Font Size: 

In many demographic applications the information of interest cannot be estimated directly but indirect estimation procedures have to be employed. With its emphasis on rates and events, many methods in demography are based on distributional models for counts (and corresponding exposures), so that extensions of Poisson models are a natural starting point. For many applications the indirect estimation problem can be phrased as a Penalized Composite Link Model (PCLM). This contribution will demonstrate the usefulness of the PCLM in demographic applications.