reasons are:

• The evidence about the life-saving effects of insecticide-treated bednets is high quality. This effect is based on a Cochrane review of 5 RCTs (Pryce et al.. 2018).
• But there is no evidence of the effect of malaria prevention’s life-improving effects on subjective wellbeing.
• The income-improving effects are from historical episodes of malaria eradication 50-100 years ago, so we’re very uncertain if the effects are generalisable to present malaria-suppression efforts.
• There is very little study of the neutral point, an important input into understanding the life-saving benefits.

Depth of Analysis

Shallow. We believe that we have included most of the available evidence, but we have only spent limited time on the analysis and our analyses rely on a number of estimates that we have not vetted in depth.

Funding need

GiveWell estimated that AMF could absorb $33.2 million in 2023. 

Conclusion

Our analysis suggests that the cost-effectiveness of AMF depends heavily on one’s philosophical views. Under views more favourable to improving lives, AMF is less cost-effective than treating depression with psychotherapy. However, under views more favourable to saving lives, AMF is more cost-effective than treating depression with psychotherapy. Thus, AMF may be a good option for donors who highly value saving lives. We have made an online app that you can use to examine the cost-effectiveness of AMF under various philosophical views.

FAQs

• Does HLI take a stance on the philosophical questions of the badness of death and the neutral point?
• No, we do not take a stance. We present the range of results according to the different views.
• What is the relationship between the use of the subjective wellbeing approach and these philosophical choices?
• There is no direct link. Whatever you choose as your measure of quality of life, there are independent, further questions about how to value quantity of life. Existing measures of health (e.g., DALYs) have to contend with these issues as well.
• Are you arguing that it is better for some people not to be saved?
• On some philosophical views it is always better to be alive. On others, it depends on how good or bad the life is. On a third set, existence and non-existence aren’t comparable. As an organisation, we do not take a stand on this issue, and rather present the results on a range of assumptions. These are difficult philosophical issues, and we encourage donors to develop views for themselves. Note that prioritisation in healthcare always involves a choice between whether it is better to save lives or improve lives, given resources are limited.
• Do you account for the impact that more or fewer people on the planet has on others?
• The only ‘spillover’ factor we consider is the grief to friends and family when someone dies. We do not take a stand on whether a larger or smaller population is better, nor do we know how we would put numbers on that if we wanted to.
• What additional work do you plan to do on the issue of how to weigh quality against quantity of lives?
• We plan to conduct original research to better understand the location of the neutral point on subjective wellbeing scales. However, we do not have any immediate research plans with regards to philosophical views on the badness of death. We think this is a difficult philosophical issue that is best left up to individual donors to decide for themselves. We do plan to do some work on ‘moral uncertainty’, how it is appropriate to make decisions when we are uncertain about ethics (as opposed to being unsure about the facts).

Our reports so far

We have one moderate-depth report on AMF. We have made an online app that you can use to see the impact of your answers to the philosophical questions on the calculations of AMF’s cost-effectiveness. We have also critiqued GiveWell’s use of the evidence for income gains due to antimalarial bednets in our dozen doubts report.