Life Insurance Demand and Household Portfolio Behavior: Comment

Abstract

The recent article in this Journal by Headen and Lee [9] is an interesting attempt to extend previous work in the area of ordinary life insurance demand analysis. The authors characterize life insurance as one of several components of a household's financial asset portfolio and, therefore, suggest that decisions to purchase or not purchase insurance in the short term involve choices among portfolio alternatives. The viability of this hypothesis is explored using macro measures restricted to the household sector of the economy and relating insurance demand to the financial market behavior of households. In light of the relative paucity of published work in the area of life insurance demand analysis, an extension of this nature can be an important contribution to the literature. The authors clearly have recognized the exploratory nature of their research and some of its limitations. For example, their model incorporates only four variables thought to explain demand, while ignoring variables which Headen and Lee admit are relevant to the insurance purchase decision. However, we feel it necessary to call attention to some additional conceptual and empirical questions which were not fully discussed in the article and which must be kept in mind when evaluating the article's contribution. The objectives of this comment are to examine the conceptual dimensions of the paper, to analyze certain of the empirical procedures, and to make suggestions for extension and improvement of the model. A shortcoming of the authors' approach is the same one that they have attributed to other studies in the literature, that is, these studies have used, exclusively, aggregate economic variables and have employed a time series approach. The Headen and Lee article purports to deal with household decision-making and hypothesizes differing behavior across individual households; yet, the aggregate approach used does not afford the opportunity to investigate the behavior of individual households. There are, as a result, some potentially serious conceptual problems with the model. Assumptions are made regarding household types, household behavior, and the place of life insurance in household decision-making which are