The elephant in the bednet: a brief overview

18 November 2022

If you want to maximise the impact of your charitable donations, is it better to give to charities that improve lives or extend lives? This may seem like a philosophy-free endeavour: don’t we just crunch the numbers and see what comes out on top? In fact, we need to make some crucial philosophical assumptions to decide how to crunch the numbers and value extra years of life. Different assumptions can lead to very different conclusions.

In our new report, we set out two philosophical choices and show how they alter the cost-effectiveness of three charities considered to be some of the best ways to help others: Against Malaria Foundation (‘AMF’; insecticide-treated bednets), GiveDirectly (cash transfers), and StrongMinds (group psychotherapy for depression).

On the assumptions most generous to extending lives, AMF is 1.3x more cost-effective than StrongMinds. However, on the assumptions most generous to improving lives, StrongMinds is 12x better. The philosophy is a crucial consideration – not a rounding error – one that merits greater attention. Because people will disagree about ethical issues, it’s essential that these underlying assumptions are brought to the surface.

The report is quite detailed, so this blogpost gives a simple, high-level overview of what the problems are and why they can’t be avoided. More determined readers are welcome to venture ‘into the weeds’ at their own risk…

The first problem is how to value deaths at different ages. Is it better to save a 1-year-old or 10-year-old? Intuitively, we should save the 1-year-old. Why? Well, they have nine more years to live. At first glance, the right principle is ‘prioritise the youngest’.

Now, let’s ask: Is it better to save a 1-week-old embryo, or a 1-year-old child? The simple ‘prioritise the youngest’ principle, taken to its logical conclusion, would tell us to save the embryo. But many have the intuition that it’s more valuable to save the 1-year-old. What might explain this intuition? One thought is that the 1-week-old isn’t nearly as developed, so they don’t have as strong a connection to their later selves and therefore have the weaker ‘interest’ in living than the 1-year-old.

Thinking this way, perhaps the most valuable lives to save are not infants, but those who are a bit older, whose psychology is maturing, but still have many years of life left; hence, we should adopt a ‘prioritise older children over infants’ principle instead. The choice between these principles is particularly relevant here, given that many of the lives that AMF saves are those of very young children.

Alternatively, you could take the view of the ancient Greek philosopher, Epicurus, that death is not bad for us because nothing can be bad for us if we don’t exist. From this, we get a ‘prioritise living well, not living long’ principle. Although Epicureans won’t directly value preventing deaths, life-extending interventions can still have some value, for instance preventing grief from bereavement.

The second issue comes from how to measure quality of life. In order to compare the value of different options, we need to combine quality and quantity of life into a single number. One way to measure quality of life – a way we’re very sympathetic to – is asking people to rate their life satisfaction on a 0-10 scale (0 – not at all satisfied, 10 – completely satisfied). To determine the value of an extra year of life, we need to know how good that life is in relation to the ‘neutral point’, the location on the scale where life is neither good nor bad for someone. The problem is, it is unclear where to place the neutral point on the 0-10 life satisfaction scale.

We might think the neutral point should be 0/10. However, that would make it impossible for any life to be worse than death, even a life of unrelenting misery. Yet, we generally think life can be both better and worse than death.

Alternatively, we could take the neutral point to be 5/10, the middle of the scale. Average life satisfaction in many low-income countries is below 5, this implies, counter-intuitively, that many people have currently negative lives and that, possibly, it would be bad to save them.

A third option is to put the neutral point somewhere between 0 and 5. But this is arbitrary, absent any further justification. A couple of small surveys have been conducted to assess where people think the neutral point is, but this is fairly weak evidence and the results are inconclusive. It’s hard to know what the most sensible answer is (we’re undertaking our own survey on the subject).

Clearly then, there are some difficult choices to make—choices that are both consequential and unavoidable. The report explains these choices in more detail, shows what happens when we crunch the numbers in different ways, and sets out what further work is needed. We hope that, by the end of it, readers will have a better understanding of the issues they face and what they think it would be best to do.