Analysis of sequences of household migration in the United States

Different migrants types differ in their socio-economic and demographic characteristics. They also show large differences in their outmigration probabilities. Yet, explanatory analyses of the reasons for such different behaviors are not conclusive. The purpose of this dissertation is to look for a more satisfying answer to the question of why people with different migration histories exhibit different migratory behaviors. This is done within a theoretical framework that emphasizes an integration of both the human capital theory and the life course perspective on migration, and acknowledges the household as the decision-making unit in migration analysis.;The main data source used is the Survey of Income and Program Participation (SIPP). The SIPP is a multi-panel longitudinal survey that is nationally representative of households in the civilian noninstitutionalized population. The principal advantages of SIPP over other data set are: its large sample size, recentness, and the possibility to measure covariates before the move.;The preliminary univariate analysis indicates that families at risk of a secondary move are almost four times as likely to migrate than are those who live in their place of birth. Also, most of the covariates have the expected relationship with household outmigration probabilities.;A dichotomous logit model is used to analyze the determinants of outmigration for different types of move in a multivariate setting. The model is later generalized to the polytomous case in order to contrast determinants of onward migration from those of return migration.;In general, the life cycle variables were more likely to be significant determinants of primary outmigration than of secondary outmigration. Conversely, human capital variables were often found to be stronger predictors of secondary outmigration. However, the advantage of the full model, which includes both the life cycle and the human capital variables appeared clearly from the analysis.;This study demonstrates the importance of using the household as the unit of analysis. Many of the variables show opposite coefficients when measured for the household head and the household head's spouse. In addition, the employment status of the spouse appears to be the most influential deterrent to household outmigration.