Assessing the Validity of Microsimulated Kinship Networks Using Swedish Population Registers

Liliana P. Calderón-Bernal, Max Planck Institute for Demographic Research (MPIDR)
Diego Alburez-Gutierrez, Max Planck Institute for Demographic Research
Martin Kolk, Stockholm University
Emilio Zagheni, Max Planck Institute for Demographic Research (MPIDR)

Estimating kinship networks is data-demanding and mainly possible using empirical sources or demographic models. Empirical micro-level data, like population registers, provide a realistic picture but are scarce and limited by truncation and survivorship bias. Demographic models like microsimulation use aggregated data, but only minimally account for population heterogeneity, family similarity and non-marital fertility. We analyse the differences between approaches by examining the correspondence in size and structure of kinship networks derived from empirical and microsimulated data. We compare the number of kin from grandparents to grandchildren for the Swedish population (cohorts 1915-2017), estimated based on contemporary Swedish population registers (Kolk et al., 2023) and the SOCSIM microsimulation programme. Mean numbers and distributions of most kin are similar across both sources. The microsimulation produces slightly lower numbers of kin for cohorts unaffected by truncation in the registers, but better accounts for kin of early cohorts that are under-registered due to missing parent-child links, conditioning on survival or migration. Although a systematic comparison of empirical and microsimulated kinship networks is only possible for a few countries like Sweden, our assessment of the accuracy of synthetic kinship networks may serve as a reference for research using microsimulation to study kinship in other contexts.

Keywords: Families, Unions and Households, Simulation , Civil Registration and Vital Statistics, Data and Methods

See paper.

  Presented in Session 170. Kinship: Patterns, Changes, and Determinants