by Brian Tomasik
First written: Feb. 2016; last update: 17 Jul. 2017

## Summary

It's plausible that humanity reduces more wild-animal suffering than it creates on balance. If increasing the human population reduces wild-animal suffering, then we might favor charities that save human lives in order to increase the number of humans who exist. The Against Malaria Foundation (AMF) is one example of an effective life-saving charity. A very rough calculation suggests that AMF may prevent on the order of 104 invertebrate-years of suffering per dollar donated to it, although the magnitude and even the sign (+ or -) of this figure is highly uncertain. A priori, AMF seems to me more likely to reduce wild-animal suffering than other anti-poverty charities that don't save lives because non-life-saving charities probably have less impact on the human population size (and may even decrease human fertility rates). And anti-poverty charities that involve family planning may have the opposite effect as AMF with respect to wild-animal suffering.

## Rough cost-effectiveness estimate for AMF

Cotra (2016) reports the following rough GiveWell estimates:

a $37,391 donation to AMF would, in expectation: • Prevent the deaths of ~4.08 children under the age of 5 [...] • Prevent the death of ~1 person over the age of 5 [...] • Have a financial benefit equivalent to [roughly doubling the income of] 74.782 people [for a year.] I'll ignore the effect of increasing incomes and focus on the lives saved. In a sample calculation, Cotra (2016) suggests that we might assume "a life expectancy of ~65 years at age 5", i.e., that saving a child under age 5 adds 65 years to that person's life on average. Maybe this is too optimistic; let's reduce it to ~40 years to be conservative. I'm uncertain what life expectancy to use for people over age 5, but let's say it's ~20 years, just to make up a number. Then$37,391 donated to AMF saves an expected 4.08 * 40 + 1 * 20 = 183 years of life.

We also need to consider how saving one child's life may inhibit mothers from having additional children. Estimates here are messy, but Cotra (2016) suggests an informal, rough estimate that every child saved prevents 0.37 to 0.52 future births, or maybe even fewer. I'll assume that the number is 0.5, which gives a more conservative cost-effectiveness estimate than using a lower number here.

I'll also assume no replacement effect for children over age 5, since if a child dies when she's older, this is less likely to influence her parents' fertility choices. Moreover, as Carl Shulman notes: "increased population compounds over time with the population growth rate." Saving women over 5 years of age may save the lives of future mothers, including pregnant women: "Malaria infection during pregnancy is a significant public health problem with substantial risks for the pregnant woman, her fetus, and the newborn child."

Factoring in prevented future births, we find that $37,391 donated to AMF creates a net increase of 4.08 * 40 * 0.5 + 1 * 20 = 102 years of life in expectation. Finally, let's assume, as I did here, that one person-year lived by someone in a malaria-dense country reduces invertebrate suffering by ~5 * 106 invertebrate-years. Then the cost-effectiveness of donating to AMF would be (5 * 106 invertebrate-years prevented per human-year created) * (102 human-years created) / ($37,391 donated) = 1.4 * 104 invertebrate-years prevented per dollar.

## Does saving human lives reduce invertebrate populations?

Malawi is one of the countries where AMF works. While poor, people in Malawi have significant environmental impacts that plausibly reduce net invertebrate suffering:

Most of Malawi was originally covered by forest but, over the years, people have been cutting down the trees and burning them where they fall to open up areas for farming — this is commonly known as “slash and burn” agriculture. [...] Whilst the population of Malawi was small, the environment was able to recover as the trees would regenerate but now [...] there is increased pressure on available land. [...]

Wood is the main fuel in Malawi, and 95% of homes still use wood or charcoal for cooking. Nearly everybody uses three-stone fires for cooking, and each fire consumes about three large bundles (weighing about 30kg each) of wood per week.

Burning is significant because it eliminates large amounts of biomass in a non-sentient way, preventing stored plant energy from being eaten by bacteria, fungi, and bugs.

After a friend of mine visited Kenya, he told me regarding human impact on net primary productivity: "My observation was that cultivated land is relatively unproductive compared to native vegetation in the villages I visited. Farmers tend to work on a rather small scale, though." As far as firewood, my friend reported: "I talked to a professor about this; he said that in Kenya, 70% of energy comes from wood. In the villages I visited, people tend to cook with firewood or charcoal."

Combining the above information with general trends about human environmental impacts, it seems more likely than not that an increased human population in malaria-prone countries reduces net wild-animal numbers. However, more research is needed on this issue. For example, the net impacts of climate change on wild-animal suffering are unclear. Speeding up economic development may hasten the spread of high-productivity agriculture to developing nations, which may increase total plant growth on crop fields. People inflict significant harm on animals, including insects, such as by eating them. Many other considerations like this add significant uncertainty to the analysis.

## Does AMF increase the human population?

There also remains some uncertainty about the net impact of AMF on the human population size. GiveWell tentatively suggests that "Overall, it appears that life-saving interventions unaccompanied by other improvements, where access to contraception is weak, are likely to lead to some acceleration of population growth." However, that post also notes that several prominent figures have argued the opposite, that "population growth tends to slow as child mortality declines, possibly because parents feel less need to have multiple children in order to hedge against the risk of death."

## AMF vs. other charities

It would be good to explore other charities that save lives in the developing world besides AMF. I chose AMF because GiveWell attempted a rough cost-per-life-saved calculation for it and because it's generally considered an effective organization. AMF seems superior to, e.g., deworming because AMF probably creates more life-years, while deworming mainly improves quality of life. Thus, AMF probably has bigger environmental impact. Given that I'm uncertain about the sign of reducing poverty for wild-animal suffering, the safest international-health charity from a wild-animal standpoint would be one that has much more impact on increasing human population size than it does on decreasing poverty. Indeed, just decreasing poverty might reduce human population growth, which could be net bad.

In addition, it seems plausible that family-planning charities reduce human population growth. A priori, I would expect this to increase wild-animal suffering. Thus, it seems important to avoid a blanket conclusion like "international-health charities reduce wild-animal suffering". (And even if one is only optimizing for a "total utilitarian" aggregate of human QALYs, the sign of net impact for mortality-reduction charities might differ from that of family-planning charities.)