A Critique of the Glass-Grebenik Model for Indirectly Estimating Desired Family Size.

In Volume VI of the "Papers of the Royal Commission on Population," D. V. Glass and E. Grebenik present a model for estimating distributions of desired family size. The term "intended family size" will be used to designate children ever born under these hypothetical conditions of absent sterility and perfect birth control. This is to distinguish it from the family size that a couple might report as their "ideal" or "preferred" family size. This paper does not attempt a comprehensive evaluation but focuses on three aspects. First, evidence is given that asking couples for their fertility "preferences" or for their family size "ideals" does not adequately measure "intended family size." The second question raised in this paper is whether the model has yet attained its most efficient form. The assumption that the chances of another unintentional birth remain constant irrespective of the number preceding it is not only suspect, it leads to misleading results. Third, suggestions are made for modifying the model. The modifications would complicate the model but also add to its possible uses. In the passage from early marriage to later marriage, "private ideals" and "fertility preferences" change from expectations of future fertility to recollections of past intentions. This necessitates discussing "fertility preferences" in relation to more than one period of marriage.