Assessing Validity and Scope of the Intrinsic Estimator Approach to the Age-Period-Cohort Problem

Liying Luo, University of Minnesota

In many different fields, social scientists desire to understand temporal variation associated with age, time period, and cohort membership. Among methods proposed to address the identification problem inherent in Age-Period-Cohort analysis, the Intrinsic Estimator (IE) appears to have few assumptions and yield good estimates of the age, period, and cohort effects. This paper assesses the validity and application scope of IE theoretically and illustrates its properties with simulation experiments. It shows that IE implicitly assumes a specific constraint on the true age, period, and cohort effects, which depends on the number of age, period, and cohort categories and has non-trivial implications for parameter estimation in empirical research. The author argues that the identification problem is inherent and cannot be circumvented by clever methods, so that APC techniques aiming to recover “simultaneously independent” age, period, and cohort effects are not only empirically misleading but also theoretically meaningless.