Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Levels and Trends in Contraceptive Use This chapter describes the extent to which contraception is being prac- ticed in sub-Saharan Africa. We first review sources of data on and defini- tions of contraceptive use. We then discuss what these data show regarding levels and trends in prevalence and differentials in current use by women. Next, we look at indicators related to contraceptive use: ever use, discontinuation, contraceptive knowledge, knowledge of sources, of supply and fertility pref- erences. Finally, we briefly present evidence of use from the perspective of men. Information on contraceptive use is far from complete, given that contraceptive prevalence data based on nationwide surveys are not available for all African countries.) SOURCES OF DATA ON CONTRACEPTIVE USE There are three main sources of data on contraceptive use in Africa based on representative samples at the national level: the World Fertility Surveys (WFS), the Contraceptive Prevalence Surveys (CPS), and the De- mographic and Health Surveys (DHS). Studies to date in Africa for which However, most countries that have initiated active family planning programs have also conducted surveys; thus, if data from the existing surveys were used to obtain a regional average, it would tend to overstate the actual use of modern contraception in Africa. 19
20 FACTORS AFFECTING CONTRACEPTIVE USE data are available include nine WFS (from 1977-1982), seven CPS (1982- 1984), and 12 DHS (1986-1990) (Kendall, 1979; Rutstein et al., 19921. Six populations were included in both the WFS and the DHS programs: Cameroon, Ghana, Kenya, Nigeria, Northern Sudan, and Senegal.2 For Benin, Cole d'Ivoire, Lesotho, and Mauritania, data are available from the WFS only. Populations with only DHS data include Botswana, Burundi, Liberia, Mali, Togo, Uganda, Zimbabwe, and the state of Ondo in Nigeria. Surveys using the CPS questionnaires have been conducted in selected sites in Zaire (four urban and two rural areas in 1982-1984) and in selected regions of Somalia (1983), but they do not provide national-level estimates and thus are omitted from most of the tables in this chapter. There are also numerous small-scale studies that have been conducted in the capital cities andlor selected regions of countries. (For a review of studies of fertility and knowledge, attitudes, and practice of contraception for Africa from 1960 to 1973, see Baum et al., 1974. For the period after 1983, see Gelbard et al., 1988.) However, no data from these have been included in the tables below, because it is doubtful that the results are representative or that the data collection techniques are comparable. Details on the year of data collection and of data included in this chap- ter are listed in Table 2-1. As the table indicates, despite the three major data collection efforts, for large parts of sub-Saharan Africa neither fertility (except from occasional censuses) nor contraceptive use data are available. This situation is most notable in Central Africa; with the exception of the Cameroon study and the 1982-1984 contraceptive prevalence survey in Zaire (subnational in scope), there is very little demographic information for this region of Africa. DEFINITION OF CONTRACEPTIVE USE In the DHS studies conducted in sub-Saharan countries between 1986 and 1990, at least half the current users of contraception in 5 of the 12 countries were relying on traditional methods, such as withdrawal and rhythm. There is little literature on the effectiveness of such methods, but they are considered much less effective than modern methods; thus the convention has emerged of differentiating between modern and traditional (or modern versus all methods) in reporting contraceptive prevalence results, especially in the context of Africa. In the three main types of studies conducted to date (WFS, CPS, DHS), methods defined as "modern" include oral contraceptives, intrauterine de 2Data for Cameroon were not available during the writing of this report. Some data were available for Nigeria (from the DHS country report) and are included in relevant tables.
LEVELS AND TRENDS IN CONTRA CEPTI VE USE TABLE 2-1 Data Sources on Contraceptive Use in Sub-Saharan Africa 21 Survey Program and Date Country wFsa cpSb DHsc Others Western Africa Benin 1981- 1982 Cote d'Ivoire 1980-1981 Ghana 1979- 1980 1988 Liberia 1986 Mali 1987 Mauritania 1981 Nigeria (Ondo State) 1986- 1987 Nigeria 1981- 1982 1990 Senegal 1978 1986 Togo 1986 Eastern Africa Burundi 1987 Ethiopia 1990 Kenya 1977- 1987 1984 1988- 1989 Malawi 1984 Rwanda 1983 Sudan (northern) 1978-1979 1989-1990 Somaliad 1983 Uganda 1988-1989 Zimbabwe 1984 1988-1989 Central Africa Cameroone 1978 1991 Zaired 1982- 1984 Southern Africa Botswana 1984 1988 Lesotho 1977 South Africa 1975- 1976 1981 Swazilandf 1988 aWorld Fertility Survey. bContraceptive Prevalence Survey. CDemographic and Health Survey. dThe data for the CPS in Zaire and in Somalia are not national. In Somalia, the survey was conducted for urban areas. In Zaire, the survey covered four cities and two rural areas. eThe DHS data for Cameroon are not available as of the writing of this report. fThe data for Swaziland are not presented in the tables. As noted in this chapter, Swaziland is not discussed in this report because of its small population size, estimated at under one million people.
22 FACTORS AFFECTING CONTMCEPTIVE USE vices (IUDs), female and male sterilizations, injections, condoms, diaphragms, and spermicides. The standard questionnaires explicitly mention two types of traditional methods: periodic abstinence (also called rhythm or the cal- endar method) and withdrawal. Any other method that the respondent con- siders to be a means of preventing pregnancy is recorded as "other." Other traditional methods include douche, herbs, and "gns-gris" (amulets, charms, or spells to prevent conception). Many of the country-specific question- na~res list the traditional methods common to their populations. As discussed in Chapter 1, two postpartum practices observed in much of sub-Saharan Africa are sexual abstinence and breastfeeding. Although these practices may result in lower fertility, in Africa they are generally motivated by the desire to ensure the survival of the youngest child and allow the mother to recuperate between births. Accordingly, neither is considered a contraceptive method in this report (see Appendix A for a discussion of the rationale). However, a high proportion of women in the African WFS and DHS did respond that they were using abstinence as a contraceptive method, although many were in fact practicing postpartum abstinence. For example, much of Togo's high level of traditional contra- ceptive use is due to a large proportion of women who reported using abstinence as a method (72.5 percent of these women were also in postpar- tum abstinence). In addition, more than half of the DHS did not include abstinence as a possible contraceptive, so we could not consider the use of abstinence as a method for all countnes.3 Although postpartum abstinence and abstinence are not considered contraceptive methods in this report, pe- nodic abstinence (i.e., rhythm) is. In sum, throughout this volume, the proportion currently using contra- ceptives refers to the proportion of women currently married or in union, aged 15-49 years, who are using a contraceptive method at the time of the survey (unless otherwise stated).4 The data are based on all women fulfill- ing these criteria, regardless of whether or not they are breastfeeding or postpartum abstinent. 3For those surveys that did include abstinence as a method, the proportion of women who reported practicing abstinence are given in a note to Table 2-2. 4There is evidence that contraceptive use is rising among women who are sexually active and unmarried. This topic is addressed for adolescents in a report by the Working Group on the Social Dynamics of Adolescent Fertility (1993) of the Panel on the Population Dynamics of Sub-Sarahan Africa.
LEVELS AND TRENDS IN CONTRACEPTIVE USE PREVALENCE OF CURRENT CONTRACEPTIVE USE IN SELECTED COUNTRIES Results from Demographic and Health Surveys Conducted from 1986 to 1990 23 Table 2-2 presents contraceptive prevalence rates (CPRs) for both cur- rent use and ever use of all methods combined,5 of modern and traditional methods separately, and of specific methods listed in the 12 DHS studies conducted in African countries between 1986 and 1990 for which data are available.6 Of the 12 countries, only two show a modern contraceptive prevalence rate of more than 30 percent: Zimbabwe (36 percent modern, 43 percent all methods combined) and Botswana (32 percent modern, 33 per- cent all methods combined). Kenya shows the next highest rate with 18 percent modern use (27 percent all methods combined). No other DHS country in Africa has even 6 percent of women of reproductive age using a modern contraceptive, although in Ghana and Togo the percentage currently using any method is greater than 10. The U.S. Agency for International Development denotes populations with 0 to 7 percent modern usage as having emergent family planning programs, and clearly the majority of the countries in Table 2-2 fall into that category (Destler et al., 1990~. Data from these 12 countries also reflect regional differences. The three higher use countries (Zimbabwe, Botswana, and Kenya) are all anglophone countries in eastern or southern Africa. In the four francophone countries surveyed, three of which are in western Africa (Mali, Senegal, and Togo), less than 3 percent of the women use modern contraceptive methods. These differences may reflect a growing divergence between southern and East Africa and West Africa. For example, data from the Swaziland Family Health Survey show prevalence rates for modern methods of almost 14 percent, suggesting that this nation may be following the path of the three high-use countries (Warren et al., 1992~.7 However, there are still too few high-prevalence countries in sub-Saharan Africa to be able to attribute dif- ferentials to geographical factors or different colonial histories. Moreover, these geographical or historical distinctions may be proxies for other deter- minants of demand for children and use of contraception discussed in Chap- ters 3 to 5, including socioeconomic factors, infant and child mortality rates, levels of female education, and the strength of family planning pro grams. SThe ever-use rates are discussed later in this chapter. 6Data for Ondo State are excluded when later data for Nigeria as a whole are available. 7Swaziland is not discussed in this report because of its small population size, estimated at under one million people.
24 FA CTORS AFFECTING CONTRACEPTIVE USE TABLE 2-2 Women in Union Age 15-49 Who Currently Use or Have Ever Used a Contraceptive Method, DHS, 1986-1990 (percent) Modern Methods Any Any Vaginal Method Modern Pill IUD Injection Method Condom Current Use Botswanaa 32.5 31.7 14.8 5.6 5.4 0.0 1.3 Burundia 6.7 1.2 0.2 0.3 0.5 0.0 0.1 Ghana 12.9 5.2 1.8 0.5 0.3 0.3 0.3 Kenya 26.9 17.9 5.2 3.7 3.3 0.4 0.5 Liberia 6.4 5.5 3.3 0.6 0.3 0.2 0.0 Malia 3.2 1.3 0.9 0.1 0.1 0.1 0.0 Nigeria 6.0 3.5 1.2 0.8 0.7 0.1 0.4 Senegala 4.6 2.4 1.2 0.7 0.1 0.1 0.1 Sudan (northern) 8.6 5.5 3.9 0.7 0.1 0.0 0.1 TOgoa 12.1 3.1 0.4 0.8 0.2 0.6 0.4 Uganda 4.9 2.5 1.1 0.2 0.4 0.0 0.0 Zimbabwe 43.1 36.1 31.0 1.1 0.3 0.0 1.2 Ever Use Botswanaa 61.7 60.1 49.4 17.4 15.0 1.3 10.1 Burundia 23.8 2.3 0.8 0.6 0.9 0.1 0.3 Ghana 37.0 22.5 14.9 1.3 10 8.3 4.5 Kenya 45.0 29.0 18.0 8.4 6.7 2.1 4.3 Liberia 18.8 15.9 13.5 2.8 1.7 1.0 1.7 Malia 7.6 3.4 2.7 0.5 0.2 0.4 0.4 Nigeria 14.0 8.4 4.8 1.7 1.9 0.6 2.0 Senegala 11.9 5.7 3.5 1.5 0.4 0.8 1.4 Sudan (northern) 25.2 18.6 17.3 1.6 1.1 0.3 1.9 TOgoa 33 0 10.1 3.8 1.6 1.1 3.7 3.3 Uganda 21.5 7.0 5.0 0.5 1.3 0.2 0.7 Zimbabwe 79.0 63.0 57.1 3.4 14.5 0.3 17.0 aAlthough abstinence is included as a traditional contraceptive method in these country surveys, it is not included as a method in this report (see this chapter for rationale). For the information of the reader, rates of abstinence for these countries follow. Current use: Botswana, 0.5; Burundi, 2.0; Mali, 1.5; Senegal, 6.7; and Togo, 21.8. Ever use: Botswana, 7.9; Burundi, 14.6; Mali, 12.7; Senegal, 23.5; and Togo, 59.0. SOURCE: Rutenberg et al. (1991) and different national DIES survey reports and data tapes.
LEVELS AND TRENDS IN CONTRACEPTIVE USE Traditional Methods Sterilization Any Periodic Other Female Male Traditional Abstinence Withdrawal Method 25 4.3 0.3 0.80.2 0.3 0.3 0.1 0.0 5.54.8 0.7 0.0 1.0 0.0 7.76.2 0.9 1.6 4.7 0.0 9.07.5 0.2 1.3 1.1 0.0 0.90.6 0.1 0.2 0.1 0.0 1.91.3 0.1 0.5 0.3 0.0 2.51.4 0.5 0.6 0.2 0.0 2.20.9 0.1 1.2 0.8 0.0 3.12.2 0.3 0.6 0.6 0.0 9.06.4 2.3 0.3 0.8 0.0 2.41.6 0.3 0.4 2.3 0.2 7.00.3 5.1 1.6 4.3 0.3 9.45.1 5.8 0.6 0.1 0.0 22.618.8 8.2 0.5 1.0 0.0 25.119.7 8.3 3.0 5.0 0.1 24.220.9 3.0 3.2 1.1 0.0 6.73.3. 3.4 1.2 0.1 0.0 5.72.8 0.8 2.1 0.3 0.0 8.14.3 2.8 2.3 0.2 0.0 8.63.4 1.2 8.7 0.8 0.0 14.912.1 4.0 1.9 0.6 0.0 29.420.2 13.1 1.0 0.8 0.0 17.413.5 4.8 3.0 2.3 0.2 48.77.2 41.1 10.9
26 FACTORS AFFECTING CONTRACEPTIVE USE In the three higher use countries, the pill is the most widely used mod- ern method, followed by the IUD, injection, and female sterilization (though the rank ordering of these other methods varies by country). Of note is the fact that 5 percent of women in Kenya and 4 percent in Botswana have undergone sterilization, contradicting the frequent assumption that this method is totally unacceptable in any sub-Saharan African country (see Chapter 4 for discussion). Only two countries report any use of vasectomy: Zimba- bwe (0.2 percent) and Botswana (0.3 percent). Chapter 5 provides further discussion of the use of modern methods in specific countries. Among traditional methods, rhythm is most commonly used, except in Zimbabwe and Botswana where the main traditional method is withdrawal. In no country is the current use of traditional methods greater than 10 percent. Trends in method mix to date suggest that as countries move from low contraceptive prevalence to higher levels, the use of traditional methods gives way to an increase in oral contraception (particularly in the 10-45 percent CPR range) followed by a rapid rise in female sterilization in coun- tries above the 45 percent level (Destler et al., 1990~. Quality issues change as the mix changes, particularly because the more permanent and effective methods (the IUD, NORPLANTR, voluntary surgical contraception) require clinical settings. National contraceptive prevalence rates tend to mask the substantial regional diversity that exists within countries. These differences are shown in Table 2-3. The national-level data for each country refer to the percent- age of women married or in union currently using a modern contraceptive method. The regional data that follow indicate the number of percentage points by which each region differs from the national average. The range between regions within a given country (i.e., the number of percentage points between the lowest and highest prevalence) varies from 5 percentage points (in Burundi and Senegal) to 30 percentage points (in Zimbabwe). In fact, the countries with the greatest interregional variations are those with higher than average national levels of prevalence, which result from the high levels of use (by African standards) in selected regions. For example, two of the regions with high positive deviations Harare and Chitungwiza, Zimbabwe, with 48.0 percent modern use, and the Central province of Kenya with 30.8 percent are in countries with the highest prevalence. However, Kampala, Uganda, shows the greatest positive deviation of all the regions (15.4 percentage points above the national level) with a prevalence rate of 17.9 percent.
LEVELS AND TRENDS IN CONTRA CEPTIVE USE TABLE 2-3 Regional Variations in Modern Contraceptive Prevalence Rates, DHS, 1986-1990 (average prevalence) Country and Region/Province Prevalence Rate Burundi Imbo Mumirwra/Mugamba Plateau Lowlands Range Ghana Western Central Greater Accra Eastern Volta Ashanti Brong Ahafo Other regions Range Kenya Nairobi Central Coast Eastern Nyanza Rift Valley Western Range Liberia Since Grand Gedeh Montserrado Rest of country Range Mali Kayes, Koulikoro Sikasso, Segou Mpoti, Gao, Tombouctou Bamako Range Nigeria Northeast Northwest Southeast Southwest Range 27 (1.2) 4.9 0.4 -0.4 -1.2 5.1 5.2 -2.0 -0.3 5.4 0.6 -1.3 1.3 0.0 -4.5 9.9 (17.9) 10.0 12.9 -3.1 1.6 -7.7 0.2 -7.9 20.8 (5 5) -1.6 -2.6 4.2 -1.1 6.8 (1.3) -0.5 -0.7 -0.5 4.7 6.8 6.0 -4.0 -4.8 2.8 9.0 13.0
28 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-3 continued Country and Region/Province Prevalence Rate Senegal West Central Northeast South Range Sudan (northern) Khartoum Northern Eastern Central Kordofan Darfur Range Togo Coast Plateau Central Kara Savana Range Uganda West Nile East Central West Southwest Kampala Range Zimbabwe Manicaland Mashonaland Central Mashonaland East Mashonaland West Matabeleland North Matabeleland South Midlands Masvingo Harare/Chitungwiza Bulawayo Range (2.4) 3.1 -1.8 -1.8 0.0 4.9 (s.5) 9.8 2.8 -3.4 -1.4 -4.1 -5.3 15.1 (3.1) 1.5 -0.7 -1.2 0.2 -2.8 4.3 (2.5) -2.5 -0.5 -0.1 o.s -1.6 15.4 17.9 (36. 1) -10.5 4.0 7.0 7.1 - 18.1 - 14.9 -0.9 -0.8 11.9 5.6 30.3 NOTE: Shown in parentheses for each country is the average prevalence. The figures for each region represent percentage point deviations from the national average. The range indicates the difference between the highest and the lowest regional rates. SOURCE: DHS national survey reports.
LEVELS AND TRENDS IN CONTRACEPTIVE USE Results of WFS, CPS, and Other Surveys, 1975-1990 29 Data on contraceptive use by method are presented in Table 2-4 for African countries participating in the WFS, CPS, and other national-level fertility or family planning surveys from 1975 to 1990. The countries that have had at least two surveys are discussed under the section on trends below. For the ten countries that have not had a subsequent DHS (Benin, Cameroon, Cote d'Ivoire, Ethiopia, Lesotho, Mauritania, Rwanda, Somalia, South Africa, and Zaire), these are the only available national-level data on contraceptive prevalence. Several findings merit comment. First, contraceptive prevalence was less than 6 percent at the time of the survey in all but two of these ten countries (the exceptions being South Africa with 48 percent in 1981 and Rwanda with 11 percent in 1983~. Although some change may have taken place since that time, the available data do not indicate any significant use of contraception (apart from South Africa, which is by no means typical of the region) among this group of countries. Second, even at these very low levels of prevalence, there was greater use of traditional than modern meth- ods in all countries except Ethiopia and South Africa. For example, Rwanda reported an overall prevalence of 11 percent, but only 1 percent of women were using a modern method. Third, among the users of modern methods, the most common method was the pill, except in Rwanda, where injection was the most common. This finding parallels the experience in other Afri- can countries, as well as numerous other developing countries in other re- gions (United Nations, 1989a). Trends in Modern Contraceptive Prevalence The analysis of trends in contraceptive prevalence in the region is lim- ited by the number of countries for which there are at least two points of observation over time. Apart from Kenya, a veritable demographic labora- tory in terms of the number and frequency of surveys conducted to date, only seven other countries have data from two time points: Botswana (1984 and 1988), Ghana (1979-1980 and 1988), Nigeria (1981-1982 and 1990), Senegal (1978 and 1986), South Africa (1975-1976 and 1981), Sudan (1978- 1979 and 1989-1990), and Zimbabwe (1984 and 1988-1989). In Zimbabwe, Botswana, and Kenya as well as in South Africa, there were dramatic increases in the use of modern methods between surveys (shown in Table 2-5), which is evidence of the growing acceptance of fam- ily planning in these countries during the 1980s. Modern prevalence rose an average of 1.2 percentage points per year in Kenya, 1.9 points per year in Zimbabwe, and 3.3 points per year in Botswana. In South Africa, it rose an average of 2.0 points per year between the two surveys.
30 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-4 Women in Union Age 15-49 Who Currently Use or Have Ever Used a Contraceptive Method, WFS, CPS, and Other Surveys, 1977-1990 (percent) Modern Methods Any Any Vaginal Country Method Modern Pill IUD Injection Method Condom Benin 5.9 0.5 0.2 0.1 0.0 0.1 0.1 Botswana 19.3 18.6 10.0 4.8 1.0 0.1 1.2 Cameroon 2.4 0.6 0.2 0.2 0.0 0.0 0.2 Cute d'Ivoire 2.9 1.6 0.4 0.1 0.0 0.0 0.0 Ethiopia 3.1 2.5 1.9 0.3 0.0 0.0 0.1 Ghana 6.4 5.5 2.4 0.3 0.1 1.6 0.6 Kenya 1977-1978 5.5 4.2 2.0 0.7 0.6 0.0 0.1 1984 14.2 9.6 3.1 3.0 0.5 0.1 0.3 Lesotho 5.1 2.5 1.2 0.1 0.2 0.0 0.1 Malawi 3.5 1.1 0.7 0.3 0.1 a o.o Mauritania 0.8 0.3 0.0 0.0 0.0 0.1 0.0 Nigeria 1.0 0.6 0.2 0.1 0.2 0.0 0.0 Rwanda 10.8 0.9 0.2 0.3 0.4 0.0 0.0 Senegal 1.3 0.6 0.3 0.2 0.0 0.0 0.1 Somalia (urban) Baydhaba 0.5 0.5 0.3 0.2 0.0 0.0 0.0 13urco 0.3 0.3 0.3 0.0 0~0 0~0 0 0 Hargeisa 2.6 0.8 0.7 0.1 0.0 0.0 0.0 Kismayo 0.5 0.5 0.3 0.0 0.0 0.2 0.0 Mogadishu 0.6 0.5 0.4 0.0 0.0 0.0 0.1 South Africa 1975-1976 37.0 35.0 1 1.0 4.0 1 1.0 0.0 2.0 1981 48 0 45.0 15.0 6.0 14.0 3 0b Sudan (northern)d 4.6 3.9 3.1 0.1 0.2 0.1 0.1 Zaire (urban) Kananga 11.3 2.6 0.6 0.1 0.3 0.0 0.3 Kinshasa 34.8 4.2 1.3 0.6 1.3 0.2 0.4 Kisangani 18.3 2.3 0.4 0.7 0.5 0.0 0.3 Lubumbashi 35.8 7.6 3.6 1.6 0.8 0.3 0.2 Zaire (rural) Nkara 16.9 2.1 1.2 0.0 0.2 0.0 0.0 Vanga 38.1 10.9 1.8 0.9 1.8 0.0 0.3 Zimbabwe 38.4 26.6 22.6 0.7 0.8 0.1 0.7 aMethod not asked. bTotal for vaginal methods and condom. CTotal for periodic abstinence and withdrawal. dNumbers are for ever-married women.
LEVELS AND TRENDS IN CONTRACEPTIVE USE 3 Traditional Methods Sterilization Any Periodic Other Female Male Traditional Abstinence Withdrawal Method 0.0 0.0 5.4 1.7 3.1 0.6 1.5 0.0 0.7 0.3 0.3 0.1 0.0 0.0 1.8 1.2 0.4 0.2 0.0 0.0 2.4 1.8 0.1 0.5 0.2 0.0 0.6 0.5 0.1 0.0 0.5 0.0 0.9 0.7 0.2 0.0 0.9 0.0 1.3 1.1 0.2 0.0 2.6 0.0 4.6 3.8 0.6 0.2 0.9 0.0 2.6 0.1 2.5 0.0 a a 2.4 0.1 0.2 2.1 0.2 0.0 0.5 0.0 0.4 0.1 0.1 0.0 0.4 0.3 0.1 0.0 0.0 0.0 9.9 8.4 1.4 0.1 0.0 0.0 0.7 0.4 0.0 0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.8 0.1 0.0 1.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.1 6.0 0.0 3.0 0.0 0.0 3.0 8.0 0.0 3.0 0.3c 0.0 0.3 0.0 0.5 0.4 0.1 0.0 1.3 0.0 2.0 1.6 0.4 0.0 0.4 0.0 21.5 13.9 6.9 0.7 0.4 0.0 7.9 4.5 2.5 0.9 1.1 0.0 11.3 8.8 1.4 1.1 0.7 0.0 8.5 8.3 0.2 0.0 6.1 0.0 8.1 3.6 4.5 0.0 1.6 0.1 9.7 0.6 6.5 2.6
32 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-5 Change in Modern Contraceptive Use Over Time in Sub- Saharan Regions Country and Region DHS/Other Botswana (1984, 1988) Urban Rural Ghana (1979-1980, 1988) Central, Western Greater Accra, Eastern Volta Ashanti, Brong Ahafo Northern, Upper Kenya (1978, 1988-1989) Nairobi Central, Eastern Rift Valley Coast Western, Nyanza Nigeria (1981 - 1982, 1990) Northeast Northwest Southeast Southwest Senegal (1978, 1986) Central Northeast (Fleuve, Oriental) South (Casamance) West (Dakar, Thies) South Africa (1975-1976, 1981) Sudan (northern) (1978- 1979, 1989- 1990) Khartoum Northern, Eastern Central Kordofan, Darfur Zimbabwe (1984, 1988-1989) Bulawayo Harare/Chitungwiza Manicaland M as hon al an d Mashonaland East (except Harare/Chitungwiza) Mashonaland West Masving Matabeleland North (except Bulawayo) Matabeleland South Midlands 18.6 37.1 24.9 3.3 5.3 11.3 6.0 8.5 1.1 18.8 9.0 4.7 5.1 2.6 1.0 0.0 0.3 1.4 1.5 0.6 0.4 0.0 0.0 2.3 35.0 3.7 17.6 2.9 7.1 1.5 26.6 17.1 377a 14.5 208 24.2 31.7 40.8 27.5 5.2 4.1 7.9 3.9 6.0 0.7 17.9 27.9 24.5 18.1 14.8 10.1 3.5 1.3 0.7 3.9 10.5 2.4 0.5 0.6 2.4 5.5 45.0 5.5 15.8 4.4 4.1 0.8 36.1 41.2 48.0 25.6 40.1 43.1 43.2 35.3 18.0 21.2 35.2 aIncludes Mashonaland East, West, and Harare/Chitungwiza. bIncludes Matabeleland South.
LEVELS AND TRENDS IN CONTRACEPTIVE USE 33 By contrast, there was relatively little increase in the use of modern methods in the four other countries that had data for at least two points: Ghana (increased from 3 to 5 percent in eight years), Nigeria (from 1 to 4 percent in eight years), Senegal (from 1 to 2 percent in eight years), and Sudan (from 4 to 6 percent in eleven years). Where modern contraceptive use has increased, one might expect a decrease in the reliance on traditional methods, perhaps due to substitution between the two. However, evidence from the three higher use countries is mixed. Data from Zimbabwe tend to support this substitution hypothesis: Between 1984 and 1988 the use of modern methods increased by almost 10 points, while the use of traditional methods decreased by about 3 points. However, in Botswana, modern use increased 13 points in the four years between surveys, while traditional use remained at slightly less than 1 per- cent in both surveys. In Kenya, the 12-point increase in modern method use was accompanied by an 8-point increase in the use of traditional methods (from 1 to 9 percent over 11 years). Differentials in Use At the individual level, a number of socioeconomic factors are associ- ated with current modern contraceptive use. We use DHS data to investi- gate these relationships, as illustrated in Figures 2-1 to 2-3 (see Table 2A-1 for percentages presented in these figures). The main findings are as fol- lows (see Chapter 3 for additional discussion): . The use of modern contraceptives is higher in urban than rural areas. Without exception, prevalence rates for modern methods are higher in urban than rural areas (see Figure 2-1), though in Ghana and Ondo State, Nigeria, the difference is not statistically significant.8 The difference ranges from a few percentage points in those countries where modern use is low even in urban areas to as much as 18 points in Zimbabwe. 8The bivariate comparisons presented in Figures 2-1 to 2-3 and Tables 2-7 and 2-8 were checked against multivariate results derived from logit models (the latter using unweighted data). In the logit models, the explanatory variables included the woman's age, urban or rural residence, her level of education, and also the region of residence. All of the covariates were treated as categorical in nature and were entered into the logit models as main effects (i.e., no interactions between covariates were explored). (The omitted categories were (1) residence = urban, (2) education = none, and (3) age = 15-19.) The aim was to determine if the bivariate contrasts displayed in the figures and tables provide a reliable guide to the direction and significance of effects in the presence of multivariate controls. Generally they do, except where noted in the text. See Table 2A-1 to this chapter for the significance of the effects in the multivariate model. .
34 60 50 4a 30 20 10 O ~ FACTORS AFFECTING CONTRACEPTIVE USE Percentage L ~ L~ At' , BWA BDI GHA KEN LBR MLI NGA SEN SUD TGO UGA ZWE Country Urban ~ Red FIGURE 2-1 Modern contraceptive use by residence of women currently in union (age 15-49~. NOTE: BWA=Botswana; BDI=Burundi; GHA=Ghana; KEN=Kenya; LBR=Liberia; MLI=Mali; NGA=Nigeria; SEN=Senegal; SUD=Sudan; TGO=Togo; UGA=Uganda; ZWE=Zimbabwe. Except in Botswana, Kenya, and Zimbabwe, the use of modern methods remains an essentially urban and periurban phenomenon in the region. Elsewhere in the western and eastern African countries studied, less than 5 percent of married women in rural areas report modern contraceptive use. Analyses of the earlier WFS data revealed similar tendencies (Lightbourne, 1980; Sathar and Chidambaram, 1984~. With urbanization comes greater exposure to Western ideas and greater access to family planning and health services. Except in Zimbabwe where the community-based distribution of contraceptives is strongly developed in rural areas (London et al., 1985; Zimbabwe National Family Planning Council, 1985), family planning services are far more likely to reach urban than rural residents. · The use of modern contraceptives varies markedly by region within a given country. As mentioned above and shown in Table 2-3, there are marked differences between the national average and region-specific preva- lence rates. The greatest use of modern methods is observed in regions that are very urban. For example, in the Kampala region in Uganda, the region of Nairobi and the Central province in Kenya, and the region of Harare and Chitungwiza in Zimbabwe, the proportion of married women using a mod
LEVELS AND TRENDS IN CONTRA CEPTI VE USE 60 50 40 30 20 10 o Pementage 1 BWA BDI GHA KEN LBR ELI FJGA SEN SUD TGO UGA ZWE Country _ None _ Primary _ Set+ 35 FIGURE 2-2 Current use of modern contraception by level of education of women In union (age 15-49), selected countries of sub-Saharan Africa. NOTE: BWA=Botswana; BDI=Burundi; GHA=Ghana; KEN=Kenya; LBR=Liberia; MLI=Mali; NGA=Nigeria; SEN=Senegal; SUD=Sudan; TGO=Togo; UGA=Uganda; ZWE=Zimbabwe. em contraceptive method is at least 10 points higher than for the country as a whole. Conversely, in the less urbanized regions, modern contraceptive use is lower than the national averages. Examples include Matabeleland North and South and to a lesser degree Manicaland in Zimbabwe; the Nyanza and Western provinces in Kenya; and the regions of Darfur and Kordofan in Sudan. Other explanations for these regional differences in modern contra- ceptive use are explored in Chapter 6. For traditional methods, there is less regional variation than is the case with modern methods (data not shown), which may suggest that traditional methods are part of a set of culturally prescribed practices that are less affected by forces of.modernization or access to family planning services. . The use of modern contraception increases with education. Female education is an important determinant of demographic behavior in general and of family planning behavior in particular. The African data yield re- sults similar to those obtained from DHS studies worldwide: The use of modern contraceptives increases with the educational level of the woman (see Figure 2-2~. Surprisingly, this same positive relationship between education and use is seen for traditional methods in at least half of the DHS countries in
36 FACTORS AFFECTING CONTRACEPTIVE USE Parentage 40 30 20 10 ~- - - O . , I ~ 1~19 20-24 2~29 30 34 Age - 3~39 40 44 45 49 Kenya I Boy ~ ~rnbabwe FIGURE 2-3 Current use of modern contraception by age of women (15-49 years) for the vanguard countries. Africa (data not shown). However, it is not the case in Zimbabwe and Botswana, where the better-educated women are generally using a modern method, or in Togo where the use of traditional methods is high across all categories of education. · The use of modern contraception generally increases, then decreases, with aged The variation in contraceptive prevalence by age is similar in the higher prevalence countnes: The use of a modern method among married women is lowest among women aged 15-19, gradually increases, and then decreases again toward the end of the reproductive years. The peak in the curve differs somewhat by country. In Zimbabwe, use is highest among women aged 20-34 compared to Botswana and Kenya where it tends to be highest among women aged 25-40 (see Figure 2-3~. In summary, modern contraceptive prevalence increases with the educa- tion of the woman and is higher in urban areas and among women aged 20- 39. 9For several countries, particularly Burundi, Mali, and Togo, the changes in use do not vary significantly by S-year age group. See Appendix Table 2-1 for the significance of the age effects.
LEVELS AND TRENDS IN CONTRACEPTIVE USE Reason for Use: Spacing Versus Limiting 37 It is important to understand the motivations for contraceptive use, that is, to what extent is use motivated by a desire to limit family size versus the desire to space births? In the WFS, CPS, and DHS conducted to date, there is no direct question regarding the reason for contraceptive use. Rather, respondents are asked if they desire more children and, in the case of the DHS, when they would like their next birth. This information, when com- bined with contraceptive use data, allows researchers to infer the motive for use. Those users who desire no more children are classified as limiters; those users who want more children eventually but not for at least two years, are considered spacers. (For more information on this indirect means of defining motivation, see Mauldin and Segal, 1986; Westoff, 1991; Westoff and Ochoa, 19911. At first glance the data in Figure 2-4 suggest that in at least half of the DHS countries, modern contraceptives are used for limiting rather than spacing. However, it is important to note that in the five of the six coun 100 80 60 40 20 a Percentage 'Al K 111 ~ 88 111 8R,I ELI ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ _ ~ ~ ~ ~ ~ ~ ~ ~ ~ l ~ ~ 1 ~ ~ ~ ~ 1 Ll L'l Ll L I L Ll ~ ~ ~ ~ BWA. BDI GHA KEN LBA MLI SEN SUD TGO UGA ZWE Country - Limping ~3 Spacing FIGURE 2-4 Use of modern contraception for spacing versus limiting, women in union (age 15-49), selected countries of sub-Saharan Africa. NOTE: BWA=Botswana; BDI=Burundi; GHA=Ghana; KEN=Kenya; LBR=Liberia; MLI=Mali; SEN=Senegal; SUD=Sudan; TGO=Togo; UGA=Uganda; ZWE=Zimbabwe.
38 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-6 Proportion of Women Using a Modern Method by Number of Living Children for the Higher Prevalence Countries Number of Children Country 0 1 2 3 4 5+ Botswana 13.6 30.4 37.2 41.2 33.4 31.2 Kenya 0.9 10.1 18.6 19.6 22.5 21.0 Zimbabwe 1.4 36.5 42.4 42.2 43.6 32.3 tries where limiting emerges as the primary motive, modern contraceptive use is low (less than 6 percent). In two of the three higher use countries, Zimbabwe and Botswana, the primary motive is spacing; these findings support the often-heard statement that family planning is used for spacing rather than limiting in sub-Saharan Africa. For example, Westoff (1991) maintains that data from the DHS in Africa on reproductive intentions and desired timing of births reflect the persistence of a double desire: sustained fertility, on the one hand, and appropriate spacing between successive births, on the other. By contrast, Kenya presents an intriguing case in which more than half the users openly state that they want no more children. Table 2-6 sheds light on the relationship between modern method use and current family size for Botswana, Kenya, and Zimbabwe. Only in Botswana, do a substantial proportion (14 percent) of women with no living children use modern contraception. Among those with one living child, the percentage using contraception increases sharply. Between one and two children, there is a slight additional increase in use, but above parity two, use seems to plateau (and even drops for parities of five or greater in the case of Zimbabwe). In sum, these cross-sectional results imply that most women in these countries probably do not begin using modern contracep- tion until they have at least one to two children. OTHER INDICATORS RELATED TO CONTRACEPTIVE PRACTICE Current use of a modern contraceptive method is the aspect of contra- ceptive practice that is of greatest interest both to demographers (as a proxi- mate determinant of fertility) and to family planning policymakers (as a measure of the coverage of their programs). However, several other aspects of contraceptive practice are of interest: ever use of a method, discontinuation,
LEVELS AND TRENDS IN CONTRACEPTIVE USE 39 knowledge of contraceptive methods, knowledge of sources of contracep- tion, and fertility preferences. Family planning researchers often use these variables as indicators of constraints on use in a population. Ever use of contraception can signal the degree to which contraception is accepted or tried. Low levels of knowledge of contraception may indicate that the population is unaware of fertility-limiting options. Low levels of knowl- edge of sources of contraception may suggest that access to family planning is limited and program extension is warranted. Fertility preferences indi- cate the potential demand for children and suggest regions where there is unmet need for family planning. Ever Use of a Modern Method Ever use of a modern method refers to the percentage of women ever in union who at some point during their lifetimes have used a modern contra- ceptive. Ever use by definition must equal or exceed current prevalence. The ever-use measure reflects the extent to which a given population has used contraception (and thus serves as a useful indicator in studies designed to test specific family planning interventions). Moreover, the comparison between ever use and current use gives an indication, admittedly crude, of the perseverance of users and the relative duration of contraceptive use (London et al., 1985; United Nations, 1989a). In 10 of the 12 African DHS countries, less than 30 percent of women ever in union have used a modern contraceptive method. In sharp contrast, 60 and 63 percent in Botswana and Zimbabwe, respectively, have ever used modern contraception. The rank ordering of countries is similar for ever use and current use. As shown in Table 2-2, the five African countries leading the list on current use of a modern method are (in descending order) Zimbabwe, Botswana, Kenya, Sudan, and Liberia (although in the last two, 6 percent of currently married women use one). These same countries are first in ever use with the same ranking ordering except for Liberia, which is replaced by Ghana, where "experimentation" with contraception would be in line with available information on program activity in that country (de- scribed in more detail in Chapter 5~. In Zimbabwe, Botswana, and Kenya, at least half the women who have ever used any method are currently using one. Despite a very low preva- lence rate, the data in Table 2-2 suggest that in Burundi, more than half the "ever users" of modern methods are currently using one. The difference between ever and current use is smaller in these countries than in other DHS countries, indicating that a higher percentage who try a method stick with one.
40 FACTORS AFFECTING CONTRACEPTIVE USE Differentials in ever use of a modern method among different types of women ever in a union are shown in Table 2-7.~° As with current use of a modern method, ever use is higher in urban than rural areas, and it increases monotonically with education for all countries. Ever use generally reaches its highest level among women aged 25-34. Discontinuation It is widely recognized that the rate of contraceptive use at a given time depends not only on having attracted acceptors to adopt contraception in the past but also on maintaining their motivation to continue use for a pro- longed period, in accordance with their own reproductive intentions. How- ever, data on this subject are severely lacking in sub-Saharan Africa an important research gap in understanding prevalence in this region. One promising approach to data collection is the "calendar," which has been developed and tested as part of the DHS. However, to date none of the Afnc an surveys has included this instrument, in part because it is thought to be more appropriate to countries with high prevalence. Information on continuation rates can also be obtained from program- based data; however, these data have several limitations. First, samples represent the users of specific clinic-based services, rather than a cross section of the population who are using the method (including patients of private physicians or clients who obtain methods from pharmacies). Sec- ond, there is a serious loss-to-follow-up problem; typically the dropout lev- els are moderate to high, and one can speculate that those who discontinue their clinic visits are not necessarily similar in education, ethnic group, parity, etc., to those who follow their appointments regularly. Thus, the results of the analysis of such program data are only applicable to those who continue their visits, not the full sample of people who have ever used the clinic services. Third, some countries have their record-keeping sys- tems set up to follow individual clients and thus measure continuation; however, these records are not kept with sufficient accuracy to allow mean- ingful analysis. Knowledge of a Modern Contraceptive Method The DHS surveys report "knowledge of a contraceptive method." This phrase is somewhat of a misnomer, in that the surveys measure re iOThe denominators of the proportions in Tables 2-7 and 2-S are women aged 15-49 who have ever been in a union. This denominator is used because ever use and knowledge of contraception attempt to give measures for all women who have been exposed to the risk of becoming pregnant, both now and in the past.
LEVELS AND TRENDSIN CONTRACEPTIVE USE 4 1 ported awareness of different contraceptive methods. Specifically, the in- terviewer asks the respondent to name all methods of which she or he has heard ("spontaneous mentions. For each method that is not mentioned, the interviewer names the method, gives a one line description of it, and then asks if the respondent has ever heard of it ("prompted recalled. In the family planning literature, these two questions are combined into a single variable that is conventionally labelled as "knowledge," either of a specific method or of any method. We report knowledge of methods in Table 2-8. More than two-thirds of the women ever in union have heard of at least one modern method of contraception in almost all of the 12 African DHS countries; the exceptions are Burundi (63 percent) and, more notably, Mali (29 percent). Knowledge is higher in urban than in rural areas, although the gap is small or insignificant in the higher use countries where contraception is known to more than 90 percent of women, whether urban or rural. The data also show a positive and monotonic relationship between knowledge and level of education. However, Zimbabwe is remarkable, because 94 percent of women with no education have heard of at least one modern method, which suggests that family planning programs have reached all segments of the population. Knowledge of contraception is related to age in a pattern best described as an inverted U. although the differences by age are not significant in Burundi, Mali, and Togo. Generally, women aged 15-19 are less likely to have heard of a modern contraceptive method than are women aged 20-39. However, among women 40 and over, the percentage drops markedly, re- flecting perhaps the lower educational attainment and lower need for con- traception over the life cycle among this age group. Zimbabwe is again the only exception, where at least 95 percent of women in each age group know of a modern method. 1lSome might question the validity of this self-reported "knowledge," especially when the respondent can qualify for "knowing a method" simply by a nod of the head when the inter- viewer reads those methods not previously mentioned. However, the results obtained from this series of questions from numerous DHS or DHS-type surveys tend to be highly consistent with other sources of data. For example, the levels of professed knowledge are high for those methods most widely used in a given country, as shown by program service statistics. By contrast, few respondents claim to have heard of vasectomy, a method that is not widely promoted in public programs in developing countries. Also, data on knowledge of methods tend to show a consistent pattern of increase in subsequent surveys in a given country and to correlate predictably with education, urban residence, and economic status. Thus, the problem lies not with the utility of the indicator, but rather with the label "knowledge" when in fact "awareness" is a more accurate term. In this report we have retained the term "knowledge" to be consistent with the family planning literature.
42 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2-7 Women Who Have Ever Used a Modern Method (percent) Residence Education Countrya Total Urban Rural None Primary Secondary Botswana 60.2 71.2 55.1 40.8 65.5 81.7 Burundi 2.5 26.7 1.5 1.4 4.8 28.8 Ghana 23.1 33.2 18.4 11.0 30.9 52.1 Kenya 29.1 44.4 26.1 16.1 31.1 48.1 Liberia 17.8 28.9 10.8 8.4 23.0 62.2 MaliC 3 4 11.8 0.4 1.3 12.6 56.3 Ondo 9.4 13.8 6.5 3.3 10.7 23.7 Senegal 6.5 16.7 1.1 2.3 19.5 50.6 Sudan 18.1 34.9 8.2 7.0 29.1 41.8 Togo 10.6 21.0 6.2 5.2 15.6 38.8 Uganda 7.6 34.1 4.6 2.2 7.9 37.4 Zimbabwe 62.7 76.0 56.6 47.2 62.6 76.2 NOTE: The base for the table is all women age 15-49 ever in a union. aThe Nigeria DHS of 1990 is not included because the standard recode data tape was not available during the writing of this report. TABLE 2-8 Women Who Know Any Modern Method (percent) Residence Education Countrya Total Urban Rural None Primary Secondary Botswana 94.4 99.4 92.1 85.2 98.7 99.5 Burundi 63.3 92.3 62.1 60.1 75.9 94.2 Ghana 76.8 88.4 71.5 60.5 89.8 97.2 Kenya 91.3 95.4 90.6b 82.8 94.4 98.5 Liberia 69.5 78.5 63.9 61.5 85.6 96.0 MaliC 28.8 56.5 19.2 23.1 60.8 100.0 Ondo 49.4 60.0 42.3 33.2 58.2 79.2 Senegal 68.6 87.9 58.7 64.8 87.7 97.8 Sudan 71.0 91.1 59.1 54.8 91.1 98.3 Togo 81.4 92.4 76.7 75.5 91.7 98.2 Uganda 78.5 93.6 76.9b 68.0 85.2 96.3 Zimbabwe 97.9 99.0 97.4b 94 0 98.4 99.5 NOTE: The base for the table is all women age 15-49 ever in a union. aThe Nigeria DHS of 1990 is not included because the standard recode data tape was not available during the writing of this report. bThe contrast is not significant at the .05 level in a multivariate logit model.
LEVELS AND TRENDS IN CONTM CEPTI VE USE 43 Age 15-1920-2425-2930-3435-3940-4445-49 36.658.367.868.365.454.131.2b 3.81.gb2.9b3.0b1.7b2.2b2.6b 11.619.425.627.524.826.416.9 10.222.630.235.834.828.926.7 8.118.520.222.419.215.512.5 3.83.7b4.9b4.4b2.4b1.8b0.0 7.99.5b12.59.511.77.75.9 1.94.0b6.6b9.2b10.46.36.3 5.314.118.423.320.219.418.3 9.110.4b1 l.0bl3.2b9.8b10.3b7.4b 4.45.9b8.88.510.110.44.7 41.167.770.571.863.952.440.9b bThe contrast is not significant at the .05 level in a multivariate logit model. CFor Mali, logistic regression omits cases in which the woman was age 45-49. Age 15-19 20-24 25-29 30-34 35-39 40-44 45-49 92.4 96.7b96.9 95.9 95.7b 9o.9b 84.2b 61.3 61.4b71.0b 62.9b b b 50.6b 70.5 78.7.81.4 77.8 77.5 73.8 66.9b 86.9 94.594.2 93.1 92.8 85.2 82.8b 54.2 73.772.8 74.4 69.5 67.2 63.2 32.7 33.0b34.1b 30.7b 24.3b 22.0b 14.1 40.9 52.257.4 55.5 52.5 46.2 34.2 51.2 72.671.1 74.3 71.0 65.3 66.4 67.9 75.0b72.7 73.1 69.6 67.6 65.0 76.1 82.4b84.1b 83.sb 80.6b 8o.sb 74.9b 75.3 79.5b83.4 77.5 81.2 75.4b 69.5b 96.0 98.798.4 99.1 98.0 94.9b 96.5b CFor Mali, the logistic regression omits cases in which the woman had secondary schooling or more.
44 FACTORS AFFECTING CONTRACEPTIVE USE Sources of Modern Contraceptive Methods Information on knowledge of sources of contraception is one indicator of the extent to which contraceptives are accessible in a given country (although the data pertain to perceived availability, not actual availability). As shown in Table 2-9, among women aged 15-49 who know at least one method, a high proportion (78 to 100 percent) also know a source where they can obtain contraception. However, the type of source varies markedly by country, reflecting differences in the configuration of family planning service delivery in these countries. The most frequently mentioned source is a government clinic or pharmacy, except in Senegal where three-quarters of the respondents cited a private clinic or other service delivery point. However, knowledge of a second source is much more limited; only in Ghana can at least 50 percent of the respondents name a second source (data not shown). The strong influence of government programs is reflected in the data in Table 2-9. Far more women know of government than private sources in 11 of the 12 countries. One can assume that the relatively low levels of knowl- edge of private sources in Botswana, Kenya, and Zimbabwe simply reflects the extent to which family planning services are delivered through govern- ment facilities. In contrast, at least one-third of respondents knowledgeable about at least one source mention a private source (clinic or pharmacy) in the low-prevalence countries of Senegal, Ghana, Togo, and the Sudan. Given that the pill is the most widely known and used modern method in Africa, it is also of interest to examine knowledge of at least one source of the pill. Not surprisingly, the most widely known sources of the pill are similar to the most widely known sources for any method, although the specific percentages knowing each type of facility tend to be lower (data not shown). With the growing concern about AIDS, knowledge about condoms takes on a new importance. Table 2-10 provides information on women ages 15- 49 currently in a union who have knowledge of condoms. There is a re- markably wide range in awareness of condoms (or willingness to admit this knowledge to the interviewer) among countries: In Burundi, Mali, and Sudan, less than 30 percent of women who know any modern method ac- knowledged having heard of the condom, in comparison to 92 percent in Botswana. Among women with knowledge of condoms, most (67 percent or more) also know at least one source from which they believe condoms can be obtained. Government clinics or pharmacies are the most widely cited sources in eight of the eleven countries. By contrast, private pharmacies are most frequently mentioned in the remaining three countries (Ghana, Sudan, and Togo).
LEVELS AND TRENDS IN CONTRA CEPTI VE USE Fertility Preferences 45 Data on reproductive preferences are useful indicators of fertility norms and intentions. Both the WFS and DHS collected data on ideal family size and intentions to have another child. Data from the DHS on ideal family size among ever-married women show that the preferred number in sub-Saharan Africa averages just under six. The lowest ideal number is 4.7 (Kenya) and the highest is 7.1 (Senegal). The ideal number has declined between the WFS and DHS; for the three countries that conducted both surveys, Kenya showed the greatest decline (35 percent) and Ghana showed the smallest decline (9 percent) (Westoff, 1991~. Care should be taken in using these numbers as predictors of change in fertility; they are influenced by the number of children the respondent already has. The DHS also collected information on whether married women in- tended to postpone or halt childbearing. In sub-Saharan Africa, more women indicated that they wanted to delay a future birth than avoid one (Westoff, 1991~. Between the WFS and DHS, the number of women who wanted no more children increased in the three countries that conducted both surveys. Kenya showed the greatest increase (32 percent), followed by Ghana and Senegal (10 and 9 percent, respectively) (Westoff, 19914. Such data are indicative of changes in fertility in the short run (Westoff, 1990~. Westoff and Ochoa (1991) have used data on reproductive preferences to develop measures of unmet need for family planning. They define unmet need as the proportion who are exposed to the risk of conception, are not using a contraceptive method, and say they want to delay or stop childbearing. The total demand for family planning is the proportion of women with unmet need plus the women who are currently using contraception. Results from Westoff and Ochoa show that most of unmet need is for spacing rather than limiting births in sub-Saharan Africa, in sharp contrast to other regions of the world. Table 2-11 summarizes unmet need, as calculated by Westoff and Ochoa, by subgroup for eight populations in Africa. Unmet need gen- erally increases with number of children; is greater in urban areas than rural areas, except in Botswana and Kenya; and is greatest for women with pri- mary education, except in Botswana, where it is greatest among the unedu- cated, and in Burundi, Liberia, and Uganda, where it is greatest among those with secondary education. Overall, unmet need is highest in Togo (40.1 percent) and lowest in Mali (22.9 percent), suggesting that there is a substantial demand for family planning in all eight countries. Compansons of unmet need for limiting births between the DHS and the WFSi2 in Ghana 12Comparisons are noted only for unmet need for limiting births because the WFS did not collect information on spacing preferences.
46 "s° o o Cat en .~ no 3 be so o o o Cut C) o V, o - o Cat Cut C) o ~ ~ ~ o o ~ ~ o . . v Cut of 4 - C> C) C) o ~: o CD C~ o ~S E~ 04 o ~ oo oo U~ e~ oo o . . . . . . . ~ o C~ o ~ o C ~_4 o ~ ^ U' ~ ~ ~o o ~ ~ o o o ~ ~ ~ ~ V ·- ° ° ~ ~ ° ° ° ~ ~ ~ ° ° e~ ~ ;~ Ce P~ P~ . ;> ~ - .- .- ~L) P" ~ ~ C: ;^ o ~ o o o V o .o C~ ·3 o ~ o C~ C =C o o ~ ~ ~ ~ 0 cr 0 ~ ~ 0 ~ . . . . . . . . . . . - ~ C~ ~ oo ~ o ~ U~ ~ ~ ~ ~ _ ~ ~ _ =\ t_ ~ _ ~ ~ oo ~ ~ ~ - . . . . . . . . . . . C~ _ ~ _ o ~ ~ ~ ~ ~ - ~ C~ ~ ~ ~ o o ~ o o o o o o ~ ~o . . . . . . . . . . . o o ~ o o _ ~ o o o ~ _ ~ ~ ~ ~ a~ oo ~ oo . . . . . . . . . . . C~ ~ oo ~ ~ ~ oo ~ ~ ~o ~ ~ ~ oo ~ oo ~ ~ ~ oo oo a~ oo ~ O oo ~ ~ ~ ~ ~ . . . . . . . . . . . _ o oo oo ~ ~ ~ ~ ~ oo o~ ~ ~ C C .` 0.0 C o D o C ~0 3 o 3 ~ o C~ - .~: o o o C C) CO ,0. ~L> 0 ° U~ - oc o 3 - Ct ._ e~ o E~ . . E~ o z Ct Ct o C) ·(O C) 3 Oe~ C~ s: ~ t o o c) ~ o ~: o s: 3 ~: 3 - o U' - ._ C) C~ C~ o os ._ ._ Ct C~ s~ : - o. . ~ o s: ~: ~o o o U) C~ C~ ._ o o ~q e~ ._ ~ ·Z ° ~o . ~t_ Ct ~ ._ _! ~C~ ._ o ~: CD _ . ~D ._ Ct ;> ~t C, C) .S ~_ C40 . _ (L) o oc ao o:-~ ~ ~ 3 ~ =° ~S: 04 ° C~ ~ - o ~ c: C;: ~ ° 5 C.) ~ ~ - - 0D Ct s~ ~ ;> O Ct' ~ O o ~ ~ Ct ~5 C) Z 3 c: ts p,.= Ct
47 C~ C) o V) o C) E~ Ct C) s~ o U) C~ ~4 ._ 3 o ~: o o o bC ._ 3 o C) b4 e~ P~ . . C~ o o V o Cq o U: o C~ C) Z o o ~o c: 'o o o oo ~ o ~ ~ CM ~ ~ ~ ~ . . . . . . . . . . . o ~ ~ o ~ o ~ o o , ~ o o ~ ~ ~ o o ~ o oo _ o o ~ _ ~ o o o ~ ~ o C~ V ~ CC ~ c) Ct _ s~ ~ C · ;- =: oo In ~ ~ ~ 00 0 ~ O ~ ~ . . . . . . . . . . . oo oo O oo _ ~ Ir; c~) t~ Ir~ ~ C) ._ ._ - V ~:- 5: ~ CO ~ o ~ _ .= o ~ ~ ~ _ _ oo ~ ~ ~o ', s~ ~ o ._ _. _ ~ ~ ~ ~ ~ ~ ~ o o V, ~ ~ ~ ~ _ c: ~ E C ~ o 0 0 ~ ~ := V ~ C) ;~., o .O ~ ;> ~ Ct O '-v ~ o 0 ~ o 0 ~ ~ o 0 ~ ~ . . . . . . . . . . . o o o o o o _ o o o oo ~ ~ _ ~ o ~ oo oo ~ o ~ ~ oo o~ oo _ _ ~ o C~C~ o ~ ~ vo ~ ~ ~ ~ ~ C~ _. . . . . . . . . . . - 'oo o oo ~ U~ ~ ~ ~ ~ ~ o o ,,;~ ~ o 3 o o V ~ C4 ~ o C~ ~ t_ _ t_ ~ ~ ~ o ~ ~ . . . . . . . . . c~ ~ ~ - oo ~ cr ~ ~ c~ ~ ~ C~ ~ ~ ~ ~ ~ C~ ~ ~ oo c o ~o c, ct 3 v ~mmo~o~ C) CJ C~ Ct Ct C5 Ct Ct e~ ~: 0 0 O ._ :- O P~ _ Ct ~ C ~Z O V ~: U, O Z . _ 53 ~0 _ C) .= ~ ~O O ~ `: CC ._ o O C) _ ._ O ~ O ~O ._ ~ Ct ~O ~C40 ._ ._ Ce: ;^ Z Ce c: 5: . ~. 's" O .D -0 3 ~-9 ~ In ~ _ C<S. o · ~ o Ct C . ~ o C) c ,,, ~-= E c . ~,= ." ~ 3 ~ ~o D o ~ OC O o O ~ '- "e° ~S° Z ~ ~ 3 3 ~ c) D ~ ~ ~,- C) ~ . _ ~ 4_ .. O ~ ~ ~ .. oO ~o 60 s0 ~V O ~ ~ ~ ~O Z 3 u,
48 FACTORS AFFECTING CONTRACEPTIVE USE and Kenya (data not shown) indicate that unmet need has increased, 80 and 129 percent respectively. RESULTS FROM MALE SURVEYS The evidence on levels and trends in contraceptive use presented thus far in this chapter comes from data gathered for women. Several DHS also questioned men. In Burundi, Ghana, and Kenya, a subsample of the hus- bands of the women interviewed were selected to respond to the question- naire for men. In Mali, a sample of men was selected from the households visited by the interviewers; thus, the sample included single men. Results from these surveys indicate that men were more likely to report using a contraceptive method than women, except in Ghana, where the reported use of any method was approximately the same. Contraceptive use among men was higher in urban areas; generally decreased with age, except in Kenya, where use increased and then decreased with age; and increased with education.~3 In all four countries, men were more likely to know of a method than women. In Burundi, Ghana, and Mali, of the men that knew any method, more than half of them had not talked with their wives about family plan- ning in the past year. In Kenya, 36 percent of the husbands had not talked with their wives. More than three-quarters of men in Burundi, Ghana, and Kenya approved of family planning; only 16 percent of men in Mali ap- proved. Except in Burundi~4, more than a third of the wives of husbands who approved of family planning either did not know their husbands' atti- tudes toward family planning or thought their husbands disapproved. In Burundi and Kenya, men and women indicated the same ideal family size, 5.5 and 4.8 respectively. In Ghana, men preferred more children than their wives, 7.6 versus 5.5. The DHS report for Mali did not report ideal family size for men. CONCLUSION Modern contraceptive use in sub-Saharan Africa remains low except in Botswana, Kenya, and Zimbabwe. In all countries, use of modern methods is associated with urban residence and greater education, perhaps indicating that future changes in the composition of the population might be associated with increased use. The knowledge of modern contraceptives is greater i3In Kenya, there was little difference in use between men with no education and those with primary education. i4Data for Burundi are not presented in the DHS report.
49 ~'e to two is: . 'it "o .s o at ~ ~o o ad ~ A 3 .- a: so o + it, ._ so Cal o ~ z C) o ~ o ~ C, -^ EM ~u ~ ~ ~ o Do ~ ~ ~ . . . . . . . . in ~ Do car car O ~ ~ ~ ~ ~ ~ ~ Cal Do ~ Go ~ ~ ~ o US . . . . . . . . Oo ~ ~ car ~ Do cry Cal ~ ~ ~ ~ ~ o Do Cal ~ ~ Cal . . . . . . . . ~ ~ ~ o o ~ o Go Cal ~ ~ ~ ~ ~ US Go Go . . . . . . . . ~ - , ~ ~ ~ ~ ~ ~ o o ~ ~ ~ ~ ~ o U~ ~ 00 0 0 00 C~ C~ . . . . . . . . C~ ~ ~ C~ ~ 0 ~ 0 ~ ~ ~ ~ 0 ~ ~ ~ . . . . . . . . ~ 0 oo ~ ~ C~ ~ ~ . . . ~ ~ U~ ~ ax ~ ~ . . . . 0\ 0 ~ ~q C~ cr ~ 0 ~ cr . . . . . . . . ~ ~ 0 C~ C~ . . . . . . . . \0 ~ ~ ~ 1- C'N CM C~ 0 ~ oo r~ 0 ox ~ ~ _ CM ~ ~ ~ ~ ,: Ct 0 ~ ..~: ~ 53 C<S ~ 't '_ a, C~ ~S C~ O ~o U: .. O
50 FACTORS AFFECTING CONTRACEPTIVE USE TABLE 2A-1 Differentials in the Percentage of Women Who Currently Use a Modern Method Residence Education Country Total Urban Rural None Primary Secondary Botswana 31.7 40.8 27.5 18.9 32.9 50.3 Burundi 1.2 14.0 0.8 0.6 2.2 16.9 Ghana 5.2 8.1 3.9a 3.2 6.5 10.1 Kenya 17.8 25.5 16.4 9.7 19.2 29.3 Liberia 5.5 9.7 3.1 2.5 6.6 22.1 Malib 1.3 4.9 0.1 0.4 5.5 20.3 Ondo 3.8 5 3 2 Sa 1.9 3.6 9.0 Senegal 2.4 6.7 0.3 1.0 5.9 22.1 Sudan 5.5 11.3 2.2 1.9 7.9 14.9 Togo 3.1 6.5 1.7 1.7 3.9 12.2 Uganda 2.5 12.2 1.5 0.9 2.7 11.3 Zimbabwe 36.1 48.8 30.9 25.0 34.0 52.3 NOTE: The base for the table is all women currently in union. aThe contrast is not significant at the .05 level in a multivariate logit model. bFor Mali, the logistic regression omits cases in which the woman was age 45-49. than their use, and the knowledge of government sources of supply in the three higher use countries indicates the importance of government family planning programs in those countries. Estimates of unmet need indicated that there is substantial demand for additional family planning services, especially for spacing births.
LEVELS AND TRENDS IN CONTRA CEPTI VE USE Age 51 i5-19 20-24 25-29 30-34 35-39 40-44 45-49 14.5 25.2a 36.0 33.7 37.5 35.0 14.7a 0.6 o.7a l.4a 1.sa l.3a 1 oa 1 6a 2.3 3.4a 4.3a 6.9 5 pa 9 0 4.2a 6.7 11.8 16.8 22.2 22.9 21.2 17.5 2.0 4.9a 6.7 6.3 5.1 5.7 7.1 1.4 l.4a 1.6a 1.6a l.4a 0 ga 0.0 1.7 2. la 3.oa 4.6 5.4 4.2 3.5 0.5 0.8a 25a 4.8a 3.8 1 la 2~4a 2.2 42a 5 oa 7 9 5.7 7.7 4~4 2.1 1 ga 2 sa 3.7a 4.7a 42a 3.4a 1.2 1 1a 1 ga 2.6 6.0 5.2 2.5 28.3 41.5 43.8 42.9 30.8a 261a lima
