Against Effective Altruism

February 2023

Effective Altruism (EA) as a philosophy and political movement has been gaining in popularity and scope almost exponentially for several years. The unexpected demise of Sam Bankman-Fried [0], one of the world’s most famous “Effective Altruists,” and his legacy have finally cast some scrutiny on a movement that otherwise was considered by the public to be an unmitigated good. After all, who can argue against altruism, and it being carried out effectively? 

Originally, I intended to criticize EA on strictly philosophical grounds, analyzing the core axioms of its philosophy and exposing their deficiencies, but the recent FTX scandal has warranted an expansion of this essay’s scope. I find it necessary now to also demonstrate EA’s formidable and dangerous potential as a political entity. A close inspection of Effective Altruism both as a philosophy and as a political entity reveals that not only does it fail to “do good better” [1], but also actively serves as a destructive global force.

In assessing EA’s philosophy, its axioms must be clearly enumerated. The first two are simple and uncontroversial:

1. Charities can do good in the world.
   
2. Some charities can do more good than others.

For most casual observers, EA’s principles stop here. In fact, effectivealtruism.org even claims this to be the limited scope of their philosophy by stating, “The only ethical position necessary for effective altruism is believing that helping others is important” [2]. If this were the case, there would be no need to assess the dangers of such a philosophy because one would be hard pressed to find them. But a third, fundamental axiom is missing. 

To derive it, one must look to the founder of the Effective Altruism movement, Peter Singer. Singer is a moral philosopher and professor at Princeton University. He planted the seed for EA with his seminal essay “Famine, Affluence, and Morality” [3], written in response to the starvation of Bengali refugees during the Bangladesh Liberation War in 1971 [4]. In it he argues that those in affluent countries have a moral obligation to donate to the poverty-stricken areas of the world, necessitating a flow of money from wealthy countries to poor ones. To make this charge, he divides the world into two groups: those who are dying and those who are not. He claims that  “if it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought, morally, to do it.” And so, with Singer’s help, the third, crucial axiom of Effective Altruism can be stated:

3. Money must flow from those who are not dying to those who are (i.e., rich to poor countries, developed to developing, USA to Bangladesh, etc.).

This third axiom is both crucial for defining EA’s philosophy, and for identifying its core moral flaws. With this third axiom unmasked, EA’s three fundamental shortcomings can now be assessed: its dubious effectiveness, its overly narrow optimization window, and the domestic expense associated with its exporting of empathy. Throughout the analysis, viable alternatives to EA, namely Local Altruism (LA) and Unquantifiable Altruism (UA), will be posited.

Is it Effective?

EA’s third axiom demands that organizations focus on global issues. There are a number of challenges presented by attacking problems existing on a global scale. Global system dynamics are incredibly complex, especially when compared to those of local systems. It is extremely hard, if not impossible, for EA organizations to understand and predict the long term externalities of charitable efforts in developing countries. Knowing they are tackling problems of such scale, they naturally target sufficiently small, roughly linear aspects of these issues. Under the assumption that we must tackle global issues, I agree that this is the most effective method. But in taking this approach, organizations are inevitably treating symptoms instead of causes. In “Famine, Affluence, and Morality,” Singer makes the famous drowning child analogy: “if I am walking past a shallow pond and see a child drowning in it, I ought to wade in and pull the child out. This will mean getting my clothes muddy, but this is insignificant, while the death of the child would presumably be a very bad thing.” But in treating symptoms due to the manifold complexities of global dynamics, one may save the drowning child only to place her right back on the water’s edge, waiting for her to fall in again, over and over.

Global systems also offer challenges in terms of feedback mechanisms. It is much harder for a donor to track the outcomes of his money being spent in the far reaches of Sierra Leone than in his backyard. Admittedly, almost all of the major Effective Altruist organizations today have detailed overviews of whether or not their efforts are working [5]. But one is still left to trust an organization’s public relations team, with all of its associated biases, for feedback. It is not hard to conceive of an organization feeling pressure to report positive results to the very donors who are keeping the entire operation solvent. This feedback, even if accurate, remains incredibly abstract and theoretical to the donor. He receives this information via a computer screen, not through the lived experience of his own life.

Additionally, an organization may be effective today in making progress toward achieving its far-reaching fifty-year goal, but in the long term, it may be pulled in one direction or another due to competing interests, fundraising efforts, management decisions, bureaucratic inertia, or bankruptcy, leaving an incomplete project with no accountability whatsoever. And are we to expect the donors, with their money flowing to different organizations with each passing year, to keep tabs on the continual effectiveness of dollars they spent decades ago? The very definition of monetary charity implies some sense of “give and forget” on the donor level.

When stepping back further, one may ask how even to define “effective” in this context. Here, one is confronted with Effective Altruism’s inherent adherence to utilitarian philosophy. There is heated debate in the community as to whether or not EA is analogous with utilitarianism [6]. But the third axiom of EA, sending money to poor countries to save the lives of those who are dying, seems to be congruent with, if not identical to, the utilitarian goal of maximizing human welfare. If this is so, one can leverage the same attack on EA as the many well-established [7] ones against utilitarianism. This essay does not intend to indict utilitarianism, and so I will leave this line of thought up to the reader.

Regardless, a utility function must be set, and by donating to Effective Altruist causes, the donor relinquishes this decision to current EA leaders. Instead of deciding what is to be optimized for himself, he effectively chooses by proxy. This results in a sort of opaque aristocracy, where a small group of elites can utilize any means necessary to achieve its utilitarian ends. Singer states that we must save the lives of those in starving countries so long as we don’t sacrifice something of greater moral value. But ceding power to this elite to decide what is of greater moral value is a dangerous proposition, especially when Singer himself makes suggestions such as, “From the moral point of view, the prevention of the starvation of millions of people outside our society must be considered at least as pressing as the upholding of property norms within our society.” What are we willing to sacrifice to achieve EA’s utilitarian ends? Property rights? Free markets? Constitutional law? Democratic systems of governance? These are not trivial questions.

An alternative to EA’s global approach is Local Altruism (LA), which aims to solve local problems in one’s own community. In this way, one operates on an inherently smaller and simpler scale due to the locality of the issue, whose root cause is more likely to be roughly linear in its complexity and therefore able to be directly engaged. Take for example the complexity of getting the three homeless men in your neighborhood treated for their mental conditions, housed and employed versus attacking malarial death in an entire continent. Setting an example as a Local Altruist can also scale to have the emergent property of solving issues on a global level, for if the idea propagates globally, people in every community can orient themselves in a locally altruistic way.

Furthermore, an individual who not only donates money but also time and effort to resolving the issues of his locality can see every day the result of his efforts. Over the course of decades, he can constantly adjust his actions due to feedback received from his lived experience. The direct, donor-level feedback mechanism of Local Altruism is superior to that of EA, especially as timescales grow.

Narrow Optimization Window

How can one compute the value of improving things locally that aren’t as empirical as saving the life of a child in Sierra Leone? Effective Altruists tend to fetishize quantifiability and therefore fall victim to overly narrow optimization windows. This is why the success of so many EA efforts is measured by the “number of lives saved” [8]. Donors fail to even consider that there could be another metric to optimize for. There are myriad ways that one can commit acts of service for his family, church, or neighborhood that yield unquantifiable benefits. For how can one quantify the benefit of constructing an outdoor amphitheater for one’s church? Or for cooking meals for a sick neighbor? Is it not possible that the immeasurable, diffuse benefits associated with this Unquantifiable Altruism (UA) do not, in the long term, outweigh the benefits of Effective Altruism? Unquantifiable Altruistic efforts are inherently local as well, giving rise to compounding benefits.

An Effective Altruist may respond to my call for UA with the question, “Why not both?” [9]. The answer is simple: many people cannot afford to discretize and therefore need to make a value judgement about how to spend their time and money. I’ll use a friend as an example. For the sake of anonymity, let’s call him John. John spent a month trail building last year for a summer camp. He made $4,300. $3,300 he spent to send his best friend’s cousin to camp (age fourteen, who lived in a trailer with a single mom suffering from opioid addiction and had not been to school for several years). On top of John’s efforts, his best friend took in her cousin and is putting her through school. Through their combined efforts, local and embedded in a sense of community, real change was made in this girl’s life. John did not have enough money to discretize between different forms of altruism across the spectrum. He values locality and solving issues close to home, and values his tribe more than a dying one halfway across the globe.

John also thinks that working at camp was a valuable form of service since it resulted in two new trails, allowing campers to experience the thrill of mountain biking. This flies in the face of EA’s “earning to give” concept, which demands an individual seek highly paid labor in order to give lots of money to an effective charity. Therefore, John makes two value judgments that can’t be squared with Effective Altruism: 

1. He works on an “ineffective frontier” from a monetary standpoint (according to EA), but thinks that the value of his labor transcends the dollar value associated with it, in the form of getting children interested in mountain biking.
   
2. He donates his money to a local cause in his community (supporting his friend’s cousin), one he can track progress on and has a vested interest in seeing succeed, but according to EA metrics, is also on the ineffective frontier.

To scale this concept, let’s take a family with two children making good money, $150,000 annually. After expenses like taxes, food, mortgage, debt payment, gas, and saving for retirement, the parents may have 10% of this gross income, or $15,000, leftover. They want their kids to attend private school, or perhaps college. So they put the rest of their money into a 529 or related fund. They cannot afford to discretize so easily and must make a value judgement. What they choose is to value their children’s future over that of a faceless recipient of funds in a different country.

Now imagine a more average income of something like $35,000 annually (a demographic of earners that also has some of the nation’s highest birthrates [10]); it is obvious that it is much less tenable for this economic class to have the luxury of donating across the spectrum while teetering on the brink of financial ruin at the slightest economic perturbance. This is likely why EA largely attracts well-educated, childless individuals with high income. This demographic is more likely to have the luxury of donating across the spectrum.

How does the Effective Altruist assess the value of John using his money to send a friend’s cousin to summer camp? Or of John building a mountain biking trail for campers? Or of parents paying for private school or college for their child? Or donating to a concert hall or opera house in one’s town? Singer clearly shows his opposition to these forms of altruism by stating that “Australia’s aid [to the Bengali refugees in 1971], however, amounts to less than one-twelfth of the cost of Sydney's new opera house. The total amount given, from all sources, now stands at about £65,000,000.” 

Perhaps these diffuse benefits could save thousands of lives if a child decides at summer camp to turn her life in a particular direction, or someone is inspired by an orchestral performance at his local concert hall. These are unquantifiable benefits, and therefore unattractive, if not altogether invisible, to the Effective Altruist. Unsurprisingly, in an age that worships only that which can be measured, observed, studied, and quantified, Effective Altruism suffers from an overly narrow scope of optimization that limits its ability to see the diffuse, but very real, benefits of Unquantifiable Altruism.

Exporting Altruism Comes at a Local Expense

Altruism abroad comes at a local expense. An obvious reason for this is that money is sent abroad that could otherwise have been spent locally. But there are slightly less obvious, though far more damaging, externalities associated with exporting altruism. 

Humans are fundamentally tribal beings. Singer refutes this immutable law of human nature with the claim that “The fact that a person is physically near to us, so that we have personal contact with him, may make it more likely that we shall assist him, but this does not show that we ought to help him rather than another who happens to be further away. If we accept any principle of impartiality, universalizability, equality, or whatever, we cannot discriminate against someone merely because he is far away from us (or we are far away from him).” Though one may agree with Singer on theoretical grounds, one cannot help but feel our tribal nature experientially when thinking about the meaning derived from our own children or mother or best friend as opposed to a stranger halfway across the world. We place greater value on those close to us; thus, by our very nature, we organize our relationships in a hierarchical framework that begins with our family, followed by friends, neighbors, those in our town, county, state, country, and then other countries. Returning to the pond analogy, the Effective Altruist would rather walk past ten kids at home who are not drowning, but are teetering on the edge of the pond, about to fall in, and instead go across the world to pull out fifteen Africans who are actively drowning.

The very existence of nations demonstrably proves this law of nature. For what would be the point of a nation if we truly embraced Singer’s principles of “impartiality, universalizability, [and] equality?” Why have we not realized the utopian ideal of the Global Village? Our tribal nature stands in the way. And so long as the nature of humanity does not change, there will remain intractable world views that preclude the Global Village from materializing. For its inception, one must trivialize everything that distinguishes a Salafi Muslim from a Fundamentalist Mormon or a young, secular professional in San Francisco, which would be unethical and probably impossible. When we grasp the importance that locality has on the psyche, we can see an enormous benefit emerge from Local Altruism, namely that it imbues the donor with a deeply natural and spiritual satisfaction. Not only is this benefit entirely absent from Effective Altruism, but its philosophy damages donors and their communities in various other ways that will be explored below.

There is no doubt that many people donate to feel the high of a good deed. This is a fleeting high that lacks substance in most cases, but one cannot deny its existence. This describes why no sane person literally lights his money on fire, but so many people effectively burn their dollars by donating to random charities at the grocery store or online. Effective Altruists seem to look at this as a bug, not a feature, which is why many high profile Effective Altruists like Sam Harris and Will MacAskill tout the fact that their donations are set up to disburse automatically without any thought on their part. And while Harris and MacAskill may not feel a big empathy tug from the act of donation, we must not focus on the leaders of the movement but rather the average Effective Altruist.

Earning to give as a concept is targeted at a young, highly educated cohort and demands that one seek a well paying job (like a lawyer or software engineer) in order to maximize income to donate to effective causes. As universities rapidly descend from institutions that educate well-rounded citizens, to ones that churn out highly specialized, yet intellectually one-dimensional automata, Effective Altruism is filling a sort of niche for this demographic. The young and educated of the West find themselves in high-paying, yet morally vacuous jobs, living in unnatural and dehumanizing metropolitan settings, feeling isolated and metaphysically bankrupt. EA provides a sense of moral and metaphysical grounding to this cohort. It justifies their unsustainable, unfulfilling, but high earning lifestyles with the “earning to give” concept. Donating money, which isn’t a true virtue on its own, becomes a comprehensive means of existence to the Effective Altruist. An entire ontology is created around the movement, which partly explains the religious zeal of its followers [11]. While EA leaders like MacAskill and Harris may not feel a big empathy tug, many average EA followers receive not only a high but also an entire justification for their existence by adhering to the philosophy. Singer corroborates this disturbing concept by claiming that an individual “who [donates to alleviate famine] will have to sacrifice some of the benefits of the consumer society, but he can find compensation in the satisfaction of a way of life in which theory and practice, if not yet in harmony, are at least coming together.” Ironically, EA forces its followers to contribute to and engage in the most extraordinarily materialistic areas of our “consumer society” in order to maximize profit, while alleging that a deep moral “satisfaction” will accompany this lifestyle through the act of donation.

Let us assess the philosophy’s impact on a single individual. A woman subscribing to EA pursues a career as a computer scientist in order to maximize her donation potential. She puts in sixty-hour work weeks, contributing to a mobile application designed to cognitively disarm users and sell them advertisements. She delays starting a family of her own, and is too burned out for any sort of community engagement in her time off. She is required to unnaturally discount the value of her immediate community in favor of utilitarian goals abroad. 

One must ask what effect this scenario may have on the giver and the community she lives in. What is the negative value associated with the degradation of her quality of life and, subsequently, her community? You cannot argue that this lifestyle, all but demanded of EA, if followed honestly, is not degrading to the giver’s milieu. One must look not further than the infamous Sam Bankman-Fried, recently the movement’s archetypal ideal [12], for an idea of the philosophy’s real effects on its adherents. He is obviously by no means an average Effective Altruist but he nonetheless embodies the ethos.

The inorganic lifestyle imparted on the individual by EA robs him of the deep metaphysical satisfaction that is achieved through small but consistent acts of service. The empathy tug of Local Altruism is much stronger than that of EA, and provides real, lasting value to the donor. LA has an outsized impact on the donor since he gets the internal fulfillment of feeling deeply connected and responsible for his immediate surroundings. A fundamental fulfillment can be found in serving one’s own tribe instead of one a world away. It imbues the donor with a real sense of mutual responsibility. He can implement, see, and ensure that his actions are having the desired effect on his community. It motivates the giver to make certain that not just his money but, more importantly, his time and labor are actually helping his countrymen. Daily, he will receive feedback, and his dynamic efforts will make lasting change. This provides both the recipient and the benefactor with real positive value. It is impossible to achieve this benefit with Effective Altruism’s cool rationalism and extreme distance between its benefactor and recipients of aid. On top of this, the Local Altruist does not have the moral obligation to pursue a morally vacuous, high-paying job. He can choose work with implicit meaning and congruence with his larger moral outlook. EA’s local cost is obvious when it demands its adherents adopt a spiritually destructive career outlook, and it robs them of the nourishment associated with seeing the tangible effects of one’s local altruistic efforts.

This points to a more existential risk that our society faces, one that Albert Camus famously coined “The Human Crisis” [13]. We are increasingly alienated from our communities and our own humanity. The modern world has created a mechanistic and inhumane environment dominated by an overly technical way of thinking. Even the very human act of charity has been reduced to “a research field” [14] by Effective Altruism. And so we are encouraged not to give to our communities, nor to work on first-order problems that are right in front of us, but instead to add on layers of technique and abstraction. To work high-paying jobs, to give that money to an organization which in turn will rationally compute the most “effective” means of helping another group far away, and impart this technique on said group. Camus chillingly elucidates this technical process at large by stating that “In sum, one no longer dies, one no longer loves, and one no longer kills, except by proxy. This, I suppose, is what is called good organization.” 

Summary of Local and Unquantifiable Altruism’s Benefits

Local and Unquantifiable Altruism give the donor back his humanity. They encourage him to live directly for others, not by proxy. And so we see the philosophies of Local Altruism and Unquantifiable Altruism having the following benefits over Effective Altruism:

1. Confront inherently simpler challenges, allowing donors and organizations to attack root causes of more linear systems instead of symptoms.
   
2. Receive direct and long-term feedback at the donor-level.
   
3. Set examples that can propagate and scale.
   
4. Encourage career and life outlooks that are not inhumane or degrading to the individual and his community.
   
5. Receive deep metaphysical fulfillment through servicing one’s own locality, with the acknowledgment that human nature is fundamentally tribal.
   

EA as a Political Entity

Beyond its philosophical shortcomings, Effective Altruism has grown and morphed into a dangerous political entity. Even back in 1971, Singer wished to extend the scope of the movement far beyond the will of the private citizen, asserting that “that giving privately is not enough, and that we ought to be campaigning actively for entirely new standards for both public and private contributions to famine relief.” Contemporary advocates of the movement are also strong advocates of governmental support for their causes. One must consider if the majority of the modern west holds the same view as this privileged elite. In the United States, an already sagging trust in institutions has been further crippled by theatrical elections and opaque responses to pandemics. It would be hard to imagine a population consenting to an increase in taxes or equivalent drop in public goods and services provided in order to support utilitarian schemes abroad conceived by this same elite. And while broad swaths of the nation would be opposed to EA’s expansion into the political realm, it is easy to see why many of the rich and powerful rave about the brilliance of the movement—it furthers their goal of consolidating power into the technocratic apparatus.

In practice, the Effective Altruist elite has created a predatory scheme to support its efforts, which are ultimately self-serving. The activist class targets high earners as an income base for their organizations. Individuals donate to these charities that are vetted and “certified” by the very same activist class that benefits from these donations. An entire industry has been created around the idea of giving, and it has grown exponentially under the guise of “rational choice.” Coincidentally, this concept of “rationality” is extraordinarily appealing to the ever-secularized cohort of students churned out by our universities. This activist class cannot and will not advocate for the end of a consumer society, the abolishment of the rich, the effort of individuals spent locally, or the pursuit of metaphysically meaningful careers because all of these undermine their bottom line, and would ultimately bankrupt their scheme. The FTX collapse verifies all of these claims about the corruption and danger of Effective Altruism.

Even if practiced perfectly with no bad actors, the Effective Altruism movement is morally flawed and on the ineffective frontier for doing good in the world. Unfortunately, but not surprisingly, the movement is far from flawlessly executed, and has become a dangerous political entity rife with the same corruption and evil we bemoan our corporations for perpetuating. At least corporations are honest in that their ultimate goal is to make money. Effective Altruism is a greater evil in that it acts under the auspices of a greater good and a moral duty, thereby coercing millions to fall in rank.

It does not have to be this way. Local and Unquantifiable Altruism offer true alternatives to Effective Altruism that don’t rely on a predatory governing body or large sums of money. They are forms of altruism that can organically propagate and scale. Instead of maintaining starry-eyed visions of the illusory Global Village, we can actually create a functioning set of global villages, plural, which I do think is a tenable future.