The standard view within effective altruism is that a person’s contribution to a particular focus area or intervention is determined by their marginal impact—that is, the impact that the contribution adds to the pre-existing impact.
The marginal impact should be clearly distinguished from the average impact of contributions to areas or interventions. If the marginal impact is diminishing—as it normally is—, the marginal impact is lower than the average impact, whereas if the marginal impact is increasing, the converse is true.
For instance, suppose that 1,000 people have contributed to a certain area, and that each of them have produced 10 expected units of value, but that the expected value of an additional person’s work is 5 units of value. Then the marginal expected impact of your work, which should determine your decision of whether you should work on the area, is 5 units of value. That is substantially lower than the average value of work on the area (including your work), which is close to 10 expected units of impact.
A corollary of the notion that you should try to maximize your marginal impact is that the area which is the highest absolute priority—i.e. which area should receive the most global resources—may not be the greatest marginal priority. If the absolute priority area is already heavily invested in, and if it exhibits diminishing marginal returns, other areas which are less heavily invested in may be higher marginal priority.
Library of Economics and Liberty. 2016. Margins and thinking at the margin.