Third, it is widely recognized that the NSFG underreports abortion. Many unintended pregnancies are resolved by abortion, as discussed later in this chapter;2 therefore, if one wants to know, for example, how many pregnancies were unintended in a given year, data on the number of both births unintended at conception and the number of abortions would be needed—data that the NSFG does not have in full measure. Jones and Forrest (1992) suggest that the number of abortions reported in the 1988 NSFG represents only 35 percent of the number actually obtained. Important attempts have been made to supplement the NSFG data with more complete abortion information in order to provide an understanding of the overall level of unintended pregnancy (Forrest and Singh, 1990). The NSFG remains the key data set, however, for tracking the intention status of births, providing comprehensive information on the families in which such births occur and on many other issues as well.

Percentage Rates of Unintended Pregnancy

Figure 2-1 shows the best available estimate of the percentage of pregnancies that are unintended. These data are based on the 1988 NSFG, supplemented by abortion data from 1987 compiled by The Alan Guttmacher Institute and the Centers for Disease Control. As that figure shows, in 1987, 57 percent of all pregnancies were unintended at the time of conception. This figure of 57 percent includes pregnancies that were aborted as well as both mistimed and unwanted pregnancies that led to live births.3 Only 43 percent of all pregnancies in that year were intended at conception and resulted in live births.4


Much research in this field tacitly assumes that all pregnancies ending in abortion were unintended at the time of conception. Although this is accurate as a general finding, it is important to acknowledge that some very small portion of abortions are obtained for pregnancies that were intended at conception, but subsequently became problematic because of the diagnosis of a serious genetic defect in the fetus, for example, or some other troubling turn of events (Torres and Forrest, 1988). This dynamic, however, affects only a small percentage of abortions and does not change the overall estimates presented in this report.


Miscarriages are excluded from this analysis and all others in this report because the number of pregnancies ending in miscarriage is not well established and because there is no information on the distribution of miscarriages by intention status.


These data are generally consistent with information beginning to emerge from the 1988 National Maternal and Infant Health Survey, which also examined the intention status of births to both married and unmarried women (Kost and Forrest, 1995).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement