We offer here an updated estimate, as of 2000, of the unmet need for contraception in the developing world and in the former Soviet republics. The exercise is limited to numbers and proportions of women in need, as estimated from survey data. We focus on married and unmarried women, on youth aged 15-19 and 20-24, and on the implications of broadened definitions of unmet need.
METHODS AND DATA SOURCES
Surveys conducted since 1990—predominantly as part of the Demographic and Health Survey (DHS) series or by the Division of Reproductive Health of the U.S. Centers for Disease Control and Prevention—cover women in most of the developing world. Of the 55 surveys used for this analysis, 17 took place from 1990 to 1994, and 38 in 1995 or later (see appendix). A total of 115 countries are covered in the estimates, including all developing countries with a population of more than one million, except those where unmet need by the DHS definition* is negligible: China (including Hong Kong), Singapore, Taiwan and the Republic of Korea. For countries without surveys, we assumed that the proportion of women with unmet need was equal to the average (unweighted) proportion with unmet need among countries in the same region for which information was available. (Countries with surveys account for 85% of unmet need in Asia, 90% in the Central Asian republics, 67% in Latin America, 52% in North Africa and the Middle East, and 83% in Sub-Saharan Africa.) Overall, 80% of estimated global unmet need comes from countries with information, since most large countries have recently been surveyed.
Wherever possible, we developed separate estimates for currently married (including cohabiting), never-married and previously married women, and for women aged 15-19 and 20-24, as well as those aged 15-49. To estimate the numbers with unmet need, we applied the proportions of women with unmet need to the number of women in each subgroup by age or marital status.
Global and Regional Estimates
An estimated 113.6 million women in the developing world have an unmet need for contraception—105.2 million married women (of whom 55.4 million wish to space births and 49.8 million wish to limit further childbearing) and 8.4 million unmarried women (Table 1). In addition, an estimated nine million women in other regions have an unmet need: 4.6 million in Russia, 3.6 million in eastern Europe, 0.6 million in the Caucasus and 0.2 million in the Baltic republics. Thus, a total of 122.7 million women in developing countries and the former Soviet republics have an unmet need for contraception. (These figures omit users of traditional methods, who are quite numerous in the former Soviet Union and elsewhere.)
Asia contains 61 million married women with unmet need, or 58% of the total for the developing world, reflecting the inclusion of several countries with very large populations (India, Indonesia, Pakistan and Bangladesh). Sub-Saharan Africa contains 24 million (22% of the total), mostly because of the large populations of Nigeria, Ethiopia, South Africa and the Democratic Republic of the Congo. Latin America contributes 11 million married women with unmet need (11%), nearly half of whom live in Mexico and Brazil. North Africa and the Middle East account for only about eight million (8%), and the Central Asian republics, with their smaller populations, have a total of 1.1 million (1%).
In the developing world as a whole, the prevalence of unmet need among married women is about equally accounted for by women who wish to space and those who wish to limit births, but important variations exist. In Sub-Saharan Africa, 65% of unmet need is for spacing; in Latin America, by contrast, only 42% is (not shown). In Asia, spacing and limiting needs are nearly equal. Such differences call for different kinds of contraceptive supplies and different budgetary considerations.
Unmarried women add to unmet need, accounting for 7% of the developing world total (not shown). By region, however, this proportion varies widely—from 4% in Asia to 16% in Sub-Saharan Africa. Low proportions reflect that unmarried women tend not to be sexually active or tend to practice contraception if they are having intercourse. Only for Sub-Saharan Africa is information available separately for never-married women (2.6 million) and previously married women (1.9 million). In that region, 85% of unmet need among never-married women is for spacing, since most of these women are young. Among previously married women, who are mainly older and have had more children, only 42% of unmet need is for spacing. Contraceptive use among adolescents, which is surprisingly high in some countries (15% in Botswana, 23% in Togo and 27% in Cameroon), somewhat reduces unmet need among unmarried women.
The proportion of currently married women with unmet need in various regions of the developing world in 2000 ranges from 11% to 24% (Table 1). Sub-Saharan Africa's figure of 24% is about half again the average for the developing world overall (17%); other regional figures range from 11% in the Central Asian republics to 14% in Latin America and 16% in North Africa and the Middle East and in Asia. The most recent estimate for three of the four regions is well below the estimate for 1996 (see below). Country-specific declines identified in recent surveys reflect growth in the proportions of couples using modern methods, which shrinks the residual group still having an unmet need.
For unmarried women in the developing world, the proportion with unmet varies by region around the overall figure of 3%, ranging from 1% to 10%.
Comparisons with Earlier Estimates
Other estimates of unmet need have often cited 120 million, referring just to women who are currently married, and just to the developing world. That figure is probably based on a computation in the early 1990s that extrapolated population-weighted data from 50 surveys in developing countries to other countries.1 Other work estimated an upper limit of 150 million,2 but it included all traditional method users, as well as an estimate for the entire Soviet Union that also included traditional method users;3 in effect, this work estimated unmet need for modern contraception. A much broader definition was developed to include the usual DHS categories and women who especially need protection against pregnancy because their age, high parity or short birth intervals would increase their risk of poor outcomes.4 This change considerably elevated the proportions of women with unmet need in 28 countries where DHS conducted surveys from 1985 to 1990. Another study used the 1987 Sri Lanka DHS to explore a variety of health-based definitions, obtaining greatly enlarged proportions with unmet need under certain definitions.5
For currently married women in the developing world, the estimate of 105.2 million with unmet need represents only a small increase in recent years. An estimate prepared from 45 similar surveys in 1996, excluding not only China but also Central Asia, showed 101.7 million married women with unmet need (Table 2).6 When Central Asia is excluded from our total, the current estimate is 104.1 million, or just 2% more than the 1996 estimate. The difference reflects considerable growth in the population base, since the proportions with unmet need have declined, as documented in repeated surveys within numerous countries.7 (As we have noted, these declining proportions are assigned also to countries lacking surveys, under the assumption that they share the known trend.) Population growth has been substantial: According to the UN estimates, in the developing world outside of China, the number of women aged 15-49 grew by 13% between 1995 and 2000.
Overall, however, the proportion in need fell from 19% to 17%. Declines have had to be quite substantial, especially in the larger countries, to more than offset the growth in the number of women of reproductive age. India, for example, showed a decline in unmet need from 16.5% to 15.8% between the 1992-1993 and 1998-1999 national surveys. That translates to a reduction of more than 2.5 million women. Another, minor influence that might tend to lower the proportion with unmet need is that the UN has reduced its estimates of the number of women aged 15-49. However, the reductions are very small—less than 1% between the 1996 and 1998 estimates for the developing world outside of China, only 4% for Sub-Saharan Africa and nearly zero for the other regions (except Latin America, which showed a slight increase). The UN estimates for 2000 show a further reduction of 1% in the number of women 15-49.
Estimates for Women Aged 15-19 and 20-24
Because of the special importance of young women, we have extracted data on those aged 15-19 and 20-24 for special attention. Information is available for currently married women in these age-groups in all regions (Table 3), but data on unmarried women at those ages are severely limited, except in Sub-Saharan Africa. (There, among never-married women, the proportion with unmet need is 7% at ages 15-19 and 11% at ages 20-24; among previously married women, the proportions are 15% and 16%, respectively.)
These two age-groups account for 33% of unmet need among married women (not shown), or 34.9 million women. The proportion falls well below the average in North Africa and the Middle East (23%), as well as in the Central Asian republics (28%). For Sub-Saharan Africa, the figure is 31%; for Latin America, 31%; and for Asia, 35%. Young married women represent a surprisingly large group and deserve careful program attention. They tend to be at low parities, but many are interested in limiting births.
Teenagers outnumber women in their early 20s but are less likely to be married or cohabiting; therefore, the 20-24 age-group contains twice the number in need (23.5 vs. 11.4 million—Table 3). The differential is even more extreme in North Africa and the Middle East and in Central Asia. It is slightly less marked in Sub-Saharan Africa because cohabitation begins earlier there and the entire 15-19 age-group is considerably larger in relation to the 20-24 age-group than it is elsewhere.
Within each region, the two young age-groups have similar proportions with an unmet need, except in Latin America (Table 3). When regions are compared, the combined proportion for the two age-groups in Central Asia (16%) is similar to that in North Africa and the Middle East (18%), and the combined proportion in Asia (23%) is similar to that in Sub-Saharan Africa (26%).
From Tables 1 and 3, we see that for both young married women and all married women, Sub-Saharan Africa has the highest proportion with unmet need—about one in four for both groups. However, the other regions show differences: Young women have more unmet need by a considerable margin in Latin America (22% vs. 14%), in Asia (23% vs. 16%) and in the Central Asian republics (16% vs. 11%), but by rather little in North Africa and the Middle East (18% vs. 16%).
All estimates of unmet need would be considerably larger under the assumption that users of traditional methods have an unmet need for modern methods. Traditional methods usually have high failure rates, resulting in numerous unwanted pregnancies, abortions, maternal deaths and births. In the former Soviet republics, where modern methods have been largely unavailable, the inclusion of traditional method users would more than double the numbers with unmet need. In eastern Europe, the number would increase from 3.6 million to 9.9 million; in Russia (whose population is about equal to that of the rest of eastern Europe), from 4.6 million to 9.6 million. In one sense, there remains an unmet need wherever contraceptive technology is faulty and users rely on defective methods for lack of better alternatives. For clarity, however, the alternative definitions used should be made fully explicit.
It may be argued that the inclusion of traditional method users is undesirable, since it implies that unmet need should relate to the use of highly effective contraception. Indeed, the DHS definition excludes women who became pregnant accidentally while using a method, since they were merely in need of better contraception, not any contraception. The counterargument is that when the technology is imperfect, resulting in unwanted pregnancies, a true residual unmet need remains. Because traditional method users are of proven motivation, are at risk of unwanted pregnancies and are numerous in many countries, they should not be ignored, especially since a poor service environment is often responsible for their lack of access to more effective methods.
An entirely different issue is that the DHS definition of unmet need omits amenorrheic women who say they wanted the last birth, regardless of their intentions regarding a future birth. Action programs wishing to offer information and services to all women who want to postpone pregnancy require inclusive estimates, especially since delayed assistance often means an unwanted early conception that results in an unsafe abortion or an unwanted child. Substantial proportions of births occur before they are wanted and soon after the previous birth, creating a high-risk situation for the mother and infant.8 Nearly all amenorrheic women say in surveys that they wish to delay their next pregnancy or birth, but among such women, the DHS definition captures only those who say the last birth was mistimed or unwanted.
Analyses of 27 DHS surveys illustrate the impact of including women with postpartum amenorrhea:9 In these 27 countries, the proportion of married women with unmet need would increase by roughly half, from 22% to 33% (Table 4). In Sub-Saharan Africa, where spacing is dominant in most countries, the increase in the unmet need estimate is quite marked, from 26% to 43%. A separate tabulation for the spacing and limiting groups confirms this pattern (not shown).
The new estimate of 122.7 million women with unmet need reflects declines in the proportions with unmet need in many countries; additional declines have probably occurred since the surveys were conducted in the 1990s (see Appendix Table 1, page 142.) A further consideration is that the latest (2000) UN estimate for women aged 15-49 in the developing world outside of China is 1% below the 1998 UN estimate used here, which would slightly reinforce a reduced global total for unmet need.
The global figure allows for unmet need among unmarried women in all regions and includes approximate figures for all former Soviet republics. It does not include users of traditional methods, even though these users are quite numerous in many countries. The alternative definitions of unmet need offered here and elsewhere would enlarge the global figure.
Also excluded are unknown numbers of persons in need of contraceptive protection in China and a few other Asian countries with near-ceiling contraceptive use and low fertility. Given China's size, it undoubtedly contains a substantial subgroup of women who are sexually active, lack contraceptive protection and do not want to become pregnant, especially if the women are unmarried adolescents. (This is true also in the United States and elsewhere in the West.) In addition, needs exist among women who have had recent abortions and are not otherwise accounted for, and among men who are not associated with female survey respondents or who, if surveyed, would report an unmet need when their female partner would not.
In surveys, some women listed as having an unmet need say they will not use a method; in one compilation of 39 national surveys,10 an average of 61% of women with unmet need (26-83% at the country level) say they do not intend to use a method. However, the same surveys make clear that these women are generally balanced by many others who want a birth within two years (and so are classified by DHS as not having an unmet need) but who say they intend to use a method within the next year.
The overall proportion of couples with unmet need has fallen slightly as contraceptive use has increased, even though desired family size is now below earlier levels. The proportion remains substantial, however, and the population base has grown. Millions are still without protection against pregnancies that they do not want.
The surveys used in this study (see Appendix Table 1) provide the proportions of women with unmet need; the numbers of women aged 15-49, 15-19 and 20-24 are 1998 UN estimates. The proportions in each marital-status category come from the UN database on marriage patterns; from U.S. Census Bureau estimates; and from Ross J, Stover J and Willard A, Profiles for Family Planning and Reproductive Health Programs, Glastonbury, CT, USA: The Futures Group International, 1999. Special tabulations of DHS surveys were made to obtain estimates of unmet need among never-married and previously married women, and among women aged 15-19 and 20-24.
• Marital status.
Estimates of unmet need for women who are currently married or in union are available from all 55 surveys. Surveys including unmarried (never-married and previously married) women are available only for Sub-Saharan Africa and for Latin America and the Caribbean; some of the latter (those conducted by the Division of Reproductive Health, U.S. Centers for Disease Control and Prevention) depart from the DHS definition by omitting unmet need among pregnant and amenorrheic women. Surveys covering unmarried women are not available for Asia (except the Philippines) or for North Africa and the Middle East. The proportion of unmarried women with unmet need is probably small in both regions, but the absolute numbers, especially in Asia, may be large. After reviewing the few fragmentary studies available and considering the rather low estimates in Latin America and the Caribbean, we decided to use an assumption that 2% of unmarried women in Asia and in North Africa and the Middle East had an unmet need for contraception. More unmarried women than this are sexually active (even by the narrow DHS standard of sexual activity in the last month), but some practice contraception, some are infecund and some want a child within two years.
• Eastern Europe and the Russian Federation.
The United Nations lists 10 countries in eastern Europe. National surveys are available for six of these: Bulgaria, Czech Republic, Hungary, Moldova, Romania and Ukraine. Applying the average unweighted proportion with unmet need in these six to three countries with no surveys (Belarus, Poland and Slovakia) produces an overall estimate of 3.6 million women with unmet need. For the 10th, Russia, comparable surveys exist only for three major cities; letting these represent the urban sector and taking the rural-urban ratio from the six other surveys produces an estimate of 4.8 million women in unmet need for all of Russia.
• Caucasus and Baltic republics.
The following proportions of women with unmet need, from the Georgia survey, were applied to Azerbaijan and Armenia (for which survey data have recently become available): 0.1% for never-married women, 4.4% for those previously married and 21.3% for those currently married or in union. Currently married women represent 98% of the final total. For the Baltic republics, the unweighted average from surveys in Latvia and Lithuania was applied to Estonia.