An implicit ambivalence-indifference dimension of childbearing desires in the National Survey of Family Growth

Abstract

1. IntroductionIt is common in fertility surveys to collect data about the respondents' motivations and intentions regarding future childbearing. It is also common to collect data about the degree to which each of any previous pregnancies was intended and wanted. These data are collected in order to inform fertility forecasts, as well as to provide information about the family planning status of births in the population of interest. Because of society's special concern with preventing the individual and family suffering associated with pregnancies that are not wanted, considerable attention has been devoted to understanding their antecedents and social contexts.In recent years it has become clear that there are two different approaches to determining when a is unwanted. One approach asks respondents whether they wanted to have a baby before the occurred, thereby defining an in terms of their preconception desires. This approach is well represented by the National Survey of Family Growth's (NSFG's) series of questions about intended, mistimed, and pregnancy, where these three constructs all reflect preconception phenomena (London, Peterson, and Piccinino 1995). The other approach asks the respondents what they wanted and felt after the had occurred, thereby defining an in terms of their postconception desires and feelings. This approach is well represented by Miller's (1974, 1994a) series of questions about a pregnancy's intendedness and wantedness, where intendedness reflects the amount of intention that went into the respondent getting pregnant and wantedness reflects the degree of motivational and emotional acceptance or rejection of being pregnant. Numerous other researchers have adopted the postconception approach as well (for example, Adler 1992; Sable and Herman 1997; Blake et al. 2007; East, Chen, and Barber 2012). While in many cases there is little or no difference in the assessment of wantedness between the two approaches, in others, especially when preconception desires are weak, conflicted, and/or largely unconscious but the is welcomed, the difference may be appreciable. This discrepancy raises the question of how the two assessments of wantedness are related to each other. In addition, because both of these assessments are commonly made after the pregnancy, and in most cases after the birth has occurred, the question of retrospective bias is also raised.In previous research, Miller and Jones (2009) used data from the NSFG to test a multivariate model of the relationship between women's preconception desires for a (reported within four years of conception) and their postconception wantedness of that pregnancy. Their analysis showed that preconception desires were the primary predictor of postconception wantedness and completely eliminated preconception intentions from the prediction regression in the general model, as well as in a number of the context specific models. These authors questioned whether retrospective bias may have contributed to this predictive strength. However, they found that inclusion of a time-since-the-pregnancy variable as a moderator in the analysis showed that the length of retrospection had no effect on the predictive strength of preconception desires in their models, leading the authors to conclude that whatever retrospective bias was present, it did not invalidate their findings.Miller and Jones (2009) also discussed how the meaning of the construct of unwanted pregnancy had evolved during the twentieth century. In the first four decades of that period when Margaret Sanger and others in the nascent family planning movement were struggling to legalize birth control and make it available to the poor, the construct had its postconception meaning. During the next six decades the construct progressively assumed its preconception meaning as demographers progressively focused first on predicting family size, then on the determinants of both the number and timing of births, and finally settled on the three category preconception measure with the launching of the NSFG. …