Estimating Complete Migration Probabilities from Grouped Data: A Methods Protocol for a Human Internal Migration Database

Sigurd Dyrting, Charles Darwin University
Andrew J. Taylor, Charles Darwin University

Internal migration intensities, along with fertility and mortality rates, are the primary processes used to understand the dynamics and spatial structure of population change. Incorporating migration into demography's quantitative framework allows a description of population change in size and composition across both time and space but, while mathematical and conceptual frameworks have been developed, researchers in internal migration lack a public repository of historical age-origin-destination-specific migration probabilities harmonised to a common format and spanning a range of countries. A solution to this data problem in spatial demography will require a robust method for inferring migration probabilities from census and survey data subject to significant levels of uncertainty due to both small-sample noise and age aggregation. In this paper we extend the P-TOPALS and P-spline methods for estimating migration probabilities to the case of grouped ages with the aim of developing a methods protocol for a Human Internal Migration Database. We illustrate the method by estimating complete age-origin-destination migration probabilities for more the 50 countries using microdata samples from IPUMS International.

Keywords: Data and Methods, Internal Migration and Urbanization, Spatial Demography, Mathematical demography

See extended abstract.