Terminology

Women often informally discuss their personal feelings about the timing of the pregnancies they have had—whether a baby came just a bit too early, whether a pregnancy occurred at a time when it interfered with future plans but would have been wanted at a later time, whether a baby had been desperately sought for several years and was the answer to heartfelt prayers, or whether a pregnancy was not wanted at any time. Some women feel ambivalent or may disagree with their partners, one wanting a pregnancy and the other preferring to wait. Books, movies, dormitory discussions, real-life arguments, pillow talk, and conversations over the back fence focus on the complexities of wanted, unwanted, and mistimed pregnancies. It is difficult, however, to quantify people's feelings and sort them into categories that hold comparable meaning over time and across varied social groups. When an unmarried high school sophomore reports that her pregnancy occurred too early, her assessment of ''too early" probably means something different from the report of a married engineer who wanted a baby after working for several years but got pregnant her first year on the job.

Information on the demography of unintended pregnancy—the subject of this chapter—is dominated by data from the National Survey of Family Growth (NSFG), a federally-sponsored survey that has developed quite specific terminology and definitions to measure "unintended pregnancy." Over the past four decades, a series of questions has been regularly asked of women by this survey and its predecessors, the Growth of American Families surveys and the National Fertility Studies, in an effort to learn more about women's plans and intentions at the time they became pregnant. In these surveys, women are asked about pregnancies during the previous 5 years, including whether contraception was being used at the time the woman became pregnant. The relevant NSFG questions appear in Appendix C, along with additional material on the history and future plans of this survey.

On the basis of these questions, pregnancies in the NSFG are defined as

• intended at conception: wanted at the time, or sooner, irrespective of whether or not contraception was being used; or
• unintended at conception: if a pregnancy had not been wanted at the time conception occurred, irrespective of whether or not contraception was being used.

Among unintended pregnancies, a distinction is made between mistimed and unwanted:

• mistimed conceptions are those that were wanted by the woman at some time, but which occurred sooner than they were wanted; and
• unwanted conceptions are those that occurred when the woman did not want to have any more pregnancies at all.

The labels and terms from the NSFG have many limitations and ambiguities that should be mentioned. First, it is important to emphasize that the NSFG questions are designed to probe feelings at the time of conception, not at the time of birth. This distinction is important because a woman's feelings can change in many ways over the course of pregnancy (Miller, 1974). For example, Poole and colleagues (1994) report that among a sample of low-income women who were queried both early and late in their pregnancies, 12.5 percent reported a positive shift in attitude toward the pregnancy, and 10 percent reported a negative shift. In particular, an unintended pregnancy can result in a much anticipated birth and a cherished child. To emphasize this critical distinction, this report does not refer to unwanted births, for example, but rather to births resulting from unwanted pregnancies or conceptions; similar precision is used with the term mistimed.

The distinctions that the NSFG draws between intended and unintended and between mistimed and unwanted carry important implications. For example, unintended pregnancies are far more likely to end in abortion than intended pregnancies, and births resulting from unwanted conceptions appear to carry particular risks for both mother and child. In Chapter 3, these consequences of unintended pregnancy are discussed in detail.

It is also important to note that knowledge about pregnancy intentions derived from the NSFG comes entirely from women. Bumpass (1994) has reported data about partners' preferences, as provided by female respondents in the 1988 National Survey of Families and Households (see also Williams, 1994), noting that in more than one-fourth of the cases in which a woman described a birth as resulting from an unintended pregnancy, she reported that her spouse or partner had either wanted the birth at that time or was indifferent to the timing. Such disagreement may be a factor in the occurrence of pregnancies that women do not intend; this issue of partner interaction is taken up in more detail in Chapter 6.

It is also apparent that the NSFG survey questions, and similar ones used in other surveys, are often not able to measure the complicated mix of feelings that can surround pregnancy, as described by recent ethnographic research in particular (Musick, 1993). It is not uncommon to learn from a pregnant woman, for example, that she both did not "intend" to become pregnant but also was not using contraception, or that perhaps she wanted to be pregnant but was less enthusiastic about having a child (Luker, 1975). In Chapter 6, these complexities are explored in detail.

Other caveats with regard to the NSFG measures of pregnancy intention have been raised. For example, the questions about intendedness are retrospective, asking about all pregnancies in the preceding 5 years, which may make the answers offered subject to distortion and recall error. The questions do not distinguish how many months or years pregnancy timing was off, nor do they reflect the intensity of the woman's feelings about the timing of conception. In commenting on data sources other than the NSFG, Chapter 3 and Appendix G also address these generic problems in measuring the intention status of a given pregnancy, noting in particular that the difference between an intended and an unintended conception is often more complicated than these terms imply. Well aware of these nuances, the 1995 NSFG cycle will use an elaborate approach to establishing the intendedness of a given conception (see Appendix C), recognizing that the concept of intended versus unintended is more a continuum than an either/or matter. Consistent with the new approach being taken by the NSFG, one of this report's recommendations is that researchers develop more refined and differentiated measures of intention status that can accommodate important concepts like ambivalence, denial, and confusion and that can address the feelings of men and couples as well as individual women (Chapter 9).

These various caveats about the NSFG measures are not meant to suggest that this national survey and other similar ones fail to capture important information about intention at the time of conception. In fact, despite errors at the individual level, the general validity of the aggregate levels and trends in these measures has been repeatedly demonstrated. The same questions have been used with older and younger women as well as married and unmarried women and across a period of years, ranging from the time before legal abortion became widely available in the 1970s through to the 1990s. In addition, very few women are unable to answer the series of questions; less than 1 percent fall into a residual "undetermined" category. If there is any bias in the NSFG on the issue of intendedness, available evidence indicates that the direction of this bias would be in underreporting unintended pregnancy (Ryder, 1985). That is, the NSFG data may well provide overestimates of intended pregnancies and underestimates of unintended pregnancies.

Finally, three other important limitations in the data available from the NSFG should be noted. First, most research and analysis regarding pregnancy intention are based on the 1988 NSFG, which is by now somewhat dated, predating in particular the introduction and use of both injectable and implantable hormonal contraceptives. Nonetheless, the 1988 survey remains a highly credible source of information on intention status.¹ Although there was a telephone reinterview of the 1988 NSFG sample in 1990, it reported a less-than-optimal response rate, and thus its results are used sparingly in this report. The next NSFG survey, being completed in 1995, will provide more current data to extend the rich information already available.

Second, the marital status groupings used by the NSFG in its fertility questions have changed over the years. For example, the 1965 and 1970 National Fertility Studies were conducted only among currently married women. It was not until the 1973 NSFG that separated, divorced, and widowed women were also queried along with currently married women. From 1982 on, however, the NSFG has sampled all women aged 15–44, irrespective of marital status.

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

Figure 2-1

All pregnancies by outcome, 1987 (miscarriages excluded). Source: Forrest JD. Epidemiology of unintended pregnancy and contraceptive use. Am J Obstet Gynec. 1994;170:1485–1488.

Figure 2-2 is limited to unintended pregnancies only (i.e., intended pregnancies are not included), showing the outcomes of these pregnancies. One striking fact that this figure reveals is that more than half of all unintended pregnancies ended in abortion in 1987. The figure also shows that the majority of births from unintended pregnancies were from mistimed rather than unwanted pregnancies.

Figure 2-2

Unintended pregnancies by outcome, 1987 (miscarriages excluded). Source: Forrest JD. Epidemiology of unintended pregnancy and contraceptive use. Am J Obstet Gynec. 1994;170:1485–1488.

In absolute numbers, these proportions mean that in 1987, of the 5.4 million pregnancies that were estimated to have occurred, about 3.1 million were unintended at the time of conception. Within this pool of unintended pregnancies, some 1.6 million ended in abortion and 1.5 million resulted in a live birth. Only 2.3 million pregnancies in that year were intended at the time of conception and resulted in a live birth.

Women at Risk of Unintened Pregnancy

In analyzing which women contribute to this pool of unintended pregnancies, it is helpful to begin with the profile of women "at risk" of such pregnancies. The definition of women at risk of unintended pregnancy is multifacted. They are women who (1) have had sexual intercourse; (2) are fertile, that is, neither they nor their partners have been contraceptively sterilized and they do not believe that they are infertile for any other reason; and (3) are neither intentionally pregnant nor have they been trying to become pregnant during any part of the year.⁵ In 1990, about 31 million women met these criteria and could therefore be considered at risk of unintended pregnancy (Table 2-1). These 31 million women represented about half of the 62 million women in the reproductive age range, defined as ages 13–44.

TABLE 2-1

Number of Women Aged 13–44 and Estimated Number and Percentage of Women at Risk of Unintended Pregnancy, by Age, According to Poverty Status, 1990 (in 1,000s).

Not surprisingly, among women of reproductive age, the highest proportions at risk of unintended pregnancy are found at ages 18–29, the age range in which most women are fertile, have usually begun sexual activity but often prefer to delay pregnancy, and are generally too young to seek sterilization (Figure 2-3). Seventy percent of women in this age category are at risk of unintended pregnancy. This proportion drops dramatically among women age 30 and over, many of whom have been sterilized or have partners who have been sterilized. Nevertheless, nearly 12 million women aged 30–44 remain at risk of unintended pregnancy, compared with nearly 17 million women aged 18–29.

Figure 2-3

Proportion of women in age groups at risk of unintended pregnancy. Source: Henshaw SK, Forrest JD. Women at Risk of Unintended Pregnancy, 1990 Estimates: The Need for Family Planning Services, Each State and County. New York, NY: The Alan Guttmacher Institute; (more...)

Within these various age categories, economic status clearly affects the level of risk (Table 2-1). The association between the proportion of women at risk of unintended pregnancy and economic status varies with the age of the woman. Among females aged 13–19, a higher proportion of teens from families in poverty are at risk, primarily because of earlier initiation of sexual activity among low-income teens. Among teens aged 15–17, for example, 46 percent of those with incomes below the poverty level were at risk of unintended pregnancy, compared with about a third of teens with family incomes 2.5 times the poverty level or above. On the other hand, among women in their 20s, slightly higher proportions of affluent women are at risk (three-fourths of women 2.5 times the poverty level or more versus two-thirds of women below the poverty level). Among women aged 30–44 in all income groups about 4 in 10 are at risk of unintended pregnancy.

The proportion of women at risk of unintended pregnancy varies only slightly by race and ethnicity (data not shown). Non-Hispanic blacks and Hispanics are somewhat more likely to be at risk (52 percent for both groups) than non-Hispanic whites (49 percent). No similar data are available for Native or Asian Americans.

Women Who Have Unintended Pregnancies

As noted above, the most recent data show that 57 percent of all pregnancies were unintended at the time of conception (Figure 2-1). This number includes pregnancies that were aborted, as well as both mistimed and unwanted pregnancies that led to live births. Only 43 percent of all pregnancies were intended at conception and resulted in live births.

Women of all socioeconomic, marital status, and age groups contribute to this pool of unintended pregnancies, as shown in Table 2-2. (This table displays pregnancies by both intention status and outcome for several demographic groupings, and also includes the actual number of pregnancies involved in each demographic category.) As this table shows, even among currently married women, 4 in 10 pregnancies were either mistimed or unwanted. However, the incidence of unintended pregnancy varies substantially by marital status, age, and economic group, being higher among unmarried and low-income women and among women at either end of the reproductive age span. (Variations in unintended pregnancy according to patterns of contraceptive use are discussed in Chapter 4.) Pregnancies during the teen years are particularly likely to be described as having been unintended. In 1987, for example, 82 percent of the pregnancies experienced by teenagers aged 15–19 were unintended, as were 61 percent of the pregnancies experienced by women aged 20–24.⁶ Similarly, among women over age 40, 77 percent of the pregnancies experienced were unintended, as were 56 percent of the pregnancies to women aged 35–39. However, even among women between the ages of 25 and 34, between 42 and 45 percent of all pregnancies were described as having been unintended when they occurred.

TABLE 2-2

Estimated Proportions of Pregnancies (Excluding Miscarriages) by Outcome and Intention, Percentage of Pregnancies Unintended, and Percentage of Unintended Pregnancies Ending in Abortion, 1987, by Marital Status, Age at Outcome, and Poverty Status at Interview. (more...)

Marital status, which is highly correlated with age, is also strongly related to whether a pregnancy is unintended. The vast majority of pregnancies to never-married women—88 percent—are unintended. Moreover, 69 percent of the pregnancies among formerly married women are unintended. Nonetheless, even among currently married women, 4 in 10 pregnancies are unintended, as noted earlier.

Economic status is also strongly correlated with pregnancy intention. Among women whose family incomes fell below the poverty level in the 1988 NSFG, 75 percent of pregnancies were described as unintended, compared with 64 percent among women with incomes between 100 and 200 percent of the poverty level. Among women whose incomes exceeded 200 percent of the poverty level, 45 percent of all pregnancies were unintended.

Consistent with the higher rates of unintended pregnancy among women in poverty, unintended pregnancy is much more common among black than white women. Data from the 1990 NSFG show that the percentage of births reported as unintended at time of conception was 62 percent among black women; among white women, the comparable figure was 41 percent⁷ (Piccinino, forthcoming).

Among some smaller subgroups, the proportions of pregnancies that are unintended may be appreciably higher than for the nation as a whole. Groups for whom this appears to be the case include, for example, women who are homeless, teenagers who have dropped out of school and engage in multiple high-risk behaviors, of which sexual intercourse without contraception is only one, and women who are heavy abusers of alcohol and illegal drugs (Centers for Disease Control and Prevention, 1994; Armstrong et al., 1991).

Trends in Unintended Pregnancy

A key question is whether unintended pregnancy is becoming more or less prevalent. The data available to answer this question are limited because of the underreporting of abortion within the NSFG, as noted above. Forrest and Singh (1990) supplemented the 1982 and 1988 NSFG data on births with more complete information on abortions and found that the overall incidence of unintended pregnancy increased very slightly between 1982 and 1987, from 55.5 to 57.3 percent. Over those same years, however, there was a larger increase in the proportion of births resulting from unintended pregnancies, primarily because of a small decrease in the proportion of unintended pregnancies that were aborted.⁸ Put another way, given an unintended pregnancy, slightly fewer women obtained abortions in 1987 than in 1982, resulting in more children being born who were unintended at the time of conception (Forrest and Singh, 1990). Similar data are not available after 1987 on this complex relationship among pregnancies, abortions, and births.

However, using data from the NSFG, it is possible to track trends in births derived from unintended pregnancies, both those that were mistimed and those that were unwanted at conception, as summarized in the next several sections. The first section briefly presents overall trends; subsequent, more detailed sections present trends by marital status because, as noted above, the marital status groupings used by the NSFG in its fertility questions have changed over the years.

Overall Trends for All Women

Data from the large pool of ''ever-married women" show that, broadly speaking, the proportion of births resulting from unintended pregnancies decreased steadily in the 1970s through to 1982. Between 1982 and 1988, however, the picture changed. Looking at births to women aged 15–44 in all marital status groupings (Figure 2-4), the percentage of births unintended at conception increased (from 37 to 39 percent). Data from the 1990 NSFG telephone reinterview suggest that the increase continued through to 1990 (rising from 39 to 44 percent). That is, 44 percent of all births in 1990 were the result of unintended pregnancy. Preliminary data from the 1993 National Survey of Families and Households (Wave 2) indicate that the increase may extend into the early 1990s as well (Figure 2-5). Given the serious consequences that can burden births derived from unintended pregnancy (Chapter 3), this overall trend is troubling.

Figure 2-4

Intention status at conception of recent live births (births within exactly 5 years of interview date or within 2 years of interview date for 1990) to all women ages 15–44: United States, 1982 to 1990. Note: Because of rounding, totals may not (more...)

Figure 2-5

Intention status at conception among births in 1986–1988 (National Survey of Family Growth) and 1988–1993 (National Survey of Families and Households, Wave 2), among non-Hispanic white women ages 25–44. Source: Bumpass L. Unpublished (more...)

Trends Among Currently Married Women

Data on the intention status of pregnancies that result in live births to currently married women cover the longest time period—from 1965 to 1990—and are given in Figure 2-6. The overall proportion of births unintended at conception decreased steadily from 1965 to 1982 (from 55 to 30 percent). In particular, births unwanted at conception dropped dramatically over this period, decreasing from 24 percent in 1965 to 7 percent in 1982. Between 1982 and 1988, however, the percentage of births derived from unintended pregnancies among these currently married women increased, from 30 to 33 percent, largely because of an increase in births unwanted at conception. Between 1988 and 1990, the proportion of births unintended at conception increased again, to 37 percent, this time entirely because of an increase in births mistimed at conception. The broad trends displayed in Figure 2-6 appear to be statistically significant, at least at the extremes (e.g., 1982 versus both 1965 and 1990). Trends in unintended pregnancy among the slightly larger group of "ever-married" women (data not shown), which includes those who are separated, divorced, or widowed, follow the same patterns seen among currently married women.

Figure 2-6

Births in the preceding 5 years among currently married women aged 15–44 at interview, by intention status at conception, 1965–1990. Note: In calculating percentages, it is assumed that missing data are distributed proportionately across (more...)

Trends Among Never-Married Women

As one might predict, a larger proportion of births to unmarried women (although certainly not all) are unintended at the time of conception than is the case for currently married and ever-married women; but interestingly, trends in births derived from unintended pregnancies among these women follow a different pattern. Between 1982 and 1990, both currently married and ever-married women experienced a steady increase in births unintended at conception, as noted above. By contrast, the percentage of births to never-married women derived from unintended pregnancy decreased between 1982 and 1988, and then appeared to increase between 1988 and 1990. Changing proportions of births derived from mistimed and unwanted pregnancies fueled these fluctuations (Figure 2-7).

Figure 2-7

Percentage of births in the preceding 5 years to never-married women aged 15–44 by intention status at conception, 1982–1990. Source: Centers for Disease Control and Prevention, National Center for Health Statistics. National Survey of (more...)

Understanding the intention status of births to never-married women is particularly important given the increasing public policy focus on both nonmarital childbearing and childbearing among unmarried teenagers, as noted in Chapter 1.⁹ One important question in this policy arena is whether births to these two groups are consciously undertaken (intended) or not. In fact, available data show that the overwhelming majority of births to both never-married women generally and to unmarried teenagers in particular were the result of unintended pregnancies: in 1990, 73 percent of births to never-married women were the result of unintended pregnancies (Figure 2-7; Williams and Pratt, 1990), and in 1988, 86 percent of births to unmarried teenagers (i.e., under 20 years of age) were unintended at conception (The Alan Guttmacher Institute, 1994:Figure 32). Clearly, most births to women who have never been married and the vast majority of births to unmarried teenagers are not the result of conscious intent.

Trends in Births Unwanted at Conception

The broad trends noted in the preceding sections mask some subgroup differences that are important to underscore, especially as regards births from unwanted pregnancies, which carry particularly serious risks (Chapter 3). Table 2-3 shows data on the intention status of births to ever-married women (which excludes the important group of never-married women) broken out by both income and by race (black, white) over the period 1973–1988. The intention status of births among ever-married women below the poverty level followed the same pattern as that of all ever-married women from 1973 to 1988; that is, the proportion of births derived from unintended pregnancies decreased from 1973 to 1982 and then increased in 1988.

TABLE 2-3

Intention Status at Conception of Births in the Last 5 Years to Ever-Married Women and to Ever- Married Women with Incomes Below the Poverty Level, by Race, 1973–1988 (in Percent).

Following trends in unwanted pregnancies, however, reveals more dramatic changes. Although births from unwanted pregnancies among all ever-married women showed a slight increase between 1982 and 1988 (increasing from about 8 to 10 percent), births among women below poverty in this group registered much steeper increases (rising from about 12 to 21 percent). This means that by 1988 one of every five births to ever-married women below the poverty level was the result of an unwanted pregnancy—a level approaching the 1973 level of one in four, with most of the progress in the interim having effectively been erased. For black women below poverty, the trend is more pronounced. In 1973, 44 percent of births among this group resulted from an unwanted pregnancy; by 1982, this percentage had dropped to 21 percent. By 1988, however, this number was sharply up again, to 35 percent; thus, in 1988, about one birth in three to ever-married black women below poverty was due to an unwanted pregnancy.

As noted above, these figures exclude data for never-married women, who are known to have very high proportions of births that were either mistimed or unwanted at conception (see Figure 2-7). Accordingly, it is reasonable to believe that the proportion of births unwanted at the time of conception among all women below poverty is probably even higher than the levels noted in the preceding paragraph, which is addressed only to those below poverty who have ever been married.

The Role of Abortion

Reflecting the high proportion of pregnancies that are unintended, U.S. abortion rates are also high. That is, the nation's high abortion rate can be viewed as reflecting high levels of unintended pregnancies. More than a million and a half abortions occurred in the United States every year during the 1980s; in both 1991 and 1992, the total number was about 1.5 million (Figure 2-8). As already shown in Figure 2-2, over half of all unintended pregnancies end in abortion.

Figure 2-8

Number of reported abortions and rate of abortions per 1,000 women aged 15–44, United States, 1973–1992. Source: Henshaw SK, Van Vort J. Abortion services in the United States, 1991 and 1992. Fam Plann Perspect. 1993;26:100–106. (more...)

International Comparisons

International comparisons help to put many of these numbers into perspective. Data from other industrialized democracies demonstrate that the rate of unintended pregnancy experienced among teenagers in the United States is considerably higher than that in other countries and is often higher for adults as well (Jones et al., 1988, 1985). Figure 2-9 compares the level of intended and unintended pregnancies in the United States with the levels in a number of other industrialized nations.¹⁰ It is apparent from Figure 2-9 that the rate of intended pregnancy¹¹ was fairly constant across countries, ranging only between 1.18 and 1.39 pregnancies per woman. However, there is substantial variation across nations in the unintended pregnancy rate, with only France having rates of unintended pregnancy comparable to those found in the United States. Levels of 0.28 in The Netherlands, 0.63 in Great Britain, 0.79 in Canada, and 0.80 in Sweden illustrate that a variety of nations have achieved rates of unintended pregnancy that are considerably less—some a great deal less—than the level of 1.31 in the United States. The information in Figure 2-9 is obviously dated; nonetheless, it clearly suggests that lower rates can be achieved in the United States.

Figure 2-9

Total pregnancy rate and its components, the intended pregnancy rate and the unintended pregnancy rate, selected countries, selected years. Note: The total pregnancy rate is the average number of pregnancies (i.e. births plus abortions) that a woman would (more...)

U.S. Abortion Rates Versus Other Western Democracies

Although the U.S. abortion rate has declined slightly since the early 1980s (Figure 2-8), the rate remains high compared with that in other Western democracies.¹² Table 2-4 depicts the abortion rate (number of abortions per 1,000 women aged 15–44) for a number of developed industrialized nations. Compared with those nations, the U.S. abortion rate is two to four times higher. Few other Western democracies have abortion rates that even approach those of the United States; the U.S. rate is more than 50 percent higher than the Norwegian rate, for example. The abortion rate for U.S. whites is also presented to confirm that the higher U.S. rate is not explained by the higher incidence of abortion among minority women in the United States. The data show that the U.S. abortion rate is higher than the rate found in other Western nations among whites as well as among all U.S. women.

TABLE 2-4

Abortion Rates per 1,000 Women Ages 15–44 by Country, 1980, 1985–1991.

A detailed study conducted by The Alan Guttmacher Institute comparing the United States with four other comparable areas illustrates the degree to which the U.S. abortion rate reflects the high proportion of pregnancies that are unintended. In the United States, women were estimated to have a total pregnancy rate (defined in Footnote 11) of 2.6—that is, an average of 2.6 pregnancies during their reproductive years. This represents a considerably higher total pregnancy rate than that found in the other study areas: 1.7 in The Netherlands, 1.7 in Quebec, 2.0 in Great Britain, and 2.1 in Ontario.

The higher total pregnancy rate in the United States is due entirely to the higher rate of unintended pregnancies (births resulting from unintended pregnancies plus abortions). And even though many unintended pregnancies are resolved by abortion, the rate of births derived from unintended pregnancy is still higher in the United States than in any of the other study areas. Moreover, compared with the four study countries, the U.S. abortion rate remains two to four times higher, as in the earlier comparison (Table 2-4).

Conclusion

Unintended pregnancy is common in the United States; the most recent estimate is that almost 60 percent of all pregnancies are unintended, either mistimed or unwanted altogether. Reflecting the high proportion of pregnancies that are unintended, abortion rates are also high; about half of these unintended pregnancies are aborted. Although unintended pregnancies occur among women of all socioeconomic, marital status, and age groups, unmarried and poor women as well as women at either end of the reproductive age span are especially likely to become pregnant unintentionally.

During the 1970s and early 1980s, a decreasing proportion of births (as distinct from pregnancies) were unintended at the time of conception; between 1982 and 1988, however, this trend reversed such that the proportion of births that were unintended at conception increased. These trends in the 1980s reflect both an overall increase in the proportion of unintended pregnancies and a decline in the proportion of unintended pregnancies that were resolved by abortion. Some data from the 1990s suggest that the proportion of births that were unintended at conception has continued to increase into this decade. In 1990, 44 percent of all births were the result of unintended pregnancy; among such subpopulations as women in poverty, black women, never-married women, and unmarried teenagers, the percentage of births resulting from unintended pregnancy was substantially higher (59, 62, 73, and 86 percent, respectively, in the late 1980s).

International comparisons show that U.S. women experience more unintended pregnancies—and therefore more abortions as well as births—than women in many other industrialized countries. Such data suggest that lower rates of unintended pregnancy could be achieved in the United States.

References

• The Alan Guttmacher Institute. Sex and America's Teenagers. New York, NY: and Washington, DC; 1994.
• The Alan Guttmacher Institute. Unpublished data. 1994.
• Armstrong KA, Kenen R, Samost L. Barriers to family planning services among patients in drug treatment programs. Fam Plann Perspect. 1991;23:264–271. [PubMed: 1786807]
• Bumpass L. Unpublished data tabulated for the Institute of Medicine Committee on Unintended Pregnancy, October, 1994.
• Canadian Center for Health Information. Therapeutic Abortions, 1991. Ottawa, Ontario: Statistics Canada;1993.
• Centers for Disease Control, National Center for Health Statistics. National Survey of Family Growth 1982 and 1988. Telephone Reinterview; 1990. (preliminary data).
• Centers for Disease Control and Prevention. Health risk behaviors among adolescents who do and not attend school—United States, 1992. MMWR. 1994;43:129–132. [PubMed: 8309460]
• Forrest JD. Epidemiology of unintended pregnancy and contraceptive use. Am J Obstet Gynec. 1994;170:1485–1488. [PubMed: 8178895]
• Forrest JD, Singh S. The sexual and reproductive behavior of American women, 1982–1988. Fam Plann Perspect. 1990;22:206–214. [PubMed: 2272379]
• Henshaw SK, Forrest JD. Women at Risk of Unintended Pregnancy, 1990 Estimates: The Need for Family Planning Services, Each State and County. New York, NY: The Alan Guttmacher Institute; 1993.
• Henshaw S, Morrow E. Induced Abortion: A World Review, 1990 Supplement. New York, NY: The Alan Guttmacher Institute; 1990.
• Henshaw SK, Van Vort J. Abortion services in the United States, 1991 and 1992. Fam Plann Perspect. 1993;26:100–106. [PubMed: 8070545]
• Jones EF, Forrest JD. Underreporting of abortion in surveys of U.S. women: 1976 to 1988. Demography. 1992;29:113–126. [PubMed: 1547898]
• Jones E, Forrest JD, Goldman N, Henshaw SK, Lincoln R, Rosoff J, Westoff CF, Wulf D. Teenage pregnancy in developed countries: Determinants and policy implications. Fam Plann Perspect. 1985;17:53–62. [PubMed: 3842654]
• Jones E, Forrest JD, Henshaw S, Silverman J, Torres A. Unintended pregnancy contraceptive practice and family planning services in developed countries. Fam Plann Perspect. 1988;20:53–67.
• Kost K, Forrest, JD. Intention status of U.S. births in 1988: Differences by mother's socioeconomic and demographic characteristics. Fam Plann Perspect. 1995;27:11–17. [PubMed: 7720847]
• Luker K. Taking Chances: Abortion and the Decision Not to Contracept. Berkeley: and Los Angeles, CA: University of California Press; 1975.
• Miller WB. Relationships between the intendedness of conception and the wantedness of pregnancy. J Nerv Ment Dis. 1974;159:396–406. [PubMed: 4436664]
• Musick JS. Young, Poor and Pregnant: The Psychology of Teenage Motherhood. New Haven, CT: Yale University Press; 1993.
• Piccinino L. Unintended pregnancy and childbearing in the United States: 1973–1990. Advance Data from Vital and Health Statistics. Hyattsville, MD: CDC, NCHS; Forthcoming.
• Poole VL, Klerman LV, Goldenberg RL, Cliver SP. Pregnancy intendedness and maternal behaviors in a low-income high risk population. Unpublished manuscript. University of Alabama at Birmingham; 1994.
• Ryder NB. The structure of pregnancy intervals by planning status. Popul Stud. 1985;39:193–212.
• Torres A, Forrest JD. Why do women have abortions? Fam Plann Perspect. 1988;20:169–176. [PubMed: 3243347]
• United Nations. Abortion Polices: A Global Review, Vol. I. New York, NY: Department of Economic and Social Development, United Nations; 1992.
• United Nations. Abortion Policies: A Global Review, Vol. II. New York, NY: Department of Economic and Social Development, United Nations; 1993.
• Westoff CF. The decline of unplanned births in the United States. Science. 1976;191:38–41. [PubMed: 1246597]
• Williams LB. Determinants of couple agreement in US fertility decisions. Fam Plann Perspect. 1994;26:169–173. [PubMed: 7957819]
• Williams LB, Pratt WF. Wanted and unwanted childbearing in the United States: 1973–1988. Advance data from Vital and Health Statistics, no. 189. Hyattsville, MD: National Center for Health Statistics; 1990.

Footnotes

1

The 1988 National Maternal and Infant Health Survey also collected data on intention status. Although these data are briefly referred to in this report, they were not relied on as heavily as the NSFG because they are not part of a long-standing series and therefore cannot provide the perspective of trends over time.

2

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.

3

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.

4

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).

5

The concept of being "at risk" of unintended pregnancy is a complicated one with several definitions. For example, much of the research on contraceptive use summarized in Chapter 4 uses a definition in which only women who have been sexually active in the last 3 months are included in the pool of at risk, whereas the definition used here refers to sexual activity at any previous time. Moreover, other researchers suggest that even those women who have never had intercourse at all should sometimes be considered at risk. For example, young adults who have never had sexual intercourse but who are involved in increasingly intimate relationships may, in fact, be at risk of unintended pregnancy.

6

Part of the unintended pregnancy phenomenon in adolescence and the early 20s may reflect the decreasing age of menarche and the rising age of first marriage, both of which, taken together, have lengthened the period during which pregnancies are particularly likely to be considered unintended. In 1890, for example, the average number of years between menarche and marriage was 7.2 years (menarche at about 15.5 years of age and marriage at about 22 years of age); in 1988 the interval was almost 12 years (menarche at about 12.5 years and first marriage at about 24 years of age) (The Alan Guttmacher Institute, 1994:Figure 1).

7

Data are reported here on births rather than pregnancies because the research summarized in Table 2-2, which is the source for the narrative text in this section, did not include differential rates of unintended pregnancy by race, but only by marital status, age, and poverty level.

8

Possible reasons for this phenomenon include the passage by states of restrictive abortion laws, decreased numbers of facilities performing abortions, changing views about the acceptability of abortion as a solution to unintended pregnancy, and increased social tolerance for nonmarital childbearing.

9

These two phenomena—nonmarital childbearing and teen childbearing—are closely related but not synonymous; in fact, most births among unmarried women (70 percent) occur among women who are not teenagers.

10

If all unintended pregnancies were prevented, it is not the case that the total fertility rate would be as low as is implied by the intended pregnancy rate, because many unintended pregnancies are mistimed and would occur later as intended pregnancies.

11

The intended pregnancy rate is derived from the total pregnancy rate. The total pregnancy rate is the average number of pregnancies (i.e., births plus abortions) that a woman would have, assuming a continuation of current age-specific rates over her reproductive lifetime. This rate can be broken down into the planned (or intended), unplanned (or unintended), and abortion rates by applying the distribution of births by planning and intention status to the levels of fertility and abortion that prevailed in the year following the survey.

12

The term "Western democracies" is used in this section to clarify that the comparisons being made are between the United States and selected European countries, plus Canada. Other nations, such as China, the former Soviet Union, countries in Eastern Europe, and some developing countries report ratios of abortions to live births that are significantly higher than those in either the United States or in other Western democracies.