Ngalinda, Innocent: Age at First Birth, Fertility, and Contraception in Tanzania

226

Chapter 7. Early Motherhood, Contraceptive Awareness and Use among Adolescents

7.1. Introduction

Adolescents broadly defined, as the young population aged 10-19 years, is a vital population segment, making up 24.2 percent of the Tanzanian population (1988 census). Adolescence is an important period in any human being’s life as this is the time of rapid growth both physically and mentally. Adolescents theoretically believed to be in the child bearing group (15-19 years), amount to 11 percent of all Tanzanian women.<35> This period of adolescence is characterised by high risk for early and unwanted sexual activity, forced marriage, and early pregnancy-related events. The challenge to the policy-makers and to social forces influencing society is to provide adolescents with more control over their lives by either making contraceptives accessible to adolescents, or delay sexual activity, marriage, and child bearing until they are ready and willing.

Adolescent fertility in Tanzania has been ignored for a long time because of the Malthusian assumption that fertility is confined to marriage. Maintaining this assumption means excluding the increasing magnitude of premarital and extramarital fertility as seen in Chapter 4. It should be noted that in Tanzania only 30 percent of ever married women had their first sexual experience within their first marriage. The remaining 70 percent had sexual intercourse prior to their first marriage. The actual level of premarital sexual activity may even be higher than the reported level because single women might feel that it is not acceptable to acknowledge that they are sexually active.


227

However, not all adolescent sexual activity is consensual. Sexual abuse, incest, and rape are troubling realities in developing and developed countries alike. In Uganda, nearly half (49 percent) of all sexually active girls in primary school reported that they were forced to have sexual intercourse. Almost 22 percent anticipate receiving gifts or money in exchange for sex (Waite, and Moore, 1978). In the United States, 7 in 10 women who had sex before age 14, and 6 in 10 who had sex before they reach age 15, report having had sex involuntarily (AGI, 1994).

Child bearing as a crucial period of human development might have serious consequences on these young mothers. It can limit educational attainment, restrict the skills young women acquire for the work force, limit their capacity to support themselves financially, negatively affect their healthy, and reduce their quality of life. As young people postpone marriage to later ages, the likelihood of beginning a sexual relationship prior to marriage increases. Consequently, so does the chance of unintended pregnancy, abortion or birth among unmarried adolescents. Moreover, their families may disown adolescents who gave birth before marriage since virginity is still considered important to a first marriage. Therefore, many are left with the responsibility of raising the children themselves which lead to prostitution among these abandoned adolescents. Besides these problems, schoolgirls are not allowed to continue attending school when they are pregnant. Therefore, many of them perform illegal abortion which can cost their life (Njau and Lema, 1988; Mashalaba, 1989; Meekers, 1990).

The cause of adolescent premarital births have been associated with the declining age at puberty (Chapter 4) that puts adolescent girls at the risk of early premarital exposure to the sexual activities. However, adolescent women who become a parent in adolescents are at greater risk of social and economic disadvantage throughout their lives than those who delay child bearing until their twenties. They are less likely to complete their education, to be employed, to earn high wages, and to be happily married. They are more likely to have larger families especially in a country of low contraceptive prevalence like Tanzania. The ability of young mothers to support themselves and their children is affected by the employment opportunities available to them. These opportunities, in turn, are largely determined by the qualifications that young mothers bring to the market. It seems reasonable to assume that caring for young children will conflict with and possibly reduce a woman's investment of time and effort in primary


228

school completion, college attendance, post-secondary training, and early work experience. If reductions in these early investments occur, they are likely to have profound, long-term consequences for the earnings and employability of the mother and, hence, for the economic wellbeing of both the mother and her children. It was found in the 1996 TDHS (Chapter 4) that a high proportion of young women (15-19 years) had unplanned births - either not wanted at the time of the birth or not wanted at all (27 percent). Faced with an unplanned pregnancy, some young women may turn to illegal abortion since abortion is prohibited in Tanzania as explained in Chapter 4. For these reasons, it is important to separate adolescents from other groups in the study of child bearing.

Economic hardships, urbanisation, and weakening of traditional structures that informed and regulated young people’s sexual behaviour have been associated with adolescent sexual and reproductive health problems. The situation has worsened due to the total lack of a reproductive health education component in the Tanzanian school curriculum. It was thought that such information might encourage promiscuity among adolescents rather than helping them to minimize the consequences of unprotected sex (Kapinga et al., 1992; Leshabari et al., 1997; Lwihula, 1996).

In Tanzania, adolescent women independent of their marital status are generally aware of contraceptive methods but very few use them. For example in the 1996 TDHS (Chapter 6) it was found that 65.4 percent of women at the age of 15-19 have heard of at least one method of family planning, but only 3.1 percent of those who know at least one contraceptive method, use any method. Only 13 percent of women age 20-24 were using efficient contraceptive methods. This indicates that adolescents are aware of modern contraceptive methods but in fact, very few use them. Therefore, in this chapter more emphasis will be on premarital first births and ways to reduce unwanted premarital pregnancies among adolescents.

7.2. Adolescents Child bearing

Adolescents engage in sexual practices early as most of them start sexual intercourse when they are 15, get married at 16 and have their first birth when they are 17 years of age ( Figure 24 ). But if we examine adolescents by using the life table approach, which


229

is relevant for the evaluation of duration variables like age at first intercourse, marriage and birth, age at first sexual intercourse, marriage and birth go up by .2 years. The main reason for using this approach is that these variables are measured in terms of the time elapsed before a particular event occurred. For example, age at first sexual intercourse is recorded as the time elapsed between the date of birth of an individual female adolescent and the date of the first sexual intercourse. However, for some individuals the time elapsed until the day of the interview may not have been sufficient for the event to have occurred. In these cases the information on duration has been censored by the interview. For censored information, life table methods and the use of current data are found to be the most suitable analytical methods and have been employed for duration analysis in this chapter.

Figure 24: Adolescents’ age at first sexual intercourse, marriage and birth in single years

Source: calculated from 1996 TDHS


230

Table 67: Adolescents’ age at first sexual intercourse, marriage and birth

T10

T25

T50

Trimean

T75

N

AFSI

12.8

14.2

15.4

15.3

16.5

707

AFM

14.0

15.0

16.2

16.2

17.3

442

AFB

14.8

15.8

16.9

16.8

17.9

361

Sources:

AFSI (Age at first sexual intercourse)- Table 17

AFM (Age at first marriage)- Table 21

AFB (Age at first birth)- Table 11

According to Tanzania data, the average age at first sexual intercourse is 15.3 years. The average age at first marriage is 16.2 years. The average age at which adolescents bear their first child in Tanzania is about 16.8 years. By the age of 15, about 10 percent of adolescents had their first live birth and about one-quarter have done so by the time they are 16 years old. In general, it is evident that adolescents are having sexual intercourse before marriage. But most of them have their first birth after marriage and the difference between these three events is one year each between the events. However, the timing of first sexual intercourse, marriage and birth is concentrated between age 15-17.

Table 68: Mean age at first birth by current age in sub-Saharan Africa and selected developing countries

Country

Year

AFSI

AFM

AFB

Benin

1996

15.3

16.2

16.9

Central African Republic

1994

14.9

15.1

16.2

Comores

1996

14.2

14.9

15.8

Côte d’Ivoire

1994

14.7

15.4

15.9

Ghana

1993

15.5

16.2

17.1

Kenya

1993

15.1

16.4

16.7

Mali

1995

15.0

15.0

16.2

Tanzania

1996

15.2

16.2

17.2

Uganda

1995

14.8

15.5

16.3

Zambia

1996

14.8

16.1

16.6

Zimbabwe

1994

16.2

16.4

16.8

Egypt

1995

-

16.2

16.9

Bangladesh

1996

-

13.9

15.7

Brazil

1996

16.2

17.9

20.4

Dominican Republic

1996

15.1

15.2

16.3

Source: calculated from DHS III

Although Meekers (1993) by using DHS II data found a large variation in the timing of first sexual intercourse in sub-Saharan Africa, this study found a higher uniformity of


231

mean age at first sexual intercourse in the mid 90s ranging between 14.2 years in Comores and 16.2 years in Zimbabwe. However, most countries’ mean age at first sexual intercourse is around 15 years. Every country follows the trend that sexual intercourse on average is first followed by marriage, and later by first birth. But the differences between mean age at first sexual intercourse and first marriage differ to some extent from 0.2 years to 1.3 years for Kenya and Zambia respectively. This indicates that in sub-Saharan Africa, there is a high incidence of premarital sexual activity which leads to premarital births because in most cases these premarital sexual intercourse takes place without effective contraception (not pregnant protected). It is believed that unmarried adolescents in sub-Saharan Africa are more likely to be sexually experienced than are those in Latin America and the Caribbean, and much more likely than those in Asia and Oceania (McCauley and Salter, 1995).

7.3. Socio-economic Factors Associated with Adolescents’ Age at first Motherhood

Adolescents in rural areas have a lower mean age at first birth than those in urban areas. But this urban-rural difference might be attributed to many factors, e.g. education. However, it is believed that women’s mean age at first birth may be rooted in the behaviour acquired during childhood. It seems that women who grew up in Dar es Salaam had the lowest mean age at first birth. According to Table 63 , adolescents who grew up in Dar es Salaam, give birth 0.9 years earlier than those who grew up in regional towns. We expected adolescents who grew up in Dar es Salaam to have the highest mean age at first birth as they live in a more modern environment than others, and have better access to mass media and contraceptives. On the other hand social control might be the lowest in Dar es Salaam.

In the Tanzanian context, the head of a household as a rule is an adult man, no matter if he is the bread earner or not. If there is no adult male resident, a woman might be the head of the household. Female-headed households are usually single-parent households. Adolescents from households headed by females have a lower mean age at first birth than adolescents from male-headed households. This may reflect that adolescents from single parents engage in unprotected sexual activities very early compared to their counterparts from families of married couples. These adolescents try to imitate the


232

behaviour of their parent since a single parent mother is more likely to have various sexual partners than a married woman.

It can be seen from Table 69 that adolescents with no education have a lower mean age at child bearing than those with complete primary or secondary education. The mean age at child bearing is 0.7 years below the average for the sample of adolescents. On the other hand, completing primary education raises the age at first child bearing by 0.1 years above the sample mean. Attendance of secondary school raises the age at first child bearing by almost one year. This confirms the hypothesis that the higher the level of education, the higher the age at first child bearing. However, the low level of education is not the cause of early child bearing. Rather the coincidence of the two; and usually a characteristic of living in impoverished and rural environments. It seems that adolescents with incomplete primary education have the same mean age at first birth as those who did not go to school. This can be expected since they left school before they reached their menarche. The reason why adolescents who have attended secondary school give first birth later can be easily understood. In Tanzania schools do not allow visible pregnant women or mothers to continue with their education. Furthermore, the possible reason, why those, who have completed primary education, give birth earlier than those with secondary education and above might be due to the absence of occupation or further education after seven years in school. Then most of primary school leavers either get married or engage in unprotected sexual activities that result in early premarital births.

While Protestants’ mean age at child bearing is 0.1 years above average (16.8 years), other religious groups have a mean age at first birth below the average. Notably, among them are Moslems with 16.7 years and Catholic adolescents with 17.0 years. Perhaps the high mean age at first child bearing observed among Protestants is due to the fact that most followers of Christian religions spend longer periods in school than Moslems.


233

Table 69: Socio-economic factors associated with adolescents’ age at first child bearing in Tanzania

 

T10

T25

T50

Trimean

T75

T90

Spread

N

Place of residence
Urban

15.1

16.1

17.1

17.1

17.9

18.7

1.4

77

Rural

14.7

15.8

16.9

16.9

17.9

18.7

1.4

284

Child Place of residence
Dar es Salaam

13.9

15.4

16.6

16.6

17.7

18.6

1.7

25

Other Urban

16.0

16.7

17.5

17.5

18.2

18.8

.9

33

Rural

14.8

15.8

16.9

16.9

17.9

18.7

1.4

302

Head of Household
Male

14.9

15.9

17.0

17.0

17.9

18.7

1.4

302

Female

14.5

15.7

16.9

16.8

17.8

18.6

1.6

58

Education
None

14.2

15.4

16.5

16.5

17.6

18.5

1.35

96

Primaryincomplete

14.4

15.3

16.4

16.4

17.4

18.2

1.4

77

Primary complete

15.3

16.3

17.3

17.3

18.1

18.8

1.3

182

Secondary +

17.1

17.4

18.0

18.0

18.6

18.9

.55

6

Religion
Moslem

14.7

15.7

16.7

16.7

17.7

18.6

1.4

126

Catholic

15.0

15.9

17.0

17.0

17.9

18.7

1.4

104

Protestant

15.1

16.2

17.3

17.3

18.2

18.8

1.3

92

Other

13.6

15.6

16.6

16.6

17.5

18.3

1.7

36

Marital
Never married

14.8

15.8

16.9

16.9

17.8

18.6

1.4

101

Married

14.9

15.9

17.0

17.0

18.0

18.8

1.5

231

Widow

-

-

Divorced

14.4

15.5

16.8

16.7

17.7

-

1.6

18

Not living together

15.0

15.8

17.0

16.9

17.6

18.0

1

10

Polygyny
Monogamous

14.9

15.8

16.9

16.9

18.0

18.8

1.5

176

Polygamous

14.9

15.9

17.1

17.0

17.9

18.5

1.3

54

Source: calculated from 1996 TDHS

It can be observed from Table 69 that women with no education and those who did not complete their seven years of schooling also had lower mean of age at first birth than those who completed their primary or secondary education and above. The mean for those with secondary education and above is 1.1 years above the sample mean. Those without education had 0.4 years below the sample average. On the other hand, completing primary education raises the age at first birth by 0.7 years compared to those


234

who did not attend any formal education. Attending secondary school or above raises age at first birth by 1.5 years for Tanzanian adolescents. This confirms the hypothesis that the higher the level of education, the higher the age at first birth. Adolescents from rural areas had their first child bearing experience earlier than their counterparts in urban areas, except Dar es Salaam adolescent women. Since modernisation usually starts in urban areas before spreading its influence gradually into the rural areas, in a transitional society such as Tanzania, one would expect the patterns of early child bearing to be more pronounced in rural areas than in Dar es Salaam.

Protestants have higher mean age at first birth than others (Catholics and Moslems), Moslems seem to have the lowest mean age at first birth (16.7 years). The variations may be attributed to the fact that most of the Christians spend longer periods in school than Moslems. This is in line with the findings that most of adolescent respondents (30.8 percent) who did not attend any form of formal education were Moslems.

The analysis of adolescents show that each of the background characteristics including education, age at first intercourse, religion, place of residence, and current age tend to influence adolescent age at first birth. In the following parts, the relationship between a combination of all these variables and age at first birth will be determined in a more complex way. The relationships found earlier between each of the background characteristics and age at first birth may be different when all the variables are combined. Normally, this comes about as a result of the interrelationship between the independent variables. This leads us to supplement bivariate analysis with multivariate analysis.

The background characteristics are the independent variables. The unit of analysis is ever given birth adolescent. For each background characteristic, dummy variables are created with one category being selected as the reference category. As such they are omitted from the equation. This type of regression is the one in which coefficients are compared.


235

Table 70: Regression results (unstandardised coefficients) of the relationship between adolescent’s age at first birth and some selected socio-economic variables

Variable

Name

Dummy

Unst. coef.

Std. err.

Current age

Age (continuous)

V012

0.548***

0.066

Education

None

RC

Primary incomplete

E2

0.183

0.180

Completed primary

E3

0.440**

0.151

Secondary+

E4

0.862*

0.488

Age at first sexual intercourse

At union

I2

0.774***

0.191

<15

RC

15-17

I4

0.550***

0.161

18-19

I5

1.408***

0.244

Religion

Moslems

R1

-0.097*

0.146

Catholics

RC

Protestants

R3

0.282*

0.155

Place of residence

Urban

U1

-0.082

0.157

Rural

RC

Constant

5.985***

1.166

R2 (adjusted) = 0.361 and Durbin-Watson = 2.143

*** p<.001

** p<.01

p<.05

Source: calculated from 1996 TDHS

The relationship between current age and age at first birth as found earlier on was confirmed in the regression analysis. Current age is found to be positively related to age at child bearing: An increase in two years of age leads to a rise of one year in first birth of an adolescent; the relationship is significant.

The results in Table 70 show that education raises the age at first birth. Adolescent women who completed primary education have a higher age at first birth compared to those with no education; the difference is statistically significant. Secondary education raises adolescents’ age at first birth; the difference is statistically significance.

Age at first intercourse is positively related to age at first birth; the relationship is highly significant. A point to note is that women who experience their first sexual intercourse within marriage give birth later than those who experience sexual intercourse outside marriage and before attaining age 15. Late experiences in sexual intercourse explain much the difference in adolescent age at first birth. Education on the other hand plays a significant role.


236

The results of the regression as shown above support the hypothesis that Christians generally had their first births later than Moslems. A Moslem adolescent tends to have a birth earlier than a Catholic adolescent. Being a Protestant tends to increase the age at first birth compared to Catholic adolescents. The relation is significant. The type of the place of residence is not significantly related to age at first birth for adolescents.

Table 71 confirms the results above. Literate women are 2 times less likely to give birth before age 15 than illiterate adolescents. Protestant adolescents are 2.3 times less likely to give birth before attaining age 15 than Moslems.

Table 71: Logistic regression odds ratio predicting the relative risk that a woman had a live birth before age 15

Variable

Odds ratio

LiteracyLiterateSemiliterateIlliterate

.53*

1.55

RC

ReligionMoslemCatholicProtestantOthers

RC

.73

.43*

.85

Place of ResidenceUrbanRural

.96

RC

Age15-1617+

11.06***

RC

*** P< .001

** p< .01

* p< .05

Source: calculated from 1996 TDHS


237

7.4. Adolescent Premarital Child bearing

The results of Table 72 show that 10.4 percent of all adolescents had premarital<36> births and 1.6 had premarital pregnancies at the time of interview. Therefore, 12 percent of all adolescents had experienced premarital pregnancy in comparison to 15.8 with postmarital pregnancy experience. Of these, 6.3 percent were never married and the rest was married, widowed or divorced but had their first birth before their first marriage. Almost half of the adolescents (46 percent), who had their first pregnancy at the time of the survey, admitted that the pregnancy was unwanted. From these findings, it is evident that the prevalence of premarital births increases with advancing age.

Unmarried adolescents who got pregnant seem to experience sexual intercourse very early as 21 percent of adolescent with history of premarital births had their first sexual experience before the age of 14 and 22 percent between age of 14 to 15. The effect of the level of education on prevalence of premarital births is not clear. However, the prevalence of premarital births seems to be higher for illiterates in comparison to literate. The role of religious denominations on the occurrence of premarital births was investigated. Moslems seem to have a slightly higher proportion of adolescents with a history of premarital births followed by Catholics. Protestants have the least of all (7.4 percent). Place of residence seems to have no effect on premarital births of adolescents in Tanzania.

Table 73 shows the results of the logistic regression model predicting the effect of background variables on the likelihood that an ever-married woman had intercourse before first marriage. Literacy, age, and the age at which a woman marries have the strongest correlates of the likelihood of having had premarital sexual intercourse. Religion has a weak one. Exposure to the risk of premarital sexual intercourse increases with increasing age at which a woman first marries. Women who marry at a later age are significantly more likely to have engaged in premarital sexual relations than women who marry at a younger age. Literate women are 1.5 times more likely to have had premarital sexual intercourse than women who are illiterate: the effect is significant. This may be due the fact that modern education exposes adolescents to different values,


238

and the school environment enables them to interact more with partners of the opposite sex (Bauni, 1990; Ocholla-Ayayo et al., 1990).

Table 72: Percentage distribution of adolescents by birth status

Characteristics

Variable

Premarital

Post-marital

No births

N

Current age

15

1.0

-

99.0

289

16

5.1

3.9

90.9

408

17

5.0

4.7

90.4

343

18

16.8

13.1

70.1

358

19

20.9

33.4

45.7

335

15-19

10.4

10.5

79.1

1733

Literacy

Literate

9.4

7.9

82.7

1171

Semiliterate

8.3

11.5

80.3

157

Illiterate

11.8

19.8

68.4

399

Religion

Moslem

11.6

11.4

77.0

551

Catholic

10.1

8.7

81.2

552

Protestant

7.4

12.1

80.5

473

Others

9.3

14.7

76.0

150

Place of Residence

Urban

9.9

8.9

81.2

415

Rural

9.9

11.7

78.5

1318

Age at first sexual intercourse

<14

22.1

15.1

62.8

86

14-15

21.4

18.7

59.9

299

16-17

22.0

16.6

61.4

259

18-19

25.4

31.7

42.9

63

At first union

-

57.6

42.4

125

Source: calculated from 1996 TDHS

Catholics appear to be 1.1 times more likely to have premarital sexual intercourse than Moslems. Protestants are 1.2 times less likely to have premarital sexual intercourse than Moslems. This may be consistent with the strong emphasis on premarital virginity in the Islamic society. Urban residence has a positive effect on premarital sexual intercourse as urban residents are 1.2 times more likely to have had premarital sexual intercourse than their rural sisters; the effect is statistically significant. This finding can be attributed to the modernisation processes in urban areas which tends to disassociate people from traditional social controls. It is believed that the traditional social controls of adolescent sexual behaviour are less effective in urban areas than in rural areas (Adeokun, 1990).

In general, traditions and cultural conventions in Tanzania are weakening. Whereas in the past premarital virginity was an important cultural condition for young women to


239

become married, this seems to be disappearing. Today, it seems to become important that a woman proves her fecundity prior to establishing a union. More and more men engage into premarital sexual relationships with young women, whom they make believe that they would marry them in the future. Very often, if these women become pregnant, they are left alone and unmarried by their sexual partners, who do not want to take over the responsibility for the pregnant woman. The weakening of cultural conventions results in changes of the behaviour of people, often causing new problems and challenges for a society. Ever since premarital virginity has been losing its importance, the number of women who have to bear their children without getting married to the father increases. On the other hand, pregnancy is increasingly perceived by women as a means to pressure men to marry them.

Table 73: Odds ratio predicting the relative risk that an ever-married woman had premarital sexual intercourse

Variable

All women

Adolescents

LiteracyLiterateSemiliterateIlliterate

1.51***

1.24*

RC

1.37

.81

RC

ReligionMoslemCatholicProtestantOthers

RC

1.11

.86

1.07

RC

.81

.63

.47*

Place of ResidenceUrbanRural

1.23**

RC

1.13

RC

Age at first Marriage<1515-1718-1920-2122-2425+

.34***

.43***

.51***

.55**

.66*

RC

.66

.60

RC

*** P< .001

** p< .01

*p< .05

Source: calculated from 1996 TDHS

The effect of background variables on the likelihood that an unmarried adolescent is sexually experienced shows that sexual experience increases with age. Adolescents are six times less likely to be sexually experienced than women older than age 19.


240

Unmarried Protestant women are less likely to have had sexual experience than Moslems; the effect is statistically significant. Although the effect is not significant, Catholics are less likely to have had sexual experience than Moslems. Therefore unmarried Moslem adolescents are more likely to have sexual experience than others. Unmarried literate adolescents are less likely to experience sexual intercourse than illiterate women. Unmarried urban women are more likely to experience sexual intercourse before marriage than their rural counterparts; the relationship is highly significant. Urban adolescents are 2.2 times more likely to experience sex than unmarried rural resident adolescents.

Table 74: Odds ratio predicting the relative risk that a never married woman is sexually experienced

Variable

Odds ratio

LiteracyLiterateSemiliterateIlliterate

.66*

.75

RC

ReligionMoslemsCatholicsProtestantsOthers

RC

.89

.70*

1.12

Place of ResidenceUrbanRural

2.21***

RC

*** P< .001

** p< .01

*p< .05

Source: calculated from 1996 TDHS

Table 75 shows the result of a logistics regression model estimating the effect of background variables on the likelihood that a woman has a premarital birth. Due to the fact that the reference is the interval between age at marriage and age at first birth, this model excludes never married women. From the fact that adolescent contraceptive use has remained low in Tanzania (Chapter 6), these differentials in premarital sexual activity are reflected in the level of premarital child bearing. Adolescents are 1.09 times more likely to have had a premarital birth than adults were during their adolescence; the effect is significant.

Age at first marriage is the strongest correlate of premarital child bearing for those who eventually marry. Women who marry late are exposed to the risk of premarital


241

pregnancy for a longer period of time than women who marry early and they are much more likely to have a child before marriage than women who marry at a younger age. For example, women who marry before attaining 15 years are 17 times less likely to have a first child before first marriage than those who marry at age 25 and more; the effect is highly significant statistically.

Premarital child bearing is more common among literate women than among those who are illiterate. But the effect on the likelihood of having a premarital birth is not significant after controlling for the other variables. This finding does not imply that education per se is the cause of this high level of premarital child bearing but that it is due to the fact that literate women also tend to marry later than illiterate women. Urban residence has a significantly positive effect on premarital child bearing. This finding is consistent with the literature that suggests that premarital child bearing is a modern phenomenon that increases with socio-economic development (Cherlin and Riley, 1986; Kulin, 1988).

Table 75: Odds ratio predicting that a woman had her first child before first marriage

Variable

All women

Adolescents

LiteracyLiterateSemiliterateIlliterate

1.01

1.18

RC

1.51

1.17

RC

ReligionMoslemsCatholicsProtestantsOthers

RC

1.04

.90

.78*

RC

.91

.63*

.73

Place of ResidenceUrbanRural

1.05***

RC

.76

RC

Current age 15-1920+

1.09**

RC

Constant


242

Age at first Marriage<1515-1718-1920-2122-2425+

.06***

.10***

.19***

.26***

.43***

RC

.10**

.57

RC

*** P< .001

** p< .01

*p< .05

Source: calculated from 1996 TDHS

Being Moslem have a positive effect on premarital child bearing compared to other denominations in Tanzania. However, the findings in this chapter seem to suggest that premarital child bearing is a result of degradation of sexual morals as adolescents are now exercising less restrain than was the case in the past, especially educated ones. Meekers (1993) argues that what has changed is not adolescent sexual behaviour and child bearing itself, but rather the social context in which this occurs.

Table 76: Odds ratio predicting that an adolescent did not like the timing of her first pregnancy

Variable

Odds ratio

LiteracyLiterateSemiliterateIlliterate

7.31*

5.23

RC

ReligionMoslemCatholicsProtestantsOthers

RC

.07**

.27

.72*

Place of ResidenceUrbanRural

.67***

RC

Age 15-1718-19

.69**

RC

Conception StatusPremaritalMarital

8.48**

RC

*** P< .001

** p< .01

*p< .05

Source: calculated from 1996 TDHS


243

On this subject, Cherlin and Riley (1986) raised the question whether adolescent mothers want to give birth. The answer to this question can be reflected in the initial analyses of the DHS data, which suggested that more than 40 percent of adolescent births in Botswana, Ghana, Kenya, Liberia, and Togo were unwanted (Njogu and Martin, 1991). This study found (Chapter 4) that a high proportion of adolescent women in Tanzania had unintended births (27 percent of all births to adolescents), either not wanted at the time the birth occurred or not wanted at all. Therefore some of the adolescents faced with an unplanned pregnancy may turn to illegal abortion. When the abortion is performed by an untrained person, or under unsanitary conditions, the procedure can result in illness, infection, infertility, and even death. Since young women often cannot pay the price of an abortion performed by a medically or paramedical trained person, they risk their lives using fatal methods. Intakes of huge amount of chroloquine tablets were often found to be the cause of death of pregnant adolescent women who wanted to terminate their pregnancies (Justesen, et al., 1992).

Literate adolescents are 7 times more likely to be displeased with their first pregnancies than illiterate adolescents. It is interesting to note that Moslems have a greater likelihood of being displeased with the timing of their first pregnancy while Catholic adolescents are 14 times less likely to be displeased with the first pregnancies than Moslem adolescents; the difference is statistically significant. Although not statistically significant, it seems than Protestant adolescents are 4 times less likely to be unhappy with their first pregnancies than Moslems. On the other hand, those who are neither Christians nor Moslems are 1.4 times less likely to be unhappy with their first pregnancies than Moslem adolescents.

It seems that rural adolescents have a greater likelihood of being displeased with their first pregnancies than their urban counterparts; the relation is statistically significant. Urban adolescent women are 1.5 times less likely to be unhappy with their first pregnancies than rural adolescents. It is not expected that younger adolescents are happier than 18-19 year-old adolescents. However, in this study it was found that adolescent women (15-17 years old) are 1.4 times less likely to be unhappy with their first pregnancies than 18-19 years old adolescent women. This might be from a fact that the older a woman grow the more she sees about the consequences of early pregnancy. Adolescent women who conceive their first pregnancy before their first marriage are


244

most likely to be displeased with the timing of the first birth. Moreover, they are 8 times more likely to be unhappy with their first pregnancies than married ones. This fact conflicts with men’s expectation that women are to prove their fertility before marriage (Dynowski-Smith, 1989; Ocholla-Ayayo et al., 1990). From this finding, it is evident that women do not share this sentiment. Furthermore, these women are expected by their parents and others not to have sex before marriage.

Table 77: Percentage of adolescents in sub-Saharan Africa and selected developing countries by marital status at the birth of the first child

Country

Year

No births

Never

married

X

Y

Z

Premarital

Conception

Premarital

Births

Marital births

N

Benin

1996

80.6

1.4

0.5

3.9

13.7

5.8

1.9

17.6

1,075

CAR

1994

71.6

6.2

0.4

3.0

18.7

9.6

6.6

21.7

1,288

Comores

1996

92.7

0.1

0.2

2.4

4.6

2.7

0.4

7.0

844

Côte d’Ivoire

1994

71.5

12.0

2.9

2.5

11.1

17.4

14.9

13.6

1,961

Ghana

1993

81.4

3.6

0.5

5.1

9.3

9.2

4.1

14.4

803

Kenya

1993

83.2

7.4

2.0

2.6

4.7

12.0

9.4

7.4

1,754

Mali

1995

66.4

4.4

0.8

6.1

22.4

11.3

5.2

28.5

1,883

Tanzania

1996

79.2

5.9

1.4

2.5

11.1

9.8

7.3

13.6

1,732

Uganda

1995

65.9

3.3

3.0

4.5

23.3

10.8

6.3

27.9

1,606

Zambia

1996

76.2

5.8

2.2

4.6

11.2

12.7

8.1

15.8

2,003

Zimbabwe

1994

85.3

2.9

1.0

3.6

7.3

7.4

3.8

10.9

1,472

Egypt*

1995

48.9

-

-

1.5

49.8

1.5

-

51.3

673

Bangladesh*

1996

49.6

-

-

0.5

49.8

0.5

-

50.3

1,416

Brazil

1996

85.7

2.7

1.3

3.7

6.7

7.6

3.9

10.4

2,464

Dominican Republic

1996

81.7

0.9

0.4

2.3

14.8

3.6

1.3

17.1

1,801

* ever married women only

X = first births occurred before first marriage for ever married women; Y = first births within marriage but less than 9 months after first marriage; Z = first birth of 9 months or more after the first marriage;

Premarital first conception = Never married +X + Y;

Premarital first birth = Never married + X;

Legitimate first conception = Z

Source: calculated from DHS III

On the other hand, scholars suggest that women who become pregnant before marriage might use pregnancy to urge the father to marry her before the birth of the child (Karanja, 1987; Obbo, 1987). A comparison between the Tanzanian situation and other countries in sub-Saharan Africa that conducted DHS III could be useful in clarifying this matter.

The percentage of sub-Sahara African adolescents who have not yet experienced a live birth varies between countries. It ranges from 65.9 percent in Uganda to 92.7 85.3 percent in Comores. These figures mean that in sub-Saharan Africa countries at the time


245

of survey more than one in five female adolescents were already mothers. The percentage of unmarried adolescents in sub-Saharan Africa with at least one live birth ranges from 0.1 percent in Comores to 12.0 percent for all adolescents in Côte d’Ivoire. Other countries with more than 5 percent of never married adolescents with experience of a live birth include Kenya (7.4 percent), the Central African Republic (6.2 percent), Tanzania (5.9 percent), and Zambia (5.8 percent). It is interesting to note that while Uganda has a low percentage of adolescents without any birth, most of adolescents eventually got married. Only 3.3 percent adolescents experienced births and were still not married at the time of the survey.

Women, who had their first birth out of wedlock but eventually got married later, range from 0.2 percent in Comores to 3 percent in Uganda. This means that in most sub-Saharan Africa countries premarital virginity is no longer a strong precondition for a first marriage as a woman would be able to get married even with a child especially in countries like Uganda. However, the share of women in sub-Saharan Africa who conceive outside marriage ranges between 2.7 percent in Comores to 17.4 percent in Côte d’Ivoire. Adolescent premarital births in sub-Saharan Africa ranges between 3.8 percent of all adolescents in Zimbabwe to 14.9 in Côte d’Ivoire. Comores and Benin have the lowest percentages (0.4 and 1.4 percent respectively).

If we concentrate on adolescents who gave birth only, premarital pregnancies seem to be a major problem in sub-Saharan Africa countries. Almost more than half of all adolescents’ first pregnancies occurring in sub-Saharan Africa countries are among unmarried adolescents. For example 72 percent of all first adolescent pregnancies in Kenya were conceived outside marriage and premarital birth amounts to 56 percent of all adolescent births. However, most of these premarital births (44 percent) among adolescents who were still single at the time of the survey.


246

Table 78: Percentage distribution of first births by adolescent marital status at the time of the birth in sub-Saharan Africa and selected developing countries

Country

Year

Never married

X

Y

Z

Premarital conceptions

Premarital births

N

Benin

1996

7.2

2.4

20.1

70.3

29.7

9.6

209

CAR

1994

21.9

1.4

10.7

65.8

33.9

23.2

366

Comores

1996

1.6

3.2

32.3

62.9

37.1

4.8

62

Côte d’Ivoire

1994

42.0

10.2

8.8

39.0

61.0

52.2

559

Ghana

1993

19.5

2.7

27.5

50.3

49.7

22.1

149

Kenya

1993

44.1

11.9

15.6

28.1

71.5

55.9

295

Mali

1995

13.1

2.4

18.2

66.5

33.6

15.5

633

Tanzania

1996

28.3

6.6

12.2

53.2

47.1

34.9

361

Uganda

1995

9.7

8.8

13.3

68.4

31.8

18.4

548

Zambia

1996

24.5

9.4

19.3

47.0

53.2

34.0

477

Zimbabwe

1994

19.4

6.5

24.4

49.3

50.2

25.8

217

Egypt *

1995

-

-

2.9

97.4

2.9

-

344

Bangladesh*

1996

-

-

1.0

98.9

1.0

-

713

Brazil

1996

18.7

8.8

25.8

46.7

53.3

27.5

353

DR

1996

4.8

2.1

12.7

80.6

19.7

7.0

330

* ever married women only

X = first births occurred before first marriage for ever married women; Y = first births within marriage but less than 9 months after first marriage; Z = first birth of 9 months or more after the first marriage;

Premarital first conception = Never married +X + Y;

Premarital first birth = Never married + X;

Legitimate first conception = Z

Source: calculated from DHS III

Countries with severe problems of adolescent premarital first births (more than 25 percent) in sub-Saharan Africa include Kenya (56 percent), Côte d’Ivoire (52 percent), Tanzania (35 percent), Zambia (34 percent), and Zimbabwe (26 percent). This means that more than one in four adolescent mothers is single. Comores have the lowest percentage of adolescent premarital first births, 5 percent, and Benin, 10 percent. From the data we might conclude that among every four first born children of adolescents in sub-Sahara African, one will have unmarried mother.

However it is interesting to note that in sub-Saharan Africa for those adolescents who conceive out of wedlock more than half give birth before their first marriage (Côte d’Ivoire, Kenya, Tanzania etc). However, in a few countries like Comores, Benin, Ghana, and Mali more than half of those who conceive before marriege, eventually marry before the first birth.


247

7.5. Awareness and use of contraceptives

Concerns about the adverse consequences of early child bearing, particularly for unmarried women, and the risks of contracting sexually transmitted diseases have led to renewed interest in the contraceptive<37> and sexual behaviour of adolescents. The fact that adolescents have an enormous impact on future population growth make an understanding of the extent to which young women are aware of and use contraceptives a significant policy issue. This section will shed light on contraceptive use among adolescents especially unmarried ones. Little however is known on the subject in sub-Saharan Africa (McDevitt, 1996).

Until recently, contraceptive information and services in Tanzania were only available to married women while single women and adolescents did not have access to these services. However, the family planning policy has now been changed such that some of these services are now available to all sexually active persons regardless of their age and marital status. Despite this positive change in policy, there is a paucity of information on the availability of these services to the adolescents mainly because they are still being provided through the MCH/FP program. As the program is adult oriented, young women tend to perceive it as being meant for adults and hence not accessible to them.

In this section therefore, the unit of analysis are all adolescent women who are sexually experienced. This is due to the fact that potential or actual contraceptive users are those who are at the risk of becoming pregnant. An adolescent who has never experienced sexual intercourse yet is not in need to use contraceptives. In the TDHS the question used was ‘which ways or methods of family planning have you heard about’. In fact, it means awareness than rather knowledge as knowledge encompasses awareness and how to use the method. For the use of contraceptives our denominator will be further refined by excluding those who were pregnant at the time of the survey.

Awareness of family planning methods among adolescents in Tanzania was 81 percent but very few of them use contraceptives. Only 6.4 percent of adolescents, who are aware of any contraceptive method, do contracept. Results have shown ( Table 79 ) that the awareness as well as the use of contraceptives tends to increase with increasing age


248

of the adolescent. It was found that adolescents who use contraceptives use mainly the pill (40 percent), condoms (33 percent), and injections (18.2 percent). Awareness of contraceptives tends to vary with the level of education, The higher the level of education the higher is the awareness of adolescents on contraceptive methods. The same applies for the use. Only 1 percent of those who have never attended school use contraceptives compared to 26.3 percent of those who has secondary or higher education.

Awareness of contraceptives varies with adolescents’ place of residence. Adolescents residing in urban centres had higher awareness of contraceptive than those residing in the rural areas. Likewise 15.5 percent of urban adolescents use contraceptives while only 3 percent of rural adolescents contracept. When we separated Dar es Salaam from the other urban areas, we find that 15.6 percent of Dar es Salaam residents are using contraceptives, mainly pill, 12.5 percent of other urban residents use contraceptives, mainly pill and injections, while only 4.5 percent rural residents use contraceptives. Religious affiliation was also found to be associated with awareness and usage of contraceptives. Female Moslem adolescents have a high level of contraceptive awareness and use compared to those of other denominations. It is interesting to note that 8.7 percent of Catholic adolescents use contraceptive. While only 0.9 percent of Protestant adolescents contracept, Protestants are more aware than Catholics. This result was not expected especially for use of contraceptives. We expected Catholics to have the lowest level of all due to the Vatican’s refusal to accept contraceptive use (Chapter 6). Perhaps new generations are against religious directives on modern contraceptive use. However, Catholics use mainly condoms. Protestants use mainly condoms, while Moslems use mainly the pill.

It seems that there is a positive relationship between age at first intercourse, marriage, and birth with the awareness of contraceptives. Contraceptive use has no clear pattern. It is very interesting to note that divorced adolescents use even more contraceptives than currently married ones. It seems that divorced adolescents had sexual partners besides their official husbands and use contraceptives not for spacing or limiting but to prevent out-of-wedlock pregnancies (refer the definition of divorced in Chapter 6). Widowed women seem not to use contraceptives as they are not accountable for anybody after the death of their spouse. Married adolescents in polygamous unions use less contraceptives


249

than those in monogamous unions. This situation can be attributed to the compensation effects or the catch-up hypothesises (Chapter 5). Most adolescents in polygamous union are the second or more in rank to the husband; 93 percent, while first wives adolescents account for 7 percent of all adolescents in polygamous union. All of them use mainly the pill.


250-251

Table 79: Percentage distribution of adolescents’ contraceptive awareness and use

Characteristic

Awareness

Use

Method Specific use

Pill

IUD

Injec

D/F/

Con

Steri

N

Age

15

72.5

2.0

100

1

16

63.8

3.3

40.0

60.0

5

17

78.5

5.9

50.0

12.5

37.5

8

18

82.1

7.7

57.1

7.1

28.6

14

19

91.3

11.3

37.0

7.4

25.9

29.6

27

15-19

81.0

7.4

40.0

5.5

18.2

34.5

55

Education

None

58.1

1.2

100

2

Primary incomplete

77.5

5.4

40.0

10.0

10.0

40.0

10

Primary complete

91.0

9.1

40.6

3.1

28.1

3.1

25.0

32

Sec +

97.4

29.4

30.0

10.0

60.0

10

Literacy

Illiterate

80.8

1.4

100

3

Semiliterate

83.1

4.5

33.3

66.7

3

Literate

90.0

11.0

37.5

6.3

16.7

2.1

37.5

48

Current school attendance

Out of school

88.0

13.5

41.3

6.5

19.6

2.2

30.5

46

In school

77.3

16.9

37.5

12.5

50.0

8

Place of Residence

Urban

94.2

17.1

41.7

8.3

19.4

2.8

27.8

36

Rural

75.9

3.5

42.1

15.8

42.1

19

Specific Place of residence

Dar es Salaam

93.8

16.7

54.5

18.2

9.1

9.1

9.1

11

Other Urban

96.2

14.3

46.2

30.8

13

Rural

77.4

5.4

33.3

3.3

16.7

3.3

30

Religion

Moslem

87.9

11.7

45.2

6.5

22.6

3.2

19.4

31

Catholic

83.5

10.2

5.0

15.0

50.0

20

Protestant

84.3

1.1

33.3

3

None/others

43.0

0

Age at first intercourse

<12

51.0

.1

12-13

76.7

6.7

75.0

25.0

4

14-15

79.7

6.0

44.4

16.7

38.9

18

16-17

88.0

8.5

34.8

8.7

17.4

4.3

34.8

23

18-19

88.9

6.3

50.0

50.0

4

At union

72.8

4.0

16.7

16.7

33.3

33.3

6

Age at first marriage

<12

25.0

0

12-13

72.2

5.6

100

1

14-15

71.9

6.3

25.0

25.0

25.0

25.0

8

16-17

84.1

5.8

76.9

7.7

7.7

7.7

13

18-19

89.0

2.4

50.0

50.0

2

Age at first birth

<12

-

-

12-13

60.0

0

14-15

88.2

7.8

50.0

25.0

25.0

4

16-17

83.7

11.0

45.0

5.0

30.0

15.0

5.0

20

18-19

89.8

3.9

60.0

20.0

20.0

5

Marital status

Never married

81.0

8.2

32.3

3.2

16.1

3.2

45.2

31

Married

80.0

5.8

50.0

11.1

16.7

22.2

18

Widowed

33.3

0

Divorced

85.7

20.0

75.0

25.0

4

Not living together

100

15.4

50.0

50.0

2

Polygyny

Monogamous

81.5

6.1

46.7

13.3

20.0

20.0

15

Polygamous

73.5

4.8

50.0

25.0

25.0

4

Key: Injec = Injections

D/F/ = Diaphragm, foam and jelly

Con = Condom

Steri = Sterilisation

Source: calculated from 1996 TDHS

Contraceptive use among adolescents in sub-Saharan Africa is still low. It is clear from Table 80 that contraceptive use among adolescents in sub-Saharan Africa ranges from 2 percent in Comores and Central African Republic to 8 percent in Zimbabwe. However, our interest is to investigate the modern contraceptive use to delay the first birth. In this sense, we have to refine the denominator. Adolescents who are sexually experienced, not pregnant and have never given birth are the ones who are at the risk of first pregnancy hence can use contraceptives to delay the first birth. It was found that between 2.4 percent in Kenya to 23.1 percent in Comoroes adolescents use contraception to delay the first birth. But if we refine our denominator further to include only adolescents who are at risk of premarital birth by eliminating those who are or


252

have been married, we find that more adolescents use contraception to avoid premarital births. It ranges from 3 percent in Kenya to 27 percent in Comores.

Table 80: Adolescent modern contraceptive use in sub-Saharan Africa countries and selected developing countries

Country

Year

Use (General)

Use to delay first birth

All

Not currently married

Benin

1996

2.1

5.1

5.8

Central African Republic

1994

2.0

4.0

5.1

Comores

1996

2.0

23.1

26.5

Côte d’Ivoire

1994

5.2

10.7

12.5

Ghana

1993

5.0

9.0

9.5

Kenya

1993

2.4

2.4

2.9

Mali

1995

3.2

7.1

12.4

Tanzania

1996

3.1

6.6

8.9

Uganda

1995

3.4

9.2

16.9

Zambia

1996

4.7

7.3

8.1

Zimbabwe

1994

7.5

10.4

12.6

Egypt

1995

15.6

0.0

0.0<38>

Bangladesh

1996

18.0

12.1

0.0<39>

Brazil

1996

13.2

45.8

44.0

Dominican Republic

1996

8.2

27.5

21.3

Source: calculated from 1996 TDHS

With the exception of Côte d’Ivoire, many sub-Saharan Africa countries have a high percentage of premarital births and a low contraceptive prevalence rate reducing the number of young women who are able to delay their first birth. It seems that there is a relationship between contraceptive use and premarital births. The other finding from Table 80 is that in sub-Saharan Africa’s married adolescents usually do not delay their first births. In contrast even married youngsters from Bangladesh, Brazil and the Dominican Republic use modern contraceptives to delay their first births. This is due to the fact that after removing from the sample currently married adolescents, the percent of contraceptive users drops from its original figure. While in the case of sub-Saharan Africa, the percentage increases after removing currently married women. This is another proof that married women in sub-Saharan Africa do not delay their first births and the meaning of marriage is to have at least one child, but usually a larger number than just one.


253

7.6. Unmet Need and Demand for Family Planning among Adolescents

It has been shown in section 7.1 that premarital child bearing is increasing and expected to increase further. In Table 79 we found that adolescent contraceptive use is still very low. Low contraceptive prevalence might arise due to the fact that the there is a proportion of adolescent women who are exposed to the risk of pregnancy but are not using contraception despite the fact that they want to limit or space their births. The proportion of adolescent women with unmet need along with the proportion of women currently using contraception provide family planning programme managers with information on the magnitude of the potential demand for contraceptives and services.

As we have already seen in Chapter 6, women with unmet need are classified into two groups: those with unmet need for spacing births, and those with unmet need for limiting births. However, the crude measure that currently married women represent the sexually active women among adolescents can give a biased information as this group represents only 23 percent of adolescents ( Table 7 ).

Many scholars have tried to estimate unmet need in different ways. Westoff and Pebley (1981) used WFS data for 18 countries to estimate the unmet need for contraception using twelve different combinations based on exposure and attitudinal criteria. The first measure of unmet need includes currently married women of reproductive age who are exposed to the risk of pregnancy but do not want any more children and are not using an effective method. The second measure is similar to the first except that woman using any form of contraception are classified as having met their need. In each of the ten other measures the definition of unmet need is made more stringent by adding exposure or attitudinal criteria. Measure three includes women, who stated that their desired number of children is less or equal to their actual number; and measure four includes only fecund and non-pregnant women.

The major limitation for Westoff and Pebley's procedure, however, is that the unmet need group includes only women who wanted no more children, that is, unmet need for limiting births. It implies that women with unmet need for spacing births are not included (assumed not to have unmet need for family planning). This can seriously


254

under-estimate the magnitude of unmet need particularly where contraceptive use for spacing purposes is widespread. Westoff and Pebley (1981) acknowledge this problem. Regardless of the shortcomings, their method has been applied elsewhere (see for example Boulier, 1984; Mturi, 1996; Shah and Ahmed, 1982).

Nortman (1982) has developed a model that incorporates both "limiters" and "spacers". Nortman defined birth spacers as women who want to postpone their next pregnancy for a period of one year or more from the time of the interview. The approach measured unmet need for contraception over a one-year period in which fractions of exposure were calculated. The estimates obtained, using data from six developing countries; Bangladesh, Colombia, Costa Rica, Korea, Mexico, and Thailand which participated in the Contraceptive Prevalence Surveys (CPS), suggest that the percentage of fecund and currently married women of reproductive age with unmet need for contraception ranges from 22 percent to 67 percent. The key feature of this method is that it incorporates time as a factor, which allows pregnant or amenorrheic women to rejoin the group of exposed women and require contraceptive protection for at least part of the time span under review (Nortman, 1982). The application of Nortman's method has been very limited because the method produces estimates which are very close to the current-status measures described below despite of its complexity of calculation and description (Westoff, 1988).

The method for estimating unmet need for contraception most commonly used is the one given by Westoff (1988) and latter refined by Westoff and his colleagues (Westoff and Ochoa, 1991; Westoff and Moreno, 1992). This method uses current-status information given by currently married women to estimate the potential demand for contraception for spacing and for limiting births. Pregnant or amenorrheic women, whose pregnancy was unintentional (mistimed or unwanted), are included in the group with unmet need in the recent past. This method has been widely used as a standard procedure for estimating unmet need for contraception using DHS data.

However, Dixon-Mueller and Germain (1992) have argued that the method over-estimates the level of unmet need as (in some cases) a considerable part of women included in the unmet need group are currently not exposed to the risk of pregnancy. For instance, Dixon-Mueller and Germain were suspicious of the findings of Goldman


255

et al. (1989) who estimated a total unmet need for Peru in 1986 to be 29.4 percent while only 7.6 percent of these women were assumed to be currently exposed to the risk of pregnancy, i.e. currently sexually active.

The argument that Westoff's approach over-estimates the level of unmet need for contraception is supported by Bongaarts (1991). Bongaarts suggests a method which adjusts Westoff's procedure downwards (based on two arguments) to give upper and lower boundaries for the actual unmet need. He illustrates the method using 15 DHS surveys. The method assumes that fulfilling the unmet need for spacing reduces the unmet need for limiting by an equivalent amount among women who will have a need for limiting births. It is therefore necessary to have an adjustment factor which is subtracted from Westoff's estimates to show the reduction in the need for limiting births resulting from satisfying spacing demands. This gives the maximum estimate. The minimum estimate is obtained by making a downward adjustment to correct the over-estimation of spacing needs. The point estimate is obtained by taking the mean of the maximum and minimum values which shows that, on the average, 17 percent of currently married women in the 15 countries studied have unmet need for contraception. Westoff's approach gave a comparable figure of 21 percent.

Bongaarts' approach has been criticised that it deals with what will be realised in the future given a steady state rather than giving an estimate of current need which programme managers require (Westoff, 1992). However, Bongaarts argued that his approach ‘...can be interpreted either as the current unrealised contraceptive use that has resulted from unmet demand in the past, or as the future rise in prevalence that could be achieved given present preferences for the timing and quantity of child bearing’ (Bongaarts, 1991:127).

In other words, Bongaarts defines unmet need as the additional contraceptive use that would be required to achieve fertility levels consistent with women's stated reproductive intentions and to eliminate all mistimed and unwanted pregnancies. In this chapter we will adopt and try to modify Westoff's (1988) methodology because the major interest of this analysis is to examine the current situation using most preferably current-status information. Furthermore, the simplicity of the method and the fact that it has been


256

applied in many DHSs makes it possible to compare the estimates. Our main task will be to identify current sexually active women.

The shortcoming of Westoff’s approach is that it distinguishes women who are currently in a union from those who are not. In doing so the approach excludes currently not married (never and formerly married) women from the calculation of unmet need. His argument that married women are more exposed to the risk of conception than are unmarried women might have been true regarding all women, but can lead to false estimate when dealing with a specific group like adolescents, since premarital sexual activities cannot be ignored. It was found in Table 78 that 47 percent of all births to adolescents were results of premarital conception.

In order to capture all sexually active adolescents, we categorised sexually experienced adolescents into two categories: the first category being those adolescents who are currently married or living with partners therefore considered sexually active. We assume that any adolescent who declares to have a sexual partner is sexually active. The second category consists of never married, widowed, and divorced adolescents. The major task is to distinguish sexually active from sexually inactive (whether still attending school or not).

In the 1996 TDHS, there were two questions used to identify those who were sexually active. Those who were not married or living with a man were asked if they had a regular sex partner. From the answers to this question, adolescents whose stated they had a regular sexual partner were categorised as sexually active. The second question that helped to identify sexually active adolescents currently not in union was the date of last sexual intercourse in the 12 months prior to the survey. Those who acknowledged to have had sexual intercourse within one month prior to the interview are assumed to be sexually active. Our assumption here is that the experience prior to an interview reflects the respondent’s lifetime experience. Combining this group of women who are currently not married and not living with partner but have some sexual contacts with those adolescents, who are assumed to be sexually active due to the fact that they are either currently married or living with a man, forms a group of sexually active adolescents who are at risk of becoming pregnant.


257

By definition, adolescent women have an unmet need for contraception if they are not using a contraceptive method but are capable of conceiving, are exposed to the risk of pregnancy, and wish to avoid or to postpone pregnancy. Therefore adolescent women currently using some form of contraception or being sterilised are assumed to have their need met. More explanations can be obtained from Chapter 6.

Figure 25 shows the distribution of sexually active adolescent women according to the components of unmet need for family planning. The figure shows that 65.7 percent of sexually active adolescent women were considered not to be in need of contraception due to the following reasons:

(i) 6.7 percent of sexually active adolescents currently use contraceptives,

(ii) 2.7 percent of sexually active adolescents seem to be infecund,

(iii) 25.9 percent of sexually active adolescents intended to become pregnant,

(iv) 30.4 percent of sexually active adolescents want a child soon.


258

Figure 25: Estimates of Unmet Need for family planning

The total unmet need for family planning among sexually active adolescent women in Tanzania is 34.2 percent; 25.6 percent have unmet need for spacing births and 8.6 percent want to limit births. In Chapter 6 it was found that the unmet need for all women in Tanzania was 27.5 percent, unmet need for spacing being dominant (15.7 percent). This finding shows that in Tanzania there is a higher prevalence of unmet need for family planning among adolescents than older women of more than 20 years.

The total demand for family planning is obtained by adding the unmet need to the current use of contraception. The estimated total demand is therefore 40.9 percent (34.2 percent unmet need and 6.7 percent current contraceptive users). This implies that if every sexually active adolescent woman in Tanzania in need of family planning uses a method, the contraceptive prevalence rate would raise from the current figure of 6.7 percent to 40.9 percent.


259

7.7. Concluding remark

We have observed that adolescents in Tanzania engage in sexual activities at early ages. In the average, adolescent experience first sexual intercourse when there are 15 years. Data from DHS III also showed that mean age at first sexual intercourse ranges from 14.2 years to 16.2 years in sub-Saharan Africa countries. Mean age at first marriage ranges from 14.9 years to 16.4 years. Mean age at first birth in Tanzania was found to be 2 years more than age at first sexual intercourse. In other sub-Saharan Africa countries, the difference was found to be between 0.6 years to 1.6 years. In Brazil it is 4.2 years after first sexual intercourse.

We have observed a high rate of premarital adolescent pregnancies, 47 percent of all adolescent women in Tanzania. Some 12.2 percent of adolescents who became pregnant out of wedlock got married before the birth of their first child. 6.6 percent got married after the birth of their first child and the remaining 28.3 were still not married at the time of the survey. In Kenya 72 percent of all adolescent births were conceived as premarital pregnancies although 15.6 percent got married before the birth. This study found premarital births to be a serious problem among unmarried adolescent in sub-Saharan Africa. In all adolescent births in Kenya, 44 percent of all young mothers (15-19) were singles at the time of the survey, 42 percent in Côte d’Ivoire, 28 percent in Tanzania, and 25 percent in Zambia. Children born out-of-wedlock to adolescent mothers generally are severely disadvantaged because their mothers tend to be illiterate, poor, and in poor health condition (Gyepi-Garbrah, 1985a).

In this study we found that literate women had a higher likelihood of having a premarital sexual intercourse than illiterates. This is due to the fact that marriage is delayed for literate women, so they have a long time before being married. However a single literate woman is less likely to experience sexual intercourse than an illiterate single woman. Literacy raises age at first birth for adolescents. Literate adolescents on the other hand are more likely to be displeased with the timing of their first birth than illiterate adolescent women. The more years an adolescent attends school, the higher the age at first birth.

An urban woman had a higher likelihood of experiencing premarital intercourse than a rural woman. Altogether single urban women are more likely to experience sexual


260

intercourse than their rural counterparts. However, urban adolescents are less likely to be displeased with the timing of the first birth.

Protestant single women are less likely to experience sexual intercourse compared to Moslems. Moslem adolescents are more likely to experience premarital birth than Christians. Moslem adolescents are more likely to be displeased with the timing of the first child than Christians, because Moslems adolescents have the first birth earlier than Christians.

The analysis has shown that the majority of adolescents are aware of modern contraceptives (65.3 percent). However, very few use them (6.4 percent). The awareness of modern contraceptives increases with age: 85 percent of older adolescents (19 years) are aware of at least one method of modern contraceptives in comparison with 46 percent at age 15. 7.8 percent of 19-year-old women in Tanzania use modern contraceptives. The percentage of adolescents who use contraceptives decreases with lower age to 0.3 percent of adolescents at age 15. The fact that awareness of family planing among adolescents is high yet contraceptive use remains as low as 3.1 percent is a dilemma. This shows that awareness alone cannot make sure that adolescents use contraceptives. Further studies need to be carried out to determine factors which might be in play for adolescents not to be using contraceptives although they are aware of contraceptive methods.

Adolescents residing in urban areas were more aware, and use contraceptives to a higher degree than those in rural areas. This might be due to the fact that health facilities are more concentrated in urban areas than in rural areas. Several kinds of mass media too are concentrated in urban areas.

Moslem adolescents are aware and use contraceptives more than other religious affiliates. However, although Protestants are more aware of modern contraception methods than Catholics, it is Catholics who use contraceptives more than Protestant adolescent women in Tanzania. This finding was not expected, as it is known that the Catholic Church is generally opposed to modern contraception.

Adolescent awareness and use of contraceptives did not show any relationship with age at first sexual intercourse, marriage, and birth. Adolescents, who were formerly married,


261

had a higher awareness and use than those who had never been married. This might be due to the fact that those who were formerly married are probably more careful and cautious in preventing pregnancies, as they are not currently married.

High levels of education and literacy were related to contraceptive usage. The explanation for this finding is that it is adolescents attending schools who are more exposed to the adverse effects of unwanted pregnancies. However, adolescents who were not in school anymore are more aware of contraceptives than those who are still in schools. Those who are still in school use contraceptives more than those who do not attend school anymore. The explanation could be that those who are still attending school are afraid of being expelled from school if they become pregnant. Those out of school probably include those who had a premarital pregnancy and were expelled from school and therefore had nothing to lose. Most of those in schools use condoms, the pill and injections. In this study it was found that each country studied, unmarried adolescents use more contraceptives to delay first birth than married adolescents.

The total unmet need for family planning among adolescent women in Tanzania was found to be 34.2 percent. Unmet need for spacing being dominant 25.6 percent. The total adolescent demand for family planning amounts to 37.3 percent of all adolescents.

In general it was found in this study that a gap between first intercourse and first marriage among adolescent women has increased across age cohorts. This increase is predominantly a result of age at first sexual intercourse and rising of age at first marriage. The consequence of this larger gap, in combination with the overall growth in the number of adolescents, is the increasing number of adolescent women who are exposed to the risk of premarital pregnancy. This will be associated with negative outcomes such as school dropping or unsafe abortions.


Fußnoten:
<35> Bureau of Statistics (1992), Population Census\|[lsquor]\| Basic Demographic and Socio-economic characteristics, 7.
<36> Premarital in this study means out of wedlock. In fact we are not sure if those women who became pregnant before marriage, subsequently marry the same men who were responsible for their pregnancy.
<37> The term contraception here will mean modern methods of contraceptives.
<38> All sampled adolescent respondents were married and are not contracepting to delay the first birth.
<39> All sampled adolescents were ever married, 1 was divorced, 27 widowed and the rest were married. Those who are using contraceptive were 66 which meant there were married adolescent women who are contracepting.


[Titelseite] [Abkürzungsverzeichnis] [1] [2] [3] [4] [5] [6] [7] [8] [Bibliographie] [Anhang] [Selbständigkeitserklärung]

© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.

DiDi DTD Version 1.1
a subset from ETD-ML Version 1.1
Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML - Version erstellt am:
Mon Jan 17 19:31:05 2000