This essay is a commentary on a Research Note by James Trussell, Barbara Vaughn and Joseph Stanford entitled "Are All Contraceptive Failures Unintended Pregnancies? Evidence from the 1995 National Survey of Family Growth."

An investment company headed by Nobel laureates recently went so spectacularly into debt that the government had to rescue it, lest its failure threaten the entire global banking system. Explaining the situation, an observer noted that a series of unusual events had occurred. "The fault isn't with the models," he insisted.1 This is a useful reminder to all of us—researchers, clinicians and politicians—to be constantly aware of how often human behavior refuses to conform to our models.

The results of Trussell, Vaughan and Stanford likewise make clear that we should use our models with caution. As they point out, both "unintended pregnancies" and "contraceptive failures"—two touchstones of contemporary policy initiatives—are more complex than they appear at first glance.

Part of the problem resides in the history of the instrument used to measure unintended pregnancy. The series of questions used to measure intention status in the National Survey of Family Growth (NSFG) grew out of earlier knowledge, attitude and practice surveys that had been systematically refined in the period just after World War II. Given the demographic circumstances of the era, notably the largely unanticipated baby boom, these items were originally intended to measure "surplus" fertility, or the number of children a couple had in excess of planned family size.

Over the last quarter-century, however, the fertility at issue in American public policy has not been that of the married mother in her 30s and 40s who produces more children than planned, but that of the young woman who has a child before she has planned to. The need to preserve comparability between various cycles of the NSFG means that the fertility questions in the most recent survey, which have undergone extensive revision, bear the marks of having once measured a very different part of the family-building cycle. Meanwhile, the shift in political and policy interest away from the end toward the beginning of the cycle means that the NSFG is measuring a much more complicated social reality. For an earlier generation of women (and men), these surveys did indeed measure what they were supposed to, namely fertility, or how many children a couple actually wanted. But for an increasing number of American women (and men), something entirely different is at stake—not fertility, but motherhood (or fatherhood, as the case may be).

To the extent that more and more women are contemplating, not "excess" births at the end of the family-building cycle, but unexpected pregnancies at the beginning, what is really being tapped is the willingness to enter a new social role, not how extended that role should be.

Drawing on the work of decision theorists, I submit that the decision whether or not to be a parent is a very different kind of decision—a threshold decision—and touches on a much more complicated decision tree than the decision to have a completed family size of four or five children.2 Other events in the larger social world have conspired to further complicate this decision. As parenthood becomes increasingly severed from marriage in all industrialized countries, deciding about a pregnancy means a decision not only as to whether one wants to become a parent, but whether one wants to become a parent with a particular partner inside or outside marriage.3 In decision-making terms, what has happened is that now both parties must decide whether they want to continue a pregnancy or wait for a more promising situation to come along, however that is defined.

Add to these issues the fact that the women and men coming to the end of their reproductive lives at the close of this century have watched the fundamental status of pregnancy and birth change before their very eyes. Until 1960, the year in which the contraceptive pill was approved for general use in the United States, every sexually active person lived continually with the possibility of becoming pregnant. (As an indicator of this revolution's enormity, witness the fact that "the pill" is the only pharmaceutical development in history to which we can refer generically: "the" pill, as if there were only one.) A few privileged and skillful people could more or less control their fertility, but even they—and the society in which they lived—did not view this power as a given but as luck.

For the baby boomers reaching late middle age, this world changed almost overnight: By 1965, just five years after its introduction, the pill had become the most commonly used contraceptive among married couples.4 Truly effective contraception, backed up by legally available abortion, meant that for the first time in history, people had to decide actively whether to have a child, rather than passively let nature take its course.

While this development surely affected women who had already decided to be mothers and needed help in limiting their family size, its most profound impact was on a generation of women now permitted to decide if and when they would become mothers. The decision was given unusual salience in that it occurred in the context of changing roles for women, growing concern about overpopulation and shifting perceptions of gender-assigned responsibilities for the practice—and success—of contraception.5

What Trussell and his colleagues have detected in the NSFG is a structural shift in the meaning of sexual and reproductive choices in the lives of an increasing number of Americans. More and more people are faced with decisions, not about "fertility," but about motherhood (and parenthood more generally) in a social and political climate in which motherhood provokes antagonistic political battles about "family values."

Layered atop normal human ambivalence about sex, pregnancy and childbearing is the reality that our cultural, social and political structures have not caught up with the twin revolutions that the pill and the most recent wave of the women's movement have ushered in. What are the rules now? What does it mean if a couple knowingly have sex without using a contraceptive? Are they signaling commitment or carelessness? How do people use potential parenthood to test themselves and their relationships?

Given the profundity of the revolution in intimate life, and the sheer scope of social change brought about by the unprecedented control of conception, it is not surprising that Trussell, Vaughan and Stanford have found that a substantial number of contraceptive "accidents" may have been something else. Becoming "accidentally" pregnant permits people to duck the onerous responsibility of having to decide whether to enter into parenthood, and to do so in the only country in the developed world that permits people to become parents with virtually nothing in the way of social support.

Understanding better the consequences and meaning of the phenomenon documented by Trussell and his colleagues is one of our most urgent research tasks. Indicators such as our abortion rate and our rate of "unintended"pregnancy make clear that choices about parenthood are problem-fraught in this country. We must learn more about how these processes play out in the intimate lives of men and women and in the culture as a whole. If we do not, we will not only have to live with the highest rate of contraceptive "accidents" in the industrialized world, we will also continually face political movements that promise to resolve the complexities involved by turning the clock back to what they see as simpler times.6