The paper is organized as follows: Section B outlines the difficulties associated with the choice of the unit of analysis in longitudinal studies. Section C discusses the data used in the analysis and our technical approach. In section D we provide a framework for our analysis by presenting an overview of the extent and patterns over time of transtions in multiple program participation. Section E presents our findings regarding the impact of the choice of the unit of analysis on the measure of multiple program participation, and, finally, section F summarizes the study and presents our conclusions.