What this covers

Peter Singer and Russ Roberts debate the moral case for charitable giving in this conversation centered on Singer's book The Life You Can Save. Singer defends the claim that affluent people in developed countries have a binding obligation to donate significant portions of their income to relieve extreme poverty abroad, grounding the argument in a principle about preventing serious harm at modest cost to oneself. Roberts, hosting from an economics perspective, challenges whether distance, the limits of our knowledge about what actually helps, and the importance of human flourishing complicate or undermine that duty. The two move between philosophical foundations and practical complications, testing the coherence of Singer's position against real-world constraints.

Singer anchors his case in the Drowning Child Thought Experiment, arguing that if you would wade into a pond to save a child at the cost of ruined shoes, the geographical distance to a poor child in another country changes nothing morally. He marshals evidence that extreme poverty has dropped sharply and could be eliminated if the world's affluent adopted a progressive giving scale. On method, he embraces Effective Altruism and supports the use of randomized controlled trials to measure which interventions genuinely work—citing research by Esther Duflo, Abhijit Banerjee, and Michael Kremer. Roberts counters that Hayek's knowledge problem makes distant aid genuinely difficult to direct well, that much aid has failed or been captured by bad actors, that economic growth in China and India—not charity—drove poverty reduction, and that attempts to quantify human flourishing through economic metrics obscure what matters. The two also weigh whether earning to give in finance is ethical, how Evolutionary Psychology explains (but does not excuse) indifference to strangers, and whether open borders, though economically sound, risks political backlash that would entrench worse policies.

Sharpest takeaway

Singer argues that affluent people have a moral obligation to donate to effective charities to relieve extreme global poverty, because if one can prevent something very bad at small cost to oneself, failing to do so is wrong; Roberts presses that distance, knowledge limits, and flourishing complicate this duty.

  • The drowning-child thought experiment shows distance and lack of responsibility do not excuse failing to rescue at small cost
  • Extreme poverty has been roughly halved in a decade and modest progressive giving could lift everyone above the poverty line
  • Effective altruism demands measuring impact, not just signaling, though Roberts doubts much aid is measurably effective

The argument · threads5 threads · 28 claims
0.72

Preventing serious suffering at trivial cost to oneself is a binding moral duty regardless of distance or relationship.

4 pointscentrality 5/5
  • If you can prevent something really bad from happening without sacrificing anything remotely comparable in significance to you, then you ought to do it, and it is wrong not to do it; the death of a child is bad, so failing to rescue one at small cost (ruined shoes or its monetary equivalent given to charity) is wrong even if the child is distant and not your responsibility.

    “Essentially, the principle that I defend is if you can prevent something really bad happening without sacrificing anything that is remotely comparable in its significance to you--as the death of a child is for the child and for the parents and others who love and care for that child--then you ought to do it, and it's wrong not to do it.”

  • The fact that it is human nature not to care about distant strangers is a psychological explanation, not a moral justification; we cannot take refuge in 'it's human nature' to excuse failing to care, though it may bear on how harshly we blame people who behave as most people do.

    “We can't just take refuge in the idea, 'Well, it's human nature not to care about distant strangers.' That's not a moral justification for not caring about distant strangers.”

  • There are objectively good and bad things that can happen: pain and suffering are bad, happiness and pleasure are good, and dying without experiencing the good things one could is bad, as is the grief of the parents; this objective badness is sufficient grounds for caring and trying to prevent it.

    “I think there are objectively good and bad things that can happen. Pain and suffering are bad things, happiness and pleasure are good things... And, I think that that's sufficient to say why you should care.”

  • People who claim not to care about strangers are inconsistent because they resent it when others apply that same indifference to them; the reasons that exist for caring about a person in need apply equally whether the person is oneself or a stranger, so refusing to extend care to strangers is a failure of consistency rooted in evolved partiality.

    “most people resent that, or feel that there's something wrong with that... I think that suggests that they're being inconsistent when they consider their own case and when they consider the case of others.”

0.54

Evidence from randomized trials shows which poverty interventions actually work, making aid allocation increasingly knowledge-based and effective.

3 pointscentrality 4/5
  • Even if quantitative measures of well-being are incomplete and do not capture everything, they still add something to our knowledge and it is better to have that information than not to have it.

    “I agree that it's not complete. It doesn't provide us with everything we know. But I think it adds something to our knowledge. I think it's better than not having that information.”

  • The knowledge problem in aid is increasingly soluble because researchers like Duflo, Banerjee, and Kremer (2019 Nobel laureates) pioneered randomized controlled trials of poverty interventions, modeled on pharmaceutical drug testing against placebos or standard treatments, spawning a whole field that lets us increasingly identify which interventions actually help.

    “the Nobel Memorial Prize for Economics this year went to three researchers who have done just that... they have pioneered the development of assessing interventions to help people in poverty by the same kind of methods that pharmaceutical companies use to assess new drugs.”

  • Direct interventions are showing positive measurable results: GiveDirectly gives villagers in East Africa a guaranteed basic income supplement (around $12/month for 12 years) compared against control villages, with preliminary results positive and disproving worries that recipients stop working or waste money; Village Enterprise provides cash or assets plus business training and ongoing mentoring and group support, and worked well in five of six countries tested.

    “there's an organization called Give Directly... it was about $12 a month that they're giving people... they are guaranteeing it for, I think, it's 12 years. And that's being compared with other villages that are not receiving this.”

0.54

A progressive giving target of 1-33% based on income is pragmatically calibrated to generate large aggregate donations without deterring participation.

3 pointscentrality 4/5
  • Although a higher individual giving level might be ideal, a utilitarian recommendation should be judged by its consequences: setting the recommended giving level too high turns most people off, so recommending 5% that 100,000 people adopt yields a far greater total than recommending 50% that only 1,000 people adopt; hence the pragmatic 5% target.

    “if you make the level too high, you turn most people off... if I recommend 5%, and I'm lucky, maybe I'll get 100,000 people giving 5%. And obviously, that will yield a greater total than the smaller number giving 50%.”

  • Using a progressive 'giving table' that starts at 1% for moderate incomes and rises to 33% for those earning millions, set at levels imposing no serious hardship, the roughly one billion affluent people giving accordingly would raise several times the total of current foreign aid, more than enough to lift all 736 million people above the extreme-poverty line, even allowing for transfer costs and inflation.

    “when you total up how much this would raise if all the billion or so affluent people in the world were to give according to those levels: It's far more than the total amount of foreign aid given. It's several times the total amount of foreign aid given.”

  • When calculating how much to give as charity, the net cost of a tax-deductible gift is the amount given minus the tax saving, so one can sensibly compute percentages on gross income while accounting for deductions; Singer simply uses gross income rather than after-tax income for his giving percentages.

    “the net cost to of the gift is obviously the amount you give minus the tax saving that you get on it... I simply go by the gross amount rather than the after-tax income when I'm looking at the percentages that I give.”

0.44

Natural human partiality and political constraints limit what moral ideals can achieve, requiring reasoned compromises on borders and social standards.

7 pointscentrality 2/5
  • Although evolution gave us a nature that cares more for ourselves and close kin than for strangers, we also evolved the capacity to reason; once we have that capacity we can reflect on, criticize, and override our natural tendencies and develop ethical standards justifiable by reason that go beyond what is natural.

    “once you develop a capacity to reason, in some sense, you free yourself from these other natural elements, the psychology that is evolved through nature. And we have the ability to reflect on it, to become self-aware, to criticize the way that we have natural tendencies to act; and to develop different standards--which, I think are standards that can be justified in terms of reason.”

  • There is good psychological evidence that children grow up happier and more psychologically sound in a loving family with parents who truly care for them, while utopian attempts to provide institutional group care for children do not work well, so favoring those close to us is partly good and we have no good alternative to family at the moment.

    “there's good psychological evidence... that children are likely to grow up happier and more psychologically sound if they grow up in a loving family. The attempts to provide institutions, sort of utopian attempts to provide group care for children, don't seem to work that well.”

  • Although immigration is a powerful way to lift people out of poverty and Singer favors less restrictive borders, our evolved xenophobic streak means any democracy attempting large-scale open borders will provoke an electoral backlash; recent disasters like Trump's 2016 win, Brexit, and right-wing governments in Hungary and Poland were driven by this, so he reluctantly does not call for open borders because it would entrench far-right governments whose climate, trade, and social policies are disastrous.

    “any democracy that attempts to open its borders to immigrants on a large scale is going to get a backlash from the electorate.”

  • The desire to be part of respectable 'civilized company' motivates some right behavior but is an insufficient and conservative moral standard, because civilized company tolerates token charity and practices like eating factory-farmed animals; we must critically evaluate what standards that company accepts rather than treat its approval as the ultimate moral judgment.

    “The problem with that is, of course, that it's conservative... So, I don't think that's a sufficient motive. I think we have to be more critical about what civilized company accepts, what kinds of standards it tolerates.”

  • Political and trade structures harm the poor: enabling poor countries to keep foreign trawlers from fishing out their coastal waters, and ending rich-country distortions like subsidized American cotton being dumped cheaply, would let people earn livings as fishermen or cotton farmers—showing development is constrained by trade policy, not just charity.

    “Enable them to keep the foreign trawlers from fishing out of the areas off the coast where they're living, and then they might still be able to fish.”

  • Where people are suffering or their children are stunted by malnutrition, it is very hard for them to have flourishing lives, so relieving baseline suffering is a precondition for flourishing; only after getting people over that baseline can we focus on what makes people happy.

    “where their children are stunted by malnutrition, then it's very hard for them to actually have flourishing lives. So we need to get over that sort of baseline, and then maybe we can start thinking more about what makes people happy.”

  • Holding yourself up to public scrutiny—as Matt Wage did by publicly declaring he gives away large sums and asking others to call him a hypocrite if he slips into a luxurious lifestyle—is a good way to bind yourself to your principles, like binding yourself to the mast.

    “'And if you see me starting to slip in that direction I want you to call me out and tell me I'm a hypocrite.'”

0.40

High-earning careers and direct interventions both demonstrably save more lives than conventional moral assumptions suggest possible.

3 pointscentrality 3/5
  • Going into finance or high-earning sectors specifically to earn more and donate more can be a genuinely ethical career, because donating funds entirely new positions at nonprofits (potentially five new hires) rather than merely doing a marginally better job than the next applicant; people wrongly dismiss finance as unethical, but as Matt Wage's example shows, one can save more lives per year through giving than through direct rescue.

    “it gives you the opportunity to actually fund positions at nonprofits, so that rather than just taking the job yourself... you can give enough money to the nonprofit organization to employ an entirely new person.”

  • In the real world you should save the person drowning in front of you because you cannot know the future, but if a perfectly reliable prophet told you that saving this one person would prevent a successful career that would let you donate millions and thereby save thousands of lives, then it would be right to let the person in the pond drown, even though it would be emotionally difficult.

    “if somehow, some prophet who has proven totally reliable tells you that, if you save this person, you won't have this successful career... then I would say, you would be right to let the person in the pond drown.”

  • According to World Bank figures, the number of people living in extreme poverty fell from about 1.4 billion at the first edition of the book to 736 million by the new edition roughly 10-11 years later, nearly cutting extreme poverty in half.

    “In the first edition, I quoted World Bank figures... there are 1.4 billion people living in extreme poverty... In the new edition, the figure I quote... is 736 million. So, it's almost been cut in half in 10 or 11 years.”