Women who have had abortions.(Abortion and the Public Opinion Polls, part 2)

National information on the characteristics of U.S. women who have obtained abortions has come in the past from reports issued by the U.S. Centers for Disease Control (CDC). The characteristics for which data were available, aside from those relating to the abortion itself, have been limited to the woman's age, race, marital status and number of live births. Notably lacking has been any national information on her religion, income or political attitudes. Very little has been known about the characteristics of women who have obtained illegal abortions.

As noted in the previous article, two 1981 national public opinion surveys included questions asking women whether they had ever had an abortion. The ABC News/Washington Post poll of May 1981 asked 872 women, "Finally, if you don't mind me asking, have you yourself ever had an abortion?"; and the Yankelovich survey asked the same question without the apology. In both surveys, these questions were placed at the end of the interview.

A major concern in the use of surveys to study abortion is the possibility that women will conceal their abortion histories. Since it is possible to estimate the proportions of women in various age-groups who have ever had a legal abortion, the accuracy of response to the abortion questions in the surveys can be assessed. As Table 1 shows, in the Yankelovich poll, 13 percent of respondents aged 18-44 who answered the abortion question reported having had a legal or illegal abortion, as compared with an estimated 18 percent of all U.S. women 18-44 who have had a legal abortion since 1967. (*) Therefore, if the respondents are typical of all women in that age-group, 73 percent of those who had ever had a legal abortion acknowledged the abortion to the interviewer. In the ABC News/Washington Post survey, the proportion was much lower, 52 percent. In both surveys, women aged 18-24 provided more complete abortion reports than did women aged 25-44. This may reflect more openness about abortion among the younger wo men. (+)

The difference between the polls in the accuracy of abortion reporting may have resuited in part from a difference in the question wording. The ABC/Post question opened on an almost apologetic note, implicitly acknowledging that women may prefer not to report their abortion experience. In addition, the ABC/Post interview contained a number. of questions about religion, morality and the Moral Majority which might have made some women reluctant to admit having had abortions. (Other factors that may have contributed to the difference are the quality of the interviewing and random error.)

Both polls had relatively low rates of refusal to answer the abortion question-one percent in the Yankelovich poll and three percent in the ABC/Post poll. The demographic characteristics of the women who refused as well as common sense suggest that the refusals may have come disproportionately from women who had had abortions. If refusals were treated as affirmative responses to the abortion questions, the percentages reporting accurately would rise to 78 percent and 59 percent. However, no such attribution was made in the present analysis.

Table 2 shows the percentage of women in the combined sample who report that they have ever had an abortion, according to various characteristics of the woman. The calculations are based on the 1,002 women in the Yankelovich survey and the 853 in the ABC/ Post survey who answered the question about their abortion history and for whom age data are available. As is usual in commercial opinion surveys; respondents are weighted so that the distribution of the basic demographic characteristics of each sample, such as age and race, corresponds to national norms. The combined sample is large enough that when subgroups are compared, a difference of four percentage points is usually statistically significant at the 95 percent level of confidence. For respondents aged 18-34, a difference of eight percentage points usually indicates statistical significance, and for those 35 and over, a difference of only four points is usually needed.

The proportions are presented separately in Table 2 for women under age 35 and for those 35 and older. This age break was selected because the proportion of women who report having had abortions shows a discontinuity between ages 25-34 and 35-44. Most of the abortions obtained by women under age 35 would have occurred after abortion was legalized nationally, while the majority of abortions obtained by women over 35 would have been illegal. Women aged 35-44 could have obtained either legal or illegal abortions, but abortion was illegal when these women were in their teens and early 20s, the ages when most women obtain abortions. For example, a woman aged 40 in 1981 would have been 32 years old in 1973, when abortion was legalized nationally; since given present abortion rates, four-fifths of women who obtain abortions do so before age 30, (1) it is assumed that most of the abortions obtained by women 35-44 were illegal. (*)