The paper takes as its basis the mortality aspect of population projections with particular reference to this aspect of the report by the Royal Commission on Population (Cmd. 7695), and the Reports and Selected Papers of the Statistics Committee published as Volume II of the Papers of the Royal Commission.

The alternative assumptions regarding future mortality adopted by the Statistics Committee for use in their published population projections and the effects of the assumptions on those projections are discussed. The paper considers also the generation approach, the ultimate irreducible level of mortality approach, and remarks are made on the analysis of mortality according to cause of death. Further, general consideration is given to a number of factors the precise extent of the influence of which on future mortality is unknown, to demonstrate the insufficiency of a single forecast of mortality when dealing with the problem of population projections.

The conclusion is arrived at that neither of the mortality assumptions adopted by the Statistics Committee was at the time an extreme assumption, but that from considerations of the mortality aspect alone this Committee extended their projections too far into the future. The opinion is reached also that a method which makes some allowance for a generational influence would have been a suitable alternative to the declining mortality method adopted, for obtaining a not extremely light estimate of future mortality. The ultimate irreducible level of mortality approach is concluded to be impracticable at the present time, although the view is formed that the results of analyses of deaths according to cause of death should be of value when considering future population mortality.