What the age composition of migrants can tell us.

One of the central problems in migration measurement arises as a consequence of the different sources of migration data. Most information regarding migration is obtained from population censuses or population registers that report migration data for a given time interval. Another source of migration data is the sample survey which may be designed to provide information about both migrants and moves. Migration data produced by censuses are usually in the form of transitions. Population registers treat migration as an event and generate data on moves. A fundamental aspect of migration is its change over time. This paper aims at demonstrating how the regularities that appear in migration age compositions can be summarized in a useful manner and to suggest what such regularities may be telling us about patterns of natural increase, family relationships, and mobility levels among migrants. By disaggregating migrants into dependent and independent catagories, the paper illuminates the ways in which the age profile of migrating populations is sensitive to relative changes in dependency levels and in rates of natural increase and mobility. Just as population age compositions reflect particular characteristics of fertility and mortality regimes, so do observed migration age compositions reflect key aspects of family structure and migration patterns. Drawing on techniques used in the analysis of influence on fertility and mortality, Rogers and Castro propose procedures for adopting model migration schedules to infer migration patterns in the absence of accurate migration data. Such model migration schedules may be used to graduate inadequate data, thereby smoothing out irregularities and ascribing to the data summary measures that can be used for comparative analysis. They also may be used to interpolate to single years of age. Observed reliability of empirical migration data and indications of appropriate strategies for their correction are aided by the availability of standard families of migration schedules. Finally, such schedules may also be used to help resolve problems caused by missing data. Examples of the age-sex distribution and interstate migration are given for Mexico and Sweden; model migration proportion schedules are given for selected cities including Rio de Janeiro, Mexico City, Lagos, Greater Khartoum, Stockholm, and Kuwait.