Article ID: | iaor20061576 |
Country: | United States |
Volume: | 43 |
Issue: | 1 |
Start Page Number: | 121 |
End Page Number: | 132 |
Publication Date: | Feb 2006 |
Journal: | Journal of Marketing Research |
Authors: | Kamakura Wagner A., Du Rex Y. |
Keywords: | markov processes |
Household life cycle has been widely used as a determinant of consumer behavior and a basis for market segmentation. However, there is considerable disagreement about how life stages should be defined and how households progress through these stages. Existing studies use a priori definitions, which are tested on a cross-sectional survey of households collected at a single point in time and thus cannot reveal the real dynamics of the household life cycle. The Panel Study of Income Dynamics provides longitudinal data on household composition in the United States for a period of 34 years; the authors use this to identify empirically the most typical stages and paths that US households have followed since 1968. They develop a hidden Markov model in which the stages of the household life cycle are taken as latent, unobservable states that are uncovered from the manifest household demographic profiles over the 34 years, assuming that households evolve through these latent stages following a first-order Markov process. The authors apply their results to classify members of another panel (Consumer Expenditure Survey) into life stages, which enables them to study the impact of the household life cycle on households' budgetary allocations, providing a comprehensive analysis of lifestyles (through expenditure patterns) over the household life cycle.