Age and Spatial Interaction and the Indirect Estimation of Migration

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

This project will develop a formal model-based approach to the indirect estimation of migration that combines information from several geographical areas, time periods, and data sources to produce estimates of a certain variable on the basis of data that might be indirectly related to its value or that may be out of date. To do so this project will draw on work carried out in four areas: indirect estimation methods in demography, spatial interaction modeling in geography, model-based approaches to the analysis of missing data in statistics, and the mathematical representation of regularities in the age profiles of migration rates. It then will apply its particular formal model-based synthesis of these contributions to indirectly estimate the interregional migration flow data reported by the past four U.S. decennial censuses, and it will assess its approach to indirect estimation by considering the adequacy with which that approach "predicts" the past observed flows using only very limited data. Although indirect estimation techniques have been applied fruitfully in studies of mortality and fertility, they have not been developed as systematically and formally for the analysis of migration flows. Yet, with the future elimination of the "long form" questionnaire from the decennial censuses and its replacement by a much smaller continuous monthly sampling survey, students of territorial mobility often will find it necessary to deal with inadequate, missing, or possibly inaccurate sample data on migration by adopting such methods of estimation. This project will address that problem and will provide a formal methodology for carrying out such estimation efforts.