Performed in safe conditions with modern methods, induced abortion carries little risk to women’s health or mortality. In representative developed countries, the risk of dying is no more than 1 in 100,000 procedures, lower than the relatively low risks associated with pregnancy and childbirth in these countries (The Alan Guttmacher Institute, 1999). Conditions are not necessarily safe, however, for women in developing countries (for further discussion, see the section under changes in the health service environment in Chapter 2) and are particularly hazardous for young women because of the circumstances under which abortions among them are likely to occur. Research suggests that while young women are not disproportionately likely to have abortions, they are more likely to delay seeking them and more likely to use an unskilled abortion provider.
National vital statistics on abortion are incomplete for all but a handful of developing countries. Even when data are available, they are often not broken down by age. Official statistics, even when they are reasonably complete, exclude abortions that take place outside the health care system. Survey data, in which women are interviewed concerning their experience of abortion, are available for some countries. However, collecting data on induced abortion through surveys is notoriously difficult, often because abortion is legally restricted, but also because, regardless of its legal status, abortion is an extremely sensitive issue (Barreto et al., 1992; Malhotra et al., 2003; Singh, Henshaw, and Berentsen, 2003). As a result, data on the incidence of abortion among young women in developing countries are largely unavailable, and the data that are available probably underreport its incidence. The available body of research on other aspects of abortion in general, and among young women in particular, is also relatively limited. Many studies are based on samples drawn from hospitals, reflecting the population of women who obtain unsafe abortions but not those who succeed in obtaining a safe procedure; many of the available studies are small scale, and many are qualitative, with few providing quantitative information on numbers, prevalence, or rates of abortion.
There were only five developing countries with national data on abortion prevalence among young women in the mid- to late 1990s: Cuba, Kazakhstan, Kyrgyzstan, Turkey, and Uzbekistan. Among these, the abortion rate (per 1,000 women) among women under age 20 ranged from virtually zero in Uzbekistan to 91 in Cuba. For women ages 20-24, the range was from 18 to 88. Cuba is noteworthy for its extremely high abortion rate among women younger than age 20; at 91 per 1,000, it exceeds that of the 19 other developed and developing countries in this study (Bankole, Singh, and Haas, 1999). Information based on other types of data is also available from a few countries. A large-scale survey of young