Comparison: Standardization and Decomposition

Abstract

Demographers regularly wish to make comparisons, either between two or more populations at the same point in time (or thereabouts) or between the same population at two or more points in time. They frequently seek to do this by way of some single summary measure calculated for each of the populations to be compared. Comparisons of summary measures can, however, be quite misleading because differences between them reflect differences in population composition, as well as differences in the underlying level, incidence or prevalence of the demographic phenomenon under scrutiny. ‘Population composition’ refers to the proportionate distribution of a population across categories of variables, or combinations of variables, which capture demographically pertinent characteristics of individuals. Such variables include age, sex, marital status, ethnicity, education and occupation, but not, for example, variables like eye colour, or whether left or right handed. The most familiar representation of population composition is the age-sex pyramid, a graphical device for summarizing the proportionate distribution of a population by age and sex, several examples of which were presented in Chap. 1 to illustrate the capacity of census data to capture demographic change over time. Composition is, however, a multidimensional concept that embraces several other variables besides these two.