Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 19
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
OCR for page 20
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.
OCR for page 21
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.
OCR for page 22
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.
OCR for page 23
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.
OCR for page 24
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.
OCR for page 25
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
OCR for page 26
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.
OCR for page 27
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
OCR for page 28
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.
OCR for page 29
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.
OCR for page 41
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.
OCR for page 42
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.
OCR for page 43
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.
OCR for page 44
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).
OCR for page 45
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.
OCR for page 46
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
OCR for page 47
OCR for page 49
OCR for page 51
Representative terms from entire chapter:
modern contraceptive
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
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
OCR for page 50
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
