This unique book introduces an essential element in applied demographic analysis: a tool-kit for describing, smoothing, repairing and - in instances of totally missing data - inferring directional migration flows. Migration rates combine with fertility and mortality rates to shape the evolution of human populations. Demographers have found that all three generally exhibit persistent regularities in their age and spatial patterns, when changing levels are controlled for. Drawing on statistical descriptions of such regularities, it is often possible to improve the quality of the available data by smoothing irregular data, imposing the structures of borrowed and related data on unreliable data, and estimating missing data by indirect methods. Model migration schedules and log-linear models are presented as powerful methods for helping population researchers, historical demographers, geographers, and migration analysts work with the data available to them.