Abstract
The importance of studying the relationship between the timing of moves and the choice of destination is discussed in terms of theories of migration decision processes and of methodological issues in the analysis of migration histories. Special attention is given to duration-of-residence effects. Statistical procedures based on the theory of competing risks and semi-Markov processes are developed and applied to residence histories of Norwegian men, 1965–1971. Subject to several qualifications, the results confirm the hypothesis that decisions of whether to move and where to go are independent. The methods can be used in many analogous problem areas where timing and multiple outcomes are of concern.