Giving What We Can focuses on finding the charities which result in the most benefit to the greatest number of individuals. But should we give more weight to benefitting the worst off? Should we care more about equality and distributive justice?

Giving What We Can aggregates the benefits to different people additively rather than giving extra weight to worse off individuals. While our top charities tend to help some of the worst off people in the world, we do not explicitly factor considerations of fairness and equity into our cost-effectiveness calculations. There are good reasons for this:

  1. It’s simpler. It can be very complex to add together different benefits and also weight them differently for different people. This is why academic literature on cost-effectiveness tends to aggregate benefits additively.

  2. It’s less controversial. Aggregating these benefits additively relies on the ethical intuitions everyone actually has. While egalitarians care about the distribution of benefits, they also care about total benefits. Different people may have very different ideas about how much weight to place on considerations of equity. It is easier to add your own weighting than to adjust ones that we have assumed.

  3. It usually comes to the same answer. Those who are worst off are usually the easiest to help as there are more “low hanging fruit” so to speak.

We are certainly not suggesting that utilitarianism is the only plausible ethical theory (as has sometimes been suggested[1][2]). But we do think that aggregating benefits additively provides a good starting point, which can then be supplemented by considerations of equity and fairness if required[3]. We are aware that many of our members (and staff) think that prioritising helping the worst off may be of intrinsic importance, for either prioritarian or egalitarian reasons (for the purposes of this post we will refer to both as egalitarian[4][5]). Recently, we’ve been thinking more about how our recommended charities fare on these ethical dimensions.

If aggregating benefits in an additive or egalitarian manner always came to the same answer (as it often does), that would be the end of the matter. Occasionally however, we may have to make a tradeoff between helping people who are worse off or helping a larger number of people. This post is intended to give the reader some additional considerations, which may help them think about how to weigh egalitarian concerns in choosing the most effective charities.

In this post, I will argue that:

  1. Additive benefits are the most important consideration when choosing an effective charity. However, when the cost-effectiveness of charities is sufficiently close, it may be reasonable to factor in considerations of fairness and equity.
  2. All our top charities fare well on considerations of fairness and equity. This is unsurprising given that they are all charities which help children with preventable diseases in poorer countries.
  3. Against Malaria Foundation (AMF) fares particularly well on fairness and equity grounds as it prevents a disease with a high individual disease burden which affects the economically disadvantaged.
  4. It is possible that there are other comparably cost-effective charities which do better than some of our charities on fairness and equity. However, AMF sets a high bar, as it is estimated to save a life for about $3,000, and children who die under the age of five are certainly among the very worst off globally.

How to make fair choices

In 2014, the World Health Organisation published Making Fair Choices on the path to Universal Health Coverage, which attempts to provide a framework for incorporating considerations of fairness and equity into healthcare prioritisation[6]. The report was authored by a diverse group of 18 health economists, ethicists and practitioners in order to represent broad ethical consensus. Several members of Giving What We Can contributed to the report[7].

Making Fair Choices suggests three different considerations when deciding which interventions to prioritise:

  1. Cost-effectiveness: How many people can be helped and how much they can be helped for a given level of resources?
  2. Priority for the worse off: Are the people helped particularly disadvantaged?
  3. Financial risk protection: How much does the intervention reduce the risk of impoverishment due to catastrophic health expenditure?[8]

Figure 1: Framework for integrating cost-effectiveness with other criteria when selecting services (Source: Making Fair Choices)

Not all these considerations should be given equal weight. The suggested decision procedure is to separate interventions into priority categories by setting overlapping thresholds for the cost-effectiveness of interventions. Where an intervention lies in the overlapping segment between different categories, considerations of priority for the worse off and financial risk protection are applied to determine which category it falls into. So while priority for the worse off is explicitly recognised as a relevant criteria for assessing which interventions to prioritise, it is only taken into account when the cost-effectiveness of two interventions is sufficiently close.

Under the example thresholds laid out in the report, all the interventions recommended by Giving What We Can fall firmly in the high priority category. That is, even if there were substantial reasons of fairness and equity to prioritise other interventions, this would only be relevant if those interventions were also within the high priority category.

However, there is still likely to be a substantial range of interventions which fall into the highly effective category and so, when the cost-effectiveness of interventions is sufficiently close, considerations of fairness and equity may come into play[9].

How to evaluate fairness and equity?

One difficulty with factoring in priority for the worst off is that people can be worse off across a number of different dimensions. One solution would be to use a covering value of wellbeing which would incorporate all these considerations (such as subjective reported well being). Unfortunately, such a covering value would be very difficult to base on reliable data and highly controversial as different people value the various aspects of welfare in different ways. This mirrors the difficulties in evaluating charities on direct hedonistic grounds. In practice, consideration of who is worse off should be based on a number of different dimensions which contribute to welfare. The case studies accompanying the report (not yet released) make some suggestions[10]:

  1. Health outcomes: In the absence of the intervention, would the beneficiary bear a large individual burden of disease? The individual burden of disease can be proxied by measures such as average years of life lost (AYLL).
  2. Social and economic status: How badly off is the individual in terms of income, education or occupation?

So how do our top charities fare on these grounds?


CharityWho do they primarily help?How badly off are they?
Against Malaria FoundationMostly children under the age of 5 who would die of malariaHealth outcomes: Very bad
  • Those who die from malaria tend to be very young, with 65 years of life lost on average[11].

Social and economic status: Very bad

  • The countries in which AMF operates are among the poorest in the world[12].
  • Even within those countries, malaria predominantly affects the poorest[13].
  • The poorest are also the groups least likely to have already purchased ITNs[14].
Schistosomiasis Control InitiativeChildren with schistosomiasis across AfricaHealth outcomes: Bad
  • The individual disease burden of schistosomiasis is far lower than that of malaria (0.5% DALY weighting. This weighting has been criticised as an underestimate as it does not account for clinical sequelae[15][16][17]. Givewell use an estimate of 2-5%).
  • However, populations who are at risk from schistosomiasis tend to live in areas with poor water and sanitation systems[18], and have generally poor health prospects[19].

Social and economic status: Very Bad

  • The countries in which SCI operates are among the poorest in the world[20].
  • Even within those countries, schistosomiasis predominantly affects the poorest,[21] primarily in areas where fishing and irrigated agriculture represent a large segment of the economy[22].
  • Part of SCI’s programme is focused on school-based deworming and it is possible that this means the very worst off (who may not attend school) do not receive treatment[23].
  • However, the possible spillover effects of deworming[24] means that those who do not receive treatment may still benefit.
Deworm the World InitiativeChildren with soil-transmitted helminth infections across India and more recently in EthiopiaHealth outcomes: Bad
  • The individual disease burden of soil-transmitted helminths is generally considered to be lower than schistosomiasis (Givewell use a weighting of 0.7%-1.7%).
  • However, populations who are at risk from STH tend to have poor health prospects[25].

Social and economic status: Very Bad

  • Those at risk from STH infection tend to be the poorest in a country[26].
  • STH infections are roughly twice as prevalent among socioeconomically worse off children[27].
  • India has a higher GDP per capita than most countries in which SCI operates[28].
  • However, due to high economic inequality in India, the worst off (who are most at risk from STH) are likely to be roughly equivalent in terms of social and economic status to those at risk from schistosomiasis in the countries in which SCI operates. We could not find reliable data to confirm this hypothesis.
  • DtWI’s programme is focused on school-based deworming and it is possible that this means the very worst off (who may not attend school) do not receive treatment[29].
  • However, the possible spillover effects of deworming[30] means that those who do not receive treatment may still benefit.
Project Healthy ChildrenPeople with micronutrient deficiencies in poorer countries. Micronutrient deficiencies have the greatest impact on children (through child development) and women (who are more likely to be anemic)Health outcomes: Bad
  • Micronutrient deficiency is a risk factor for a number of diseases including diarrhea, iron deficiency, iodine deficiency and lower respiratory infections[31].
  • These diseases have highly variable individual disease burdens but at least the worst, such as diarrhea are very serious[32].
  • On average, these diseases have a lower individual disease burden than malaria but higher than deworming.

Social and economic status: Bad

  • PHC operates in some of the poorest countries in the world[33].
  • Micronutrient deficiency tends to affect the economically worse off, who are unable to access a nutritious diet[34].
  • Those living in poor and rural areas sometimes remain unable to access centrally processed foods[35].
  • However, PHC take active steps to mitigate this problem by equipping smaller scale mills with fortification equipment.
  • Women are at greater risk from iron-deficiency anemia[36].

As it turns out, all of Giving What We Can’s top charities help people who could be considered to be among the worst off in the world. This is unsurprising given that they are all charities which help people with preventable diseases in relatively poor countries. We encourage donors who value equity and fairness highly to take these considerations into account to supplement our recommendations. In practice, however, it appears that substantial tradeoffs between total benefits and considerations of fairness are rarely required. We therefore plan to continue evaluating charities on the basis of their additive benefits.

Against Malaria Foundation helps people who could plausibly be described as the very worst off in the world (children who die under the age of five). For donors who place a very high weight on justice and equality, you could do a lot worse than donating to the Against Malaria Foundation.



Footnotes

  1. Gabriel, Iason. "Effective Altruism and its Critics." Journal of Applied Philosophy (2016). ↩︎

  2. It is understandable that effective altruism is often conflated with utilitarianism. In part, this is because prominent members of the movement such as Peter Singer and William Macaskill are academic philosophers who defend utilitarianism. It would be wrong however, to suggest that this characterisation is universal. ↩︎

  3. We are not neutral between all ethical theories. Global humanitarianism (the belief that all lives are of equal value, regardless of nationality, ethnicity and gender) is a core principle of our work ↩︎

  4. There are many philosophical differences between egalitarianism and prioritarianism. However, in practical terms, the key consideration is that both favour prioritising interventions which help the worse off if additive benefits are comparable. ↩︎

  5. Egalitarianism here refers to pluralist egalitarianism, the theory that both aggregate benefits and considerations of equality are important for their own sake. We are not referring to maximin egalitarianism, the theory that only the welfare of the least well off person in society is important. This view is widely held to be implausible as it would favour a world in which everyone was only marginally better off than the currently worst off person. ↩︎

  6. Norheim, Ole et al. Making fair choices on the path to universal health coverage: Final report of the WHO consultative group on equity and universal health coverage. World Health Organization, 2014. ↩︎

  7. Ole Norheim, Toby Ord, Nir Eyal and Alex Voorhoeve are authors of the original report. Alex Voorhoeve, Ole Norheim and James Snowden worked on the case studies to accompany the report ↩︎

  8. While important, financial risk protection is not the focus of this post ↩︎

  9. It should also be noted that the thresholds in Figure 1 are just examples and, in practice, will be set at the country-level. Suppose, for example, that the thresholds chosen for high and medium priority are such that there is substantial overlap and that a number of interventions fall between these thresholds. We then have to think about how the interventions carried out by our charities fare on fairness and equity grounds. ↩︎

  10. The report also suggests two other dimensions: health prospects, and discrimination and marginalisation. For simplicity, in this post we treat health prospects and health outcomes together, and ignore discrimination and marginalisation which is often vaguely defined or covered by other factors (for example socioeconomically worse off groups could be described as marginalised). ↩︎

  11. Based on GBD 2010 data. Calculated using health adjusted life expectancy in Tanzania. ↩︎

  12. "GDP per capita (current US$) | Data | Table - The World Bank." 2010. 12 May. 2016 <http://data.worldbank.org/indicator/NY.GDP.PCAP.CD> ↩︎

  13. "Does Malaria cause poverty or is it vice versa | MalariaWorld." 2014. 11 May. 2016 <https://malariaworld.org/blog/does-malaria-cause-poverty-or-it-vice-versa> ↩︎

  14. Onwujekwe, Obinna, Kara Hanson, and Julia Fox-Rushby. "Inequalities in purchase of mosquito nets and willingness to pay for insecticide-treated nets in Nigeria: challenges for malaria control interventions." Malaria journal 3.1 (2004): 1. ↩︎

  15. "WHO | Assessment of the age-specific disability weight of chronic …" 2008. 11 May. 2016 <http://www.who.int/bulletin/volumes/85/6/06-033035/en/> ↩︎

  16. "schistoinir - University of York." 2008. 11 May. 2016 <http://www.york.ac.uk/res/schistoinir/schistodisease.html> ↩︎

  17. King, Charles H. "Schistosomiasis japonica: the DALYs recaptured." PLoS Negl Trop Dis 2.3 (2008): e203. ↩︎

  18. "WASH Away NTDs - Centers for Disease Control and Prevention." 2010. 11 May. 2016 <http://www.cdc.gov/healthywater/pdf/global/programs/wash-away-ntd-program-brief.pdf > ↩︎

  19. "Schistosomiasis (bilharzia) - NHS Choices." 2008. 11 May. 2016 <http://www.nhs.uk/conditions/schistosomiasis/Pages/Introduction.aspx> ↩︎

  20. "GDP per capita (current US$) | Data | Table - The World Bank." 2010. 12 May. 2016 <http://data.worldbank.org/indicator/NY.GDP.PCAP.CD> ↩︎

  21. "Study of schistosomiasis control based on the equality of public health …" 2016. 11 May. 2016 <http://caod.oriprobe.com/articles/41478579/Study_of_schistosomiasis_control_based_on_the_equality_of_public_healt.htm> ↩︎

  22. "PLOS Neglected Tropical Diseases: Socioeconomic Inequalities in …" 2016. 13 May. 2016 <http://journals.plos.org/plosntds/article?id=info:doi/10.1371/journal.pntd.0004546> ↩︎

  23. Anderson, Roy M et al. "How effective is school-based deworming for the community-wide control of soil-transmitted helminths?." PLoS Negl Trop Dis 7.2 (2013): e2027. ↩︎

  24. Ozier, Owen W. "Exploiting externalities to estimate the long-term effects of early childhood deworming." World Bank policy research working paper 7052 (2014). ↩︎

  25. Hotez, Peter J et al. "Helminth infections: soil-transmitted helminth infections and schistosomiasis." (2006). ↩︎

  26. "WHO | Soil-transmitted helminth infections." 2012. 11 May. 2016 <http://www.who.int/mediacentre/factsheets/fs366/en/> ↩︎

  27. "PLOS Neglected Tropical Diseases: Socioeconomic Inequalities in …" 2016. 13 May. 2016 <http://journals.plos.org/plosntds/article?id=info:doi/10.1371/journal.pntd.0004546> ↩︎

  28. "GDP per capita (current US$) | Data | Table - The World Bank." 2010. 11 May. 2016 <http://data.worldbank.org/indicator/NY.GDP.PCAP.CD> ↩︎

  29. Anderson, Roy M et al. "How effective is school-based deworming for the community-wide control of soil-transmitted helminths?." PLoS Negl Trop Dis 7.2 (2013): e2027. ↩︎

  30. Ozier, Owen W. "Exploiting externalities to estimate the long-term effects of early childhood deworming." World Bank policy research working paper 7052 (2014). ↩︎

  31. "GBD Compare | IHME Viz Hub - Data Visualizations - Institute for …" 2014. 12 May. 2016 <http://vizhub.healthdata.org/gbd-compare/> ↩︎

  32. WHO, Geneva. "WHO methods and data sources for global burden of disease estimates 2000-2011." (2013). ↩︎

  33. "GDP per capita (current US$) | Data | Table - The World Bank." 2010. 12 May. 2016 <http://data.worldbank.org/indicator/NY.GDP.PCAP.CD> ↩︎

  34. "WHO | Micronutrient deficiencies." 2006. 12 May. 2016 <http://www.who.int/nutrition/topics/ida/en/> ↩︎

  35. "The role of food fortification in the control of micronutrient malnutrition." 2010. 12 May. 2016 <http://www.who.int/nutrition/publications/micronutrients/GFF_Part_1_en.pdf > ↩︎

  36. "Iron deficiency anaemia - NHS Choices." 2009. 12 May. 2016 <http://www.nhs.uk/conditions/Anaemia-iron-deficiency-/Pages/Introduction.aspx> ↩︎