Underlying family-size preferences and reproductive behavior

Abstract

Although about half the married women in a US survey said they wanted one or two children, underlying family-size preferences indicate a potential for many to have larger families. Using a measure of family-size preference that probes beyond single-valued statements, it is found that identical stated family-size desires or intentions may have different meanings for different people. Differences are systematically related to the stage in family building at which contraception begins, to current (or likelihood of future) use of effective methods, to the probability of continued childbearing, and to unwanted pregnancies.

PIP: Data from the 1973 National Survey of Family Growth (NSFG) were used to report on underlying preferences for number of children among U.S. married women as well as how such preferences may differ from single-valued statements of desired family and how they are related to reproductive behavior. The 7459 cases on which the findings rest come from a 4-stage probability sample of married women in the childbearing ages (15-44). Different sampling ratios were used for blacks and whites to ensure sufficient cases for detailed analysis, with a weighting system to ensure effectiveness. The data reveal that while about 1/2 the married women state that they want only 1 or 2 children, the underlying preferences of many indicate a potential for larger families; 1/3 of this group has a bias toward a moderate-size or large family, despite statements to the contrary. For the country as a whole, 45% of all wives of childbearing age have an underlying preference for a small family, 25% favor a moderate size, and 30% have a bias toward a large family. Identical statements about desired family size, or about expected or intended number of children, do not have the same meaning for everyone, in the sense that such responses may be supported by a range of underlying preferences, as reflected in the IN-scale measure. In other words, statements about desired size represent only partial information, and the extent to which underlying preferences depart from the conventional responses obtained provides further information that is important for interpretation and for fertility prediction. Various subgroups of the population differ considerably in the way their underlying preferences relate to their single-valued statements. Women in very low income or low educational strata who say they prefer or intend to have 2 or 3 children are much more likely than high status women to have underlying preferences for a large family size, suggesting that they will be less likely to achieve their preferred or intended family size.