ABSTRACT
The process of demographic measurement and modelling can be distilled into four steps. First, data are collected on people and events. Second, the data are put into standard formats. Third, demographic measures such as rates, probabilities, or means are estimated. And fourth, the measures are summarized, using indicators such as growth rates, population totals, or life expectancies.
Part I of the book focuses on steps 1 and 2. The two core concepts are demographic arrays and demographic accounts. A demographic array is a set of cross-tabulated counts, rates, or other demographic quantities. A demographic account is a collection of linked demographic arrays. The elements of arrays within an account conform to a set of fundamental accounting identities: for every subpopulation, population at the end of each period equals population at the beginning of the period, plus entries during the period, and minus exits during the period.
Demographic arrays and accounts are constructed from individual-level biographies. The construction process entails a shift in perspective, from individuals to groups. In real datasets, information on individual biographies is imperfect, so arrays have measurement errors and gaps, and accounting identities are not satisfied.
