Skip to ContentSkip to Navigation
OnderzoekUrban and Regional Studies InstitutePopulation Research Centre


Former research theme 'Demographic monitoring and forecasting'

The first theme is mainly methodological and continues the long-standing research interests of Frans Willekens and Sergei Scherbov. Models are developed that capture key features of demographic dynamics. Population heterogeneity is accounted for by stratifying the population into interdependent subpopulations, often referred to as multistate populations or multigroup populations. Stratification variables are personal attributes such as region of residence, marital or household status, employment status, etc. The dynamics are governed by transition probabilities. Uncertainties are associated with the ignorance of factors and causal mechanisms that underlie the demographic parameters, and with measurement problems such as sampling and indirect measurement. 

Two questions guide current research: 

  • how to properly account for the uncertainties and
  • how to reduce the uncertainties

To account for the uncertainties, probability distributions are used. As a consequence, demographic indicators no longer have single values (point estimates) but are accorded ranges of plausible values (interval estimates). The demographic models are no longer deterministic but stochastic models. Applications include stochastic life tables and probabilistic population projections. The prediction intervals in projections are based on past forecasting errors and/or on expert opinions - the combination of quantitative and judgmental methods. The reduction of uncertainties in demographic analysis and projection is achieved by incorporating substantive knowledge on the mechanisms governing demographic change. This research uses life-history modeling techniques and longitudinal data to predict transition probabilities and other change indicators. 

Last modified:15 November 2012 2.26 p.m.