Your search found 1 Results
New York, New York, United Nations, 1992. vii, 46 p. (ST/ESA/SER.A/127)Methods pertaining to the preparation of migration data for subnational population projections as of 1992 are explained. A brief review of sources of data for migration projections (censuses, surveys, and registration data) reveals that the requirements are base period estimates of the level or rate of migration between regions, estimates of the age and sex distribution of migrants, and any indicators that show likely future trends. In a discussion of the measurement of the volume of migration from census date, data on residence at a fixed prior time, estimates based on previous place of residence and duration of residence, and estimates of net migration of census survival/ratio methods are relevant. Estimates of the distribution of migrants by age and sex are explained based on different age and sex data: on place of residence at a fixed prior date, on place of previous residence and duration of residence, on age distributions from surveys, and from registers. Also explained is the use of model migration schedules when there is little or no information about age. Baseline migration projections for future estimates which are reasonable and account for variable rates of migration by region are discussed. The objectives desired are sometimes contradictory in that using a long time frame in order to average out random or abnormal fluctuations conflicts with continuing recent nonrandom or unusual changes so that emergent trends will be projected; objectives are also to use the most recent data available which account for shifts in migration patterns and to ensure convergence of migration rates toward equilibrium at some future point. Alternative strategies are provided as well as adjustments to provide consistent results. Adjustments involve the projection of numbers of migrants rather than rates, the use of out-migrant data on destination to adjust in-migration, and the scaling of in-migration to equal out-migration. Recommendations for data collection are presented. Internal migration data are best served by census data which asks the question about place of residence at a fixed prior time preceding the census and with a time interval designation that is of interest for projections. Single year of age and prior year questions and 5 years before are desired due to the need for short-range projections and planning. The 5-year prior place of residence question must be available by current region of residence and age and sex. Specific examples of multiregional projections are included.