The dangers of using 'negative durations' to estimate pre- and post-migration fertility.

To keep childbearing that occurs before and after migration separate from each other, many analysts apply a technique that uses 'negative durations' to estimate the childbearing risks that migrants have before they migrate. This strategy can lead to incorrect results and should be abandoned. In this research note, we use data for births and internal migration in Sweden to highlight how the two types of behaviour can be kept apart conceptually and analytically without use of 'negative durations'. The procedures used can easily be generalized to any similarly linked pair of behaviours.