Maternal Expectations and Ex Post Rationalizations: The Usefulness of Survey Information on the Wantedness of Children

In this paper we assess the value of retrospectively-ascertained information on the wantedness of children by evaluating (i) the extent to which such information provides an unbiased estimate of the excess births occurring solely as a consequence of imperfect fertility control and (ii) whether information on the wantedness of a child is a good predictor of its subsequent treatment by parents-whether unwantedness signals neglect. We formulate a dynamic model of fertility incorporating stochastic fertility control, uncertain child traits and information accumulation from which we can formulate a rigorous definition of child-specific unwantedness. Based on information on both retrospectively obtained and pre-birth information on wantedness and on children's birthweight, we find that parents are more likely to report that children are wanted ex post if they have a better birth outcome and exhibit overly optimistic expectations about their children's traits and/or risk preferences. As a consequence, published statistics on the prevalence of unwanted births over-state the true proportion due to contraceptive failure by 26 percent.