This 2-part article discusses effective altruism (EA), an increasingly prominent method of charitable giving. The first post deals with the theoretical underpinnings of the movement, as well as what it looks like in action. The second will address possible shortcomings/points of debate.

Part 1

So just what is effective altruism?

Effective altruism is a philosophy and social movement that applies evidence and reason to determine the most effective ways to improve the world. Effective altruists aim to consider all causes and actions, and then act in the way that brings about the greatest positive impact.¹

Philosophical underpinnings

Effective altruists see a moral obligation to give, and feel called to help alleviate the suffering caused by the inequity of human experience.

This imperative is borne out of concepts of global justice — such as the work of John Rawls and Thomas Pogge — but more directly the utilitarian philosophy of Peter Singer, who in 1972 published his incredibly influential “Famine, Affluence, and Morality”. In it, he outlines a system in which humans have an equal obligation to every other being, regardless of location:

It makes no difference whether the person I can help is a neighbour’s child ten yards away from me or a Bengali whose name I shall never know, ten thousand miles away. […] The moral point of view requires us to look beyond the interests of our own society.

EA vs ‘conventional’ charity

Effective altruism is not unique simply because of its global outlook, however. The truly distinguishing factor is a data-driven approach to maximising impact.

Those who donate…seek evidence about what the charity is doing […] If the evidence indicates that the charity is really helping others, they make a substantial donation. Those who give small amounts to many charities are not so interested in whether what they are doing helps others […] Knowing that they are giving makes them feel good, regardless of the impact.²

Effective altruists have little time for those ‘traditional’ organisations that lack transparency and a commitment to evidence-driven action. Heifer is one such ‘traditional’ example — while international in scope, it has been consistently targeted for a refusal to release outcome data, and its dismissal of randomised control trials; in 2013 a representative summarised the organisation’s view with the following:

These are lives of real people and we have to do what we believe is correct. We can’t make experiments with peoples’ lives.

Giving $50 to send a goat to Congo might create what psychologists call a ‘warm glow’, but until mega-charities like Heifer demonstrate that they create lasting change, EA takes the stance that the needs of those aforementioned ‘real people’ are not being best addressed.

EA in action

So what practices does EA promote to reach a better form of giving? What makes a cause worth donating to? GiveWell — a EA-associated meta-charity — is a good place to start. There are several factors that they look at when making recommendations:

Effectiveness

Simply put, does it work? Do these charities create lasting impact through their actions or is this a cause that sounds promising, but ultimately doesn’t really do much to improve lives? GiveWell demands an unprecedented level of transparency from the aid community, so that the way funds are used and the impact they create can be accurately evaluated. This dovetails with the increased use of Randomised Control Trials to determine the most effective methods of poverty alleviation.

Many will argue that attempting to measure the ’good’ done by various forms of intervention is a fruitless effort — just as economists don’t attempt to use utility for interpersonal comparison. No measure can claim total accuracy, but EA believe that some attempt at consistent evaluation is certainly better than the status quo (where there is no standardised method to speak of). Metrics like the Quality-Adjusted Life Year (QALY) used by the NHS to make healthcare decisions are an example of how hard data can improve decision making in resource-constrained settings.³

Lessons learned: PlayPumps International⁴

PlayPumps was the hottest clean water solution going in the early 2000s. Invented by South African Trevor Field, it used the power generated by children playing on a merry-go-round to power water pumps in communities that lacked a reliable clean water supply. Field’s organisation advertised aggressively, raising millions of dollars, and installed upwards of 1,800 pumps by 2009.

It seemed like a remarkable success story, except for the basic fact that no-one had really studied whether the pumps were actually useful for those who relied on them. Unfortunately, the answer was no: they were too hard for children to push, too complicated to repair, and extremely expensive. PlayPumps’ star fell almost as quickly as it had risen in the face of several damning international reports. For effective altruists it serves as a cautionary tale — good intentions are well and good, but you better back those up with results before diving into the deep end.

Impact/$

More important than simple effectiveness, though, is the cost required to generate such results. How much impact do charities generate per dollar spent? Many causes might be effective, but following EA’s tenant of doing the most good, EA want the most metaphorical bang for their buck. As in Maslow’s hierarchy of needs, EA-based giving seeks to start from the bottom. This usually results in supporting high-impact causes for the global poor such as disease eradication or direct cash handouts to immediately and dramatically improve their quality of life.⁵

The Pareto Principle: improving school attendance for girls

The Pareto Principle holds that roughly 80% of effects come from 20% of causes. EA takes much the same approach to charity — find the key that will unlock a disproportionate amount of value (e.g. maximise lives saved). The Deworm the World Initiative is a prime example of this principle in action. While there are several proven methods for increasing the school attendance rates of girls in sub-Saharan Africa, evaluating these methods for cost clearly demonstrates the superior per-dollar impact of de-worming.

Marginal utility

How much money does an organisation currently get? In order to maximise impact, you want to give to organisations that will benefit the most from donations. Cash-strapped charities can proportionally increase their production of ‘good’ on a per-dollar basis more than a large counterpart. Will MacAskill uses disaster relief as a prime example — well-publicised disasters tend to receive more money regardless of need. The 2011 Tōhoku earthquake and tsunami in Japan drew billions of dollars in support from around the world, except that the Japanese Red Cross was already well-capitalised, and issued a statement to that effect. Compare this to the 2008 Sichuan earthquake that was hushed up by the Chinese government, and subsequently received very little funding despite the lack of resources available to rescuers.

Ability to scale

Similar to the above point: how much would the charity benefit from more money? If their impact is limited by factors other than cash, then donations will not create maximum good. You want to find those organisations whose outcomes will scale alongside money received, so that an increase in donation will directly benefit the global poor, rather than contribute to bloated overhead.

Part 2

Limitations and critiques

Last year, GiveWell’s strict standards resulted in recommendations for only a select few charities. What causes were deemed most effective? Anti-malarial bed nets, Schistosomiasis control, de-worming, and direct cash handouts.

The limited scope of this list — both in the geography and types of causes it targets — provides a useful segue into a discussion of some personal reservations about the movement as a whole. Fundamentally, I agree that using data to drive better decisions about philanthropic giving is a big step forward, but I want to introduce 4 points that deserve some critical evaluation.

1. The arts

What about all of those causes that rely on donations, yet don’t contribute directly to saving lives? The arts, for example? They make the world an immeasurably richer place, yet effective altruists suggest that to support such causes over higher-impact alternatives is not only inefficient, but immoral. According to EA, to act in a fully rational (and necessarily extreme) way is to direct money to those life-enriching causes — like the arts — only once all individuals around the world are granted an equal chance at life. One can understand such a perspective, and encourage it, but it also gives pause for concern when we are compelled to forgo worldly pleasures, giving away all but enough for bare survival. Singer addresses this very question in his book, The Most Good you can Do:

Is it okay…for us to be going to the movies and drinking chai lattes while 1.4 billion people are living in extreme poverty?⁴

His answer (for me at least) is underwhelming:

Giving everything one owns to the poor is going to make it hard to earn more and thus to give more. You need to dress respectably to get a job, and today you may need a laptop and smartphone too. The best way of maximizing the amount you can give will depend on your individual circumstances and skills, but trying to live without at least a modest level of comfort and convenience is likely to be counterproductive.⁵

Or perhaps that’s really the point. We look for an easy moral system, one which tells us right and wrong, good and bad, clearly and without ambiguity. Yet what the fully developed individual can do (in the spirit of Aristotle again) is exercise judgement (phronesis), to define the moral limits of their own giving.

2. Earning to Give

Another concern is EA’s encouragement of earning to give. There are several well-publicised examples of individuals who took high-paying jobs with the intention of donating a large proportion of their income, and thus doing more ‘good’ than they would have as an employee of an NGO or charity. In terms of marginal benefit, the next-in-line for that NGO job forgone will still do a lot of good for the world, but the individual that would have worked in the hedge fund had the EA not taken that role would not have given nearly as much to charity. Good is therefore maximised by taking the hedge fund job over that at the WFP.

The organisation 80,000 Hours has been a major proponent of this sort of thinking, and while I agree once again with the premise, there is still cause for concern. Is there not a danger of becoming disconnected from the causes you choose to support, given the distance between you and those who you seek to assist? If it is declared immoral to give $10 to the homeless person you walk past every day (instead of donating it to an effective charity) does that not threaten to reduce those feelings of empathy that arise in the immediacy of suffering — feelings that are essential in encouraging consistent philanthropy in the first place? Not to mention the risk of lifestyle creep that comes as a result of taking that job in banking or finance, and living in an environment of privilege and self-gratification.

3. Existential Risks

A further reservation arises from the extreme rationalisation of ‘good’ — this time in the form of existential risk calculations. Existential risk refers to the likelihood of a catastrophic external event that threatens the survival of humanity. Nick Bostrom, who’s a major player in this field, has published a paper on the topic that classifies 6 different forms of risk:

Scope
global	Thinning of ozone layer	X
local	Recession in a country	Genocide
personal	Your car is stolen	Death
	endurable	terminal	Intensity

That X is existential risk: global in scale, terminal in effect. Much recent attention has been paid to the existential risk of a hostile super intelligent AI — a threat significant enough that individuals like Stephen Hawking, Elon Musk and Bill Gates have all publicly voiced concern. EA’s concern, then, is unsurprising given its core demographic of tech workers. Many of these individuals are now giving to support research about the existential risk of AI, rather than to direct poverty relief. This makes sense in terms of a strict decision calculus, as explained by Vox’s Dylan Matthews:

To hear effective altruists explain it, it comes down to simple math. About 108 billion people have lived to date, but if humanity lasts another 50 million years, and current trends hold, the total number of humans who will ever live is more like 3 quadrillion. Humans living during or before 2015 would thus make up only 0.0036 percent of all humans ever […] Put another way: The number of future humans who will never exist if humans go extinct is so great that reducing the risk of extinction by 0.00000000000000001 percent can be expected to save 100 billion more lives than, say, preventing the genocide of 1 billion people. That argues, in the judgment of Bostrom and others, for prioritizing efforts to prevent human extinction above other endeavors. This is what X-risk obsessives mean when they claim ending world poverty would be a “rounding error.” ⁶

Much as I find it morally troubling to ignore the suffering that occurs around me in favour of alleviating it overseas, I also find it difficult to accept that we should forgo assisting those here and now in favour of reducing the odds of future existential risk by (e.g.) .000001%.

4. Extreme philanthropy (Zuckerberg, Gates, Soros)

While I have huge admiration for the outcome-focussed ‘philanthrocapitalism’ of individuals like Bill Gates and Mark Zuckerberg, the encouragement that EA provides shifts our focus away from creating equal opportunity through structural reform, and therefore fails to address the very concept of global justice at its core. Moreover, the celebration of this ‘extreme philanthropy’, even in its most ‘effective’ data-driven form, is still a less effective way of limiting suffering than reducing inequality itself. In the classic ‘hand up’, not ‘hand out’ approach: economic development is the best form of aid. Unqualified celebration of extreme acts of giving, enabled only by massively unequal distributions of wealth, misses out on this fundamental fact.

Conclusions

I hope this overview has provided lots of food for thought. While I have my reservations, it’s still exciting to see the impact that evidence-based charity has had in the developing world, and now even in donor countries themselves; witness unconditional cash-grant programmes like GiveWell-approved GiveDirectly in Kenya, Bolsa Familia in Brazil, and basic income experiments in Finland and Canada. As Effective Altruism continues to gain momentum, more and more organisations will be called to innovate and improve, and to operate with greater transparency for those data-driven philanthropists who demand as such.

If you find this movement compelling, I encourage you to commit to donating at least 10% of your salary to effective causes over at Giving What We Can.

EA Links

Notable Individuals

William MacAskill

Toby Ord

Peter Singer

Organisations

80,000 Hours

Giving What We Can

The Life You Can Save

Centre for the Study of Existential Risk

Interesting Articles

The Trader who Donates Half His Pay

The Greatest Good

The Elitist Philanthropy of So-Called Effective Altruism

Could Effective Altruism destroy the arts?

You have 80,000 hours in your career. Here’s how to do the most good with them.

I spent a weekend at Google talking with nerds about charity. I came away … worried.

---

1. Wiki ↩︎

2. Singer, Peter (2014). The Most Good you can Do, p. 5. ↩︎

3. Anthropology has traditionally been critical of such metrics-first approaches, and often justifiably so. There is the danger that policies, frameworks, and flattering metrics are prioritised over the work of ‘aid’ or ‘development’ itself. GiveWell’s transparency demands aim to sweep away the obfuscation of bureaucracy, but certainly the international aid industry cannot claim to be anywhere near faultless. For interesting takes on some of these issues look to David Mosse’s 2011 Adventures in Aidland, or the more critical perspective in James Ferguson’s 1990 The Anti-politics Machine ↩︎

4. This example comes from Will MacAskill’s Doing Good Better; a highly recommended EA primer written by one of the movement’s founders. ↩︎

5. For those EA who believe in reducing total suffering, including that of non-humans, donating to campaigns that improve animal welfare is an area where a huge amount of good can be done for comparatively little money. ↩︎

6. Singer, Peter (2014). The Most Good you can Do, p. 27. ↩︎