Abstract A cost‐benefit analysis of an event must take account of the event's effect on population. Cost‐benefit analysts traditionally ignore these effects because they think that changes in the population are ethically neutral: neither benefits or costs. Although this view is intuitively plausible, it is false for theoretical reasons. There can be only one neutral level of lifetime well‐being. Adding to the population a person whose well‐being would be below this level is bad. However, this single neutral level might be very vague, which means that over a large range of levels of well‐being, adding a person at that level is neither determinately good nor determinately bad. This helps to restore the view that changing the population is generally ethically neutral. But neutrality of this sort turns out to have incredible implications for cost‐benefit analysis.