If you came upon a small child drowning in a pond, you ought to save the child even at considerable cost and risk to yourself. In 1972 Peter Singer observed that inhabitants of affluent industrialized societies stand in exactly the same relationship to the millions of poor inhabitants of poor undeveloped societies that you would stand to the small child drowning in the example just given. Given that you ought to help the drowning child, by parity of reasoning we ought to help the impoverished needy persons around the globe. To capture this intuition Singer proposed this principle of benevolence: If one can prevent some significant bad from occurring, without sacrificing anything of comparable moral importance, one ought morally to do so. Premature death caused by preventable disease, injury, and poverty is uncontroversially a significant bad. Donations to charitable organizations such as Oxfam can prevent many of these deaths around the world, so Singer's principle holds that we ought to donate (or take some action that is comparably efficient at saving lives).

But this principle of benevolence is far more stringent than common-sense opinion, for even after one has donated most of one's income each month to world poverty relief, one could still donate more, and should do so according to the principle. For after all, the further reduction in one's available spending money does not incur anything that is comparable in badness to the loss that occurs to those in need of charitable relief if one's extra monthly donation is not forthcoming.