| Scope and Method of Study. The study compares the empirical approaches to migration analysis: multiple regression, latent variables, and index numbers in studying the determinants of migration. Gross migration data for 77 Standard Metropolitan Statistical Areas for the 1965 to 1970 period are used along with economic, climatic, demographic, city amenity, city disamenity, and city size data to estimate the multiple regression, latent variable, and index number models. Since the latent variable model has not been used in previous migration studies, the study provides an opportunity to explore the usefulness of the latent variable model in migration analysis. With respect to the determinants of migration, the study examines the question of the importance of economic and quality of life factors in influencing gross migration.;The study recommends further research concerning the applicability of the latent variable model of migration analysis. This work should include research on the selection of indicators for the latent variables. The use of simultaneous equation models of migration is also recommended as a possible method of migration analysis.;Findings and Conclusions. The study finds that multiple regression is superior to latent variables in investigating migration. Index number analysis is superior for complex models with multi-dimensional determinants of migration. However, the latent variable model is so flexible and holds such promise, that it should receive further study. The study also found that both economic and quality of life factors are important influences on migration. |
