Contraceptive use in developing countries has been studied for decades, first to measure the proportions of women practicing contraception and using specific methods, and to identify the personal characteristics associated with use. Later studies developed algorithms to determine the relationship between contraceptive use and fertility levels; others tried to estimate the effect of access to services by calculating unmet need— the proportion of women who want to stop bearing children or to space their births but are not using amethod. On the basis of this research, the higher fertility among indigenous couples than among nonindigenous couples is generally thought to reflect a desire for more children, who represent more helping hands in the fields and the home. And although contraceptive use is expected to rise with women’s and men’s education, neither the difference in spouses’ educational attainment nor their combined levels of education have been seen as driving the final decision on whether to practice contraception or what method to use. Moreover, when contraceptive use rises steadily over time, fertility is usually expected to decrease proportionately. But things aren’t always that simple, as several analyses in this issue illustrate.

In the lead article, Catherine McNamee seeks to identify the reasons underlying the difference in the total fertility rates (TFRs) of indigenous couples (4.3 children per woman) and nonindigenous couples (3.1 children) in Bolivia [page 166]. According to data from the 2003 Demographic and Health Survey, unwanted fertility explains virtually the entire difference in total fertility. The data also show a substantial gap in unmet need between the two groups (26% vs. 19%, respectively). In logistic regression analyses of couples in which the woman wanted nomore births in the next two years, women’s use of modern contraceptives was not linked to their perception of whether their partner wanted more or fewer children than they did; however, compared with women who thought that they and their partner wanted the same number of children and those who thought their husband approved of family planning, those who said they did not know their partner’s fertility desires or did not know if he approved of contraception had significantly lower odds of use. Nevertheless, this model explained very little of the difference in use between indigenous and nonindigenous couples. According to the author, this suggests that structural barriers, such as affordability and cost, are the most important factors in the decision to use modern methods.

Between 1996 and 2006, use of modern contraceptives rose from 26% to 44% in Nepal. During this period, reliance on female sterilization remained the dominant method, but use of hormonal methods such as the injectable and the pill rose and use of male sterilization remained stable. Bina Gubhaju uses data from three Demographic and Health Surveys conducted during the decade to examine how shifts in the educational levels of husbands and wives may have influenced couples’ method choices [see article]. In general, the wife’s level of education was associated with the type of method used, but the husband’s educational attainment had more influence on the use of male sterilization and the condom. For example, men with any secondary or higher education were more likely than those with no schooling to rely on either of these methods. Moreover, couples in which the husband had at least six more years of education than the wife were more likely than those in which neither spouse had any schooling to use those methods. However, differences by educational level in the use of specific methods of family planning narrowed considerably over the study period. The author suggests that, as educational levels and contraceptive use are both increasing, a better understanding of how spouses’ relative levels of education influence method choice is necessary if the country’s family planning program is to provide for the population’s needs.

In a Comment, Kanako Ishida and colleagues examine the startling case of Ecuador, where the proportion of women using modern contraceptives rose from 57% to 73% between 1994 and 2004, while the total fertility rate (TFR) declined only slightly, from 3.6 to 3.3 births per woman, during the same period [see article]. The proportions of women in union, the length of partpartum infecundability and the proportions of births that were unwanted or mistimed remained relatively stable. However, despite the overall slight decline in fertility, two subgroups of women stood out as exceptions; the TFR among women with 0–5 years of education fell from 5.7 to 4.9 births per woman, and the rate among women in the Sierra region dropped from 3.9 to 3.1, mainly reflecting a decrease among rural women in the region from 4.8 to 3.7. According to the authors, the absence of a strong association between fertility rates and contraceptive use, and the small gap between ideal fertility and actual fertility (0.7–0.9 births) likely indicates that Ecuadoran women are adopting contraceptives to space rather than limit their births and that further fertility decline is unlikely without a decrease in ideal family size.

Also in This Issue

In a study conducted in low-income areas of three cities in Ethiopia, Annabel Erulkar and Abebaw Ferede looked for factors associated with early (before age 15) sexual initiation or unwanted sexual debut in an especially vulnerable population—out-of-school girls aged 10–19 [see article]. Of the nearly 2,000 respondents, the great majority were aged 15–19 and were migrants. One-third had had no education, and similar proportions had lost at least one parent or were socially excluded—defined as a lack of community support networks, friends or participation in a club or other group. Nearly seven in 10 respondents had ever worked for pay; compared with those in other kinds of work, domestic workers tended to be younger (24% vs. 2% aged 10–14), uneducated (49% vs. 13%) and more likely to bemigrants (98% vs. 52%). Overall, 23% of respondents were sexually experienced; for 27% of these, first sex had occurred before age 15 and for 29%, it had been coerced. Those whose first sex had been early or coerced had had significantly less education than other girls. Being amigrant, being a domestic worker and being socially excluded were all associated with both early and coerced first sex. Among sexually experienced 15–19-year-olds, education was negatively associated with both early and coerced sex, and social exclusion was positively associated with coerced sex.

In Nigeria, unsafe abortion is common and a major cause of maternal mortality, but policymakers have done little to address the problem. Of 49 politicians and officials interviewed in-depth by Friday Okonofua and colleagues in 2008, 36 knew that abortion is illegal except to save the life of the woman, but none knew the specific provision of the law [see article]. Few knew the annual number of abortions or abortion-related maternal deaths, although many knew of women who had died or almost died from an unsafe abortion. More than one third of respondents believed abortion should not be legal under any circumstances, while one-fifth supported legalizing abortion on medical grounds and the same proportion supported legalization on the grounds of rape and incest. The strategies they recommended to reduce maternal mortality included improving access to contraceptives, providing sex education, strengthening the health system and providing free pregnancy care. The authors conclude that intense public health education and advocacy targeting policymakers are needed to increase the political will to reduce abortion-related maternal mortality in Nigeria.

—The Editors