Stability and Change in Expectations about Family Size: A Longitudinal Study

Concerning the problem of how stable are a mother's expectations about the number of additional children she will have in future years, an analysis has been made concerning the stability of expectations of a panel of mothers in 3 separate interviews over a 2 year period. Responses about expected number of children were obtained in January-March 1962, from a probability sample of mothers in the Detroit area who had a 1st, 2nd or 4th baby in July 1961. A 2nd set of expectations was obtained from the same mothers in October-December 1962, and a 3rd set in September-October 1963. The data revealed that responses about numbers of children expected are remarkably stable in the aggregate over a 2 year period in 3 separate interviews for women recently having a 1st, 2nd, or 4th birth. The changes that occur for the majority are basically small ones, and often a 2nd change compensates in part or completely for an earlier change in an opposite direction. These compensating changes for individual respondents along with the compensating changes for different respondents produce an aggregate result of no change for the expectations of all women at each parity. Only the factors of the discrepancy between preferred and expected number of children and the experience of a pregnancy have an importanat influence both on whether expectations change and on the direction of change. Catholics are more changeable than non-Catholics, although there is no consistent difference in the direction of change. Various other social and economic characteristics were found to be associated with the amount of change, but most had no consistent relation to the direction of change. The results of this analysis are interpreted to be consistent with although not proof of the view that social norms about family size tend to control behavior of American couples, so that their expectation changes tend to be samll, overlapping, and compensating.