A Journal of Philosophy, Applied to the Real World

Being Good in a World of Uncertainty: A Reply to Temkin

Centre for Advanced Studies Justitia Amplificata


This reply affirms Temkin’s critical perspective on effective altruism but seeks to draw out its constructive implications. It first encourages Temkin to defend the practical urgency of global poverty in the face of doubts about aid effectiveness. It then argues for a more holistic conception of effectiveness to mitigate these doubts. It considers some alternative aid strategies that respond to this broader conception. Finally, it exhorts effective altruists to think more seriously about the reform of global institutions.


Temkin’s critique of effective altruism stands apart from others in at least three important ways. First, it issues from someone whose commitment to the core tenets of effective altruism is beyond dispute. Second, while other philosophers have fretted about the demands that effective altruism makes on altruistic agents, Temkin helpfully redirects our attention to the effects of the movement on its intended beneficiaries. Third, the account is impressively comprehensive: it illustrates in great detail the challenges that foreign assistance poses to a wide range of values, including underappreciated dimensions of political morality like voice, autonomy, and respect.1 Temkin’s is also the first critical perspective that I myself have found thoroughly convincing.

Like Temkin, I believe that global poverty makes considerable demands on those it spares, but what those demands actually amount to in practice is really quite murky (Lechterman forthcoming). Here, I want to explore some potential practical upshots of this conclusion. If one accepts that supporting foreign assistance initiatives leads us into a moral minefield, how can we best avoid, or navigate through, this dangerous terrain?

1. Why Poverty?

Some effective altruists might see the challenges of mitigating global poverty as the final nail in the coffin for this “cause area.”2 For many effective altruists, devoting resources to global poverty has become a hard sell even in the absence of Temkinian worries. This is because of the growing awareness of other catastrophes, which strike some as more morally urgent or at least more tractable. Measured against the importance of preventing the extinction of humanity, warding off the annihilation of the Earth, or reducing the horrific mistreatment of nonhuman animals, the misery of a “mere” two billion poor persons looks to some like a “rounding error” (Matthews 2015). Those who pair doubts about the relative significance of global poverty with emerging skepticism about strategies to combat it may be convinced to abandon this problem entirely.

We’d do well to pay attention to other important social problems, particularly ones that have suffered from neglect. But I suspect that Temkin would strongly resist the implication that we should thereby turn our backs on the global poor. And one thing we might ask of Temkin is a clearer justification for why global poverty should remain an urgent priority in light of these challenges. It’s not obvious to me that Temkin’s pluralist view of practical rationality can successfully supply this justification. For Temkin, acting well requires recognizing and responding to a host of virtue-based, agent-relative, and agent-neutral considerations. One who accepts this picture might still maintain that the agent-neutral reasons to mitigate existential risk are so compelling that they swamp the lingering virtue-based and agent-relative reasons to assist the global poor.

Defending the priority of global poverty in the face of competing causes remains easier for those who view our relationship to the needy through a theory of global justice, which mediates our duties to distant others. Take the fact that we participate in, and help to sustain, a global order that unfairly benefits us. Take also the fact that our prosperity depends in nontrivial ways on past wrongs that cast a long shadow (conquest, colonialism, exploitation, and so on). These facts help to generate stringent, agent-relative duties to the global poor that we don’t have to other potential targets of assistance.

2. Defining Effectiveness

Assuming the challenge of defending the priority of global poverty can be met, as I believe it can, let me turn to the question of how one might navigate the obstacles to addressing it.

Temkin argues that global poverty relief can raise each-we dilemmas, where what each of us has most reason to do (support the most demonstrably effective aid agencies) conflicts with what we all together have most reason to do (support the long-term reduction of poverty rates). I agree with Temkin that conflicts between individual and collective responsibility are key to understanding this problem (and many others). The arguments that aid initiatives undermine development are powerful. And even if these arguments prove faulty, I believe that reasons to help people directly will conflict with justice-based reasons to support the development of decent and stable institutions. Nonetheless, I also wonder whether looking at this problem from another angle might help us find ways around it.

As Temkin points out, most effective aid agencies either provide humanitarian relief or engage in discrete development projects without accounting for how these efforts will affect a country’s institutions, especially over the long term. But perhaps the real problem here lies in the definition of effectiveness. Temkin appears to assume a particular definition of effectiveness that has been common among effective altruists. This definition of effectiveness identifies it with measured impact. To be an effective organization, according to this way of thinking, one must be able to demonstrate sizeable impact on some measurable social indicator. The most reliable way to measure impact is through randomized controlled trials, and this explains why RCTs are a coveted source of information for rating agencies like GiveWell. But, as social scientists continually warn, randomized controlled trials are a limited tool (Clough 2015). They can only track certain kinds of interventions, and they can’t measure systemic or long-term effects. Organizations that score highly against these limited criteria may be counterproductive in the long run.

Consider an analogy. From a certain point of view, a diet of simple sugars looks remarkably effective. It appears to ramp up energy and trigger sensations of satisfaction. And if the tests we use can only measure local, short-term effects, sugar will look like the most effective form of nutrition around. But, as we all know, sugar’s effects are ultimately short-lived, and a diet high in sugar will destroy the body over the long term. Using the logic of practical dilemmas, we could conclude from this that our reason to consume an effective diet in sugar conflicts with other reasons we have to protect our health. But it would be more natural to say that a sugar-rich diet simply isn’t an effective form of nutrition in the first place.

The lesson here is that we can mitigate the apparent each-we dilemma in poverty relief by broadening the criteria we use to define and assess effectiveness. We shouldn’t consider an organization effective simply by virtue of its demonstrated ability to improve QALYs (or reduce DALYs). Rather, we should assess organizations on an array of criteria that also track relationships to broader development goals. Were we to do this, we’d probably encounter trade-offs between progress on different dimensions of assessment. Sometimes, we might judge that an organization with outstanding effects on QALYs deserves support despite concerns about its long-term impact. Other times, we might judge that the predicted improvements in QALYs aren’t worth the risks that a given intervention poses to institutional development. But we can’t make these kinds of comparisons without a richer understanding of effectiveness. And because existing data and methods make it difficult to measure and compare systemic effects, what one has most reason to do at this very moment may be to support further research into making these kinds of holistic evaluations.

3. Power, Autonomy, and Respect

A related problem that Temkin exposes is the undue power that comes along with foreign assistance.3 Even when they are well-intentioned and sensitively administered, funds from abroad create various pressures that tend to distort community priorities and behavior. Here the problem isn’t so much that philanthropy from abroad will interfere with long-term institutional progress (though it may). Rather, it’s that resources from abroad can create or reinforce objectionable social relationships. The existence of large sums of money creates incentives for individuals and communities to abandon careers, lifestyles, and policies that they may have preferred to pursue. To access or maintain the flow of benefits, receiving communities face pressure to ingratiate themselves and genuflect to the whims of donors and their agents. And interactions with well-heeled donors and staff members make receiving communities vividly and bitterly aware of their disadvantages. Feelings of powerlessness, humiliation, and disrespect seem not to resonate strongly with many members of the effective altruist movement. But I think it’s important to recognize that these experiences can be just as, if not more, disabling to people than poor physical health (Deveaux 2015).

These kinds of considerations have led some to support unconditional cash transfer programs. Because unconditional cash transfers limit the ability of donors to control and intrude, they appear to minimize relationships of domination and subordination. Cash transfers have much to recommend them. But they are no panacea. Aspects of domination and subordination can easily creep in when there are decisions to make about who receives transfers. Cash transfers can’t solve pressing problems that really do require technical expertise and specialized equipment, as in medicine and infrastructure. And foreign-funded cash transfer programs pose a clear threat to political participation and government accountability. If the needy can rely on funds from abroad, they have fewer incentives to make demands on the state.

So, it may be worth considering some additional ways of rendering assistance that acknowledge the values of autonomy and self-respect. Certain kinds of participatory organizational structures that give beneficiaries a voice in decision-making are one option (Krasner and Weinstein 2014). Another option is to treat foreign assistance projects as temporary demonstrations that will be handed over to local control after a given period of time (Reich 2016). A third possibility is to provide support for community organizing, which involves helping communities to identify shared interests and overcome collective action problems on their own (Stout 2010). Community organizing is attractive because it involves minimal outside interference, it aims for maximal inclusion, and it can foster the civic virtues needed for effective political participation.

4. Reforming the Global Order

As I suggested above, the global order bears a great deal of causal responsibility for the persistence of severe poverty. Even if one rejects the deontological arguments for reforming the global order, one must still recognize that the most sweeping improvements in poverty rates would most likely come from changes to international rules. Consider international laws that grant dictators property rights in natural resources (Wenar 2015), intellectual property rules that prevent poor countries from producing essential medicines (Pogge 2009), migration policies that limit the movement of labor, or trade policies that disadvantage farmers and fledgling industries in the global South (Risse 2012). These are just a few of the ways in which rich countries collectively exploit poor ones, with dire consequences. These are also massive, complicated, and controversial problems, which may help to explain why effective altruists, in search of opportunities for concrete progress, have been drawn to alternative paths. But suppose that effective altruists were to coordinate and consolidate their efforts for a few years to focus on a single aspect of the global order, one where reform might be feasible. As the movement grows in its size and in its collective wisdom, so does its potential to catalyze institutional change.

Reforming international institutions has its fair share of drawbacks. Chief among these is that the benefits it would generate would likely materialize only in the future, to the neglect of those who are suffering now (Cordelli 2016). And one might believe that the prospects of generating the collective will necessary to influence international rules are simply too dim to be worth pursuing. But it remains striking that effective altruists are so quick to embrace other complex global challenges like the risks from artificial intelligence and asteroid collisions, and so leery of thinking boldly about poverty.


Clough, Emily. 2015. “Effective Altruism’s Political Blind Spot.” Boston Review, July 14.

Cordelli, Chiara. 2016. “Reparative Justice and the Moral Limits of Discretionary Philanthropy.” In Philanthropy in Democratic Societies, ed. Rob Reich, Lucy Bernholz, and Chiara Cordelli, 244–67. Chicago: Chicago University Press.

Deveaux, Monique. 2015. “The Global Poor as Agents of Justice.” Journal of Moral Philosophy 12, no. 2, 125-150.

Krasner, Stephen D., and Jeremy M. Weinstein. 2014. “Improving Governance from the Outside In.” Annual Review of Political Science 17, 123–45.

Lechterman, Theodore M. Forthcoming. “The Effective Altruist’s Political Problem.” Polity.

Matthews, Dylan. 2015. “I Spent a Weekend at Google Talking with Nerds about Charity. I Came away…Worried.” Vox, August 10.

Pogge, Thomas. 2009. “The Health Impact Fund: Boosting Pharmaceutical Innovation without Obstructing Free Access.” Cambridge Quarterly of Healthcare Ethics 18, 78–86.

Reich, Rob. 2016. “Repugnant to the Whole Idea of Democracy? On the Role of Foundations in Democratic Societies.” PS: Political Science & Politics 49, 466–72.

Risse, Matthias. 2012. On Global Justice. Princeton: Princeton University Press.

Stout, Jeffrey. 2010. Blessed Are the Organized. Princeton: Princeton University Press.

Wenar, Leif. 2015. Blood Oil. Oxford: Oxford University Press.

1.Here I’m mainly referring to Temkin’s forthcoming book, which expands on the ideas published in this journal.

2.Effective altruists often refer to competing objects of beneficence as “cause areas.” See, e.g., www.causeprioritization.org

3.This, too, is something Temkin addresses more elaborately in the larger project from which this article draws.