Abstract

Effective Altruism (EA) is predominantly driven by quantitative cost-benefit analyses, focusing on metrics such as scale, the number of DALYs (Disability-Adjusted Life Years) saved, or the sheer volume of direct cash transfers. While this top-down data is crucial for maximising reach, it frequently misses the textured, lived experiences of the aid recipients. My essay argues that to truly maximise impact and prevent localised program failures, EA must integrate rich ethnographic data. By understanding the human experience and local cultural knowledge, EA initiatives can increase empathy, adapt to unforeseen social dynamics and ultimately improve the efficiency and reach of their interventions. 

Introduction 

During my introduction to Effective Altruism, I learnt about the impressive quantitative measures that scholars have come up with to ‘help’ the largest number of people, most effectively and efficiently. While this framework sounded fantastic in theory, my training in social anthropology left me with a sense of ambivalence regarding the implicitly paternalistic assumptions embedded within it. EA is often orientated around the belief that certain projects are objectively the “most beneficial” and it is the job of the enlightened, “effective” altruist – armed with quantitative frameworks – to ascertain and implement these interventions. This top-down methodology rarely engages with the lived realities and complex social webs of the people on the receiving end of the aid. 

For example, a core principle of EA is the moral imperative to maximise cost-effectiveness in philanthropic work. According to thinkers like Toby Ord, this is a paramount value, and metrics like the DALY should be prioritised over traditional moral considerations such as justice or historical context. While I agree that philanthropic work should be subject to rigorous auditing, I worry that EA’s pendulum has swung too far toward pure quantitative analysis, neglecting crucial qualitative data. We are frequently told that organisations like GiveDirectly and their unconditional cash transfers are the most effective method for alleviating poverty. Yet, in these top-down analyses, we rarely hear from the recipients themselves regarding their concerns, ambivalences, or desires regarding the programs. 

The blind sport of “neutral” interventions

Integrating qualitative data provides a much richer understanding of how aid actually functions on the ground, sometimes challenging the triumphalism of purely qualitative interventions. Anthropologist Mario Schmitd’s research in the area surrounding the western Kenyan market town of Kaeko provides an interesting case study. 

If we look at EA metrics, unconditional cash transfers are often viewed as neutral technological fixes that bypass corrupt structures. However, Schmidt found that half the households targeted by GiveDirectly in this area refused to participate in the program. From a top-down, purely quantitative perspective, this high refusal rate might easily be dismissed as a technical glitch, backward “superstition” or the localised cynicism of people who do not want to be helped.

However, through ethnographic observation, Schmidt situated this rejection within local understandings of the economy as a deeply relational tool. GiveDirectly failed to understand that their willingness to hand out “free money” clashed with local cultural realities. For the residents, accepting a gift is not just taking something for free and maximising profit, but is entangled with local networks of redistribution and reciprocity. Because GiveDirectly did not embed the transfer into local networks of redistribution, the recipients conceptualised the money in various other ways. For example, some interpreted the transfers as a barter with the devil or an asymmetrical gift relationship. Others viewed it as political reparations stemming from Barack Obama’s Luo heritage, attempting to recoup losses caused by a corrupt elite. 

Rather than recognising this a social issue rooted in the paradox of the “free gift” which anthropologist Marcel Mauss outlined in 1925, GiveDirectly treated the cash transfers’s indeterminacy as a mere technical problem. Their outreach tactics failed to integrate cash traders into local understandings of economic relationality. 

We might also question whether the values that EA quantitative frameworks implicitly prioritise are the ones which most reflect the needs or desires of the people receiving the aid. By defining success primarily through the lens of atomised utility maximising, treating individuals as independent actors seeking to maximise personal profit, EA risks imposing Western economic assumptions onto vastly different cultural landscapes. 

This clash of values is illuminated by anthropologist China Scherz’s work in Having People, Having Heart. The very concepts of independence and individual cost-benefit analyses that are highly prized by modern development and EA frameworks are not universally viewed as moral goods. In Central Uganda, for example, local ethics often place a high value on interdependence and enduring relationships of mutual obligation. Where a Western NGO might view an ongoing patron-client relationship as a problematic form of “dependence” that needs to be eradicated by interventions and the encouragement of entrepreneurialism, local populations frequently view the push for self-reliance as a suspect refusal by the wealthy to redistribute their resources and fulfil their social obligations. 

When EA relies strictly on frameworks like the DALY or on the volume of unconditional cash transfers, it quietly champions a specific set of values: measurable short-term utility, atomised autonomy, and transactional detachment. Ethnographic data is a necessary magnifying glass, which allows those who are motivated by altruism to zoom in, to pause and ask whether their “optimised” interventions are inadvertently ignoring, or worse, severing, the very social ties that local communities rely upon for long-term resilience, identity and care. 

Conclusion 

What I am recommending does not neatly fit into EA’s standard Importance, Tractability and Neglectedness (ITN) framework. It is epistemologically difficult to quantify exactly how many lives will be saved or improved by hiring an ethnographer, or sponsoring anthropological research. However, utilising qualitative data remains highly neglected within the EA space, and it is entirely tractable to partner with anthropologists during the deployment of aid. If a program scales mathematically but faces a 50% refusal rate on the ground because it ignored local economic understandings, the intervention is no longer “effecitve”. Using ethnographic data will ensure that the ambitions and intentions of EA quantitative models are actually being realised in reality.