ABSTRACT
In this chapter, the author focuses on age-period-cohort (APC) models for unequal span data. Most APC studies in the literature have identical time spans in age and period groups, such as five years in most public health studies, or ten years in demography or economic studies. These identical time spans make it easy to identify birth cohorts by the diagonals of the table since the individuals on the same diagonal were born in about the same years and thus belong to the same birth cohort. The author illustrates the intend-to-collapse (ITC) method with the retirement accounts data. He then provides the cohort structure and the coding by the ITC method. The author also illustrates the multiple estimators with the retirement accounts data and presents three curves of age, period, and cohort trends by linear model fitted to the log-transformed mean value of retirement accounts.
