CHAPTER 3
FERTILITY
3.1 Introduction
In the SADHS, data were collected on current and completed fertility. Utilising the birth histories of
women interviewed during the survey, this chapter provides direct estimates of current levels, trends and
differentials in fertility. During the interviews the women were requested to provide information on the
total number of sons and daughters they had given birth to that were still living with them, the number
living elsewhere and the number who had died. Interviewers obtained a birth history for each woman,
including details on each live birth separately, according to the month and year of birth, sex and survival
status. In the case of children who had died, their age at death was also recorded. The fertility indicators
reported here are based on the answers provided by women aged 15-49 years regarding their reproductive
histories.
3.2 Fertility Levels
The total fertility rate (TFR) and age-specific fertility rates (ASFRs) are a common measure of recent
fertility. The TFR is defined as the number of children a woman would have by the end of her
childbearing years if she were to pass through these years bearing children at the prevailing age-specific
rates. Table 3.1 shows the age-specific and other aggregate fertility measures calculated from the 1998
SADHS data for the three-year period prior to the survey (roughly 1995- early 1998). The TFR for South
Africa, derived from the survey data was 2.9. Using data from the 1996 population census, indirect
methods yielded a slightly higher TFR level of 3.3 for 1996, two years before the SADHS. (Udjo, 1999).
Another study utilising indirect methods and the same Census data, estimated the TFR at 3.1 (Dorrington
et al., 1999).
Fertility in urban areas (TFR=2.3) is substantially lower than in rural areas (TFR=3.9). This lower fertility
in urban areas is apparent at all ages. Peak childbearing occurs between the ages of 20 and 34. Rural
women continue to bear children at later ages than urban women. The derived crude birth rate is only 22
births per 1 000 population.
Table 3.1 Current fertility
Age-specific and cumulative fertility rates and the crude birth rate
for the three years preceding the survey, by urban-non-urban
residence, South Africa 1998
_____________________________________________________
Residence
______________
Age group Urban Non-urban Total
_____________________________________________________
15-19 56 99 76
20-24 113 178 139
25-29 123 174 143
30-34 88 149 109
35-39 53 111 74
40-44 18 50 29
45-49 1 24 9
TFR women 15-49 2.25 3.92 2.90
TFR women 15-44 2.25 3.80 2.85
General fertility rate 79 133 100
Crude birth rate 19.2 25.4 21.9
_________________________________________
Note: Rates are for the period 1-36 months preceding the survey.
Rates for age group 45-49 may be slightly biased due to truncation.
Total fertility rate expressed per woman. General fertility rate
(births divided by number of women 15-49), expressed per 1,000
women. Crude birth rate expressed per 1,000 population.
33
3.3 Fertility Differentials
Differentials in fertility are shown in Table 3.2. The SADHS data indicate a strong negative linear
association between education and fertility. Whereas women with no education had a TFR of 4.5, those
who have completed Standard 10 have 2.2 children on average (see Figure 3.1). The Northern Province
has the highest TFR (3.9), while Free State has the lowest (2.2) (see Figure 3.2). The TFR for African
women is 3.1, for coloured women 2.5 and 1.9 for white women.
Unfortunately, despite attempts to design the sample so as to over sample Asian households, the sample
of Asian women was too small to allow a reliable estimate of the TFR. Table 3.2 also allows for the
assessment of differential trends in fertility over time. The mean number of children ever born to women
aged 40-49 is a measure of past fertility. By comparing current (total) fertility with past (completed)
fertility, it is clear that substantial declines in fertility occurred among all sub-groups of the population.
For instance, among urban African women there was a decline from 3.5 to 2.4.
Table 3.2 Fertility by background characteristics
Total fertility rate for the three years preceding the survey,
percentage currently pregnant and mean number of children ever
born to women age 40-49, by selected background characteristics,
South Africa 1998
_____________________________________________________
Mean
number
of children
Total Percentage ever born
Background fertility currently to women
characteristic rate1 pregnant age 40-49
_____________________________________________________
Residence
Urban 2.3 2.7 3.2
Non-urban 3.9 4.7 4.7
Province
Western Cape 2.3 3.2 3.0
Eastern Cape 3.5 3.4 4.0
Northern Cape 2.7 3.6 3.6
Free State 2.2 3.6 3.6
KwaZulu-Natal 3.3 4.3 4.0
North West 2.4 3.2 3.7
Gauteng 2.3 2.2 3.2
Mpumalanga 3.1 4.6 4.5
Northern 3.9 4.6 4.9
Education
No education 4.5 3.3 4.9
Sub A - Std 3 3.9 4.8 4.4
Std 4 - Std 5 3.5 4.0 4.1
Std 6 - Std 9 2.7 3.0 3.2
Std 10 2.2 3.5 2.4
Higher 1.9 3.8 2.5
Population group
African 3.1 3.7 4.1
Afr. urban 2.4 2.8 3.5
Afr. non-urban 4.0 4.7 4.9
Coloured 2.5 3.5 3.2
White 1.9 2.5 2.5
Total 2.9 3.5 3.7
_____________________________________________________
Note: The number of Asian women interviewed was too small to
provide a reliable measure of the total fertility rate.
1
Women age 15-49 years
34
Less than 4 percent of women reported they were pregnant at the time of the survey. Although this under-
estimates the proportion of pregnant women, as many women in the early stages of pregnancy would not
have known that they were pregnant, the differentials in pregnancy status closely follow the differentials
in current fertility.
35
3.4 Fertility Trends
Adequate historical demographic statistics are only available for certain sections of the population. For
the white population, birth statistics are available from 1910 and for the Indian and coloured population
from about the 1940s. For the African population incomplete birth registration statistics created a vacuum
in our knowledge of demographic trends. Fertility rates for the African population were calculated with
the aid of census statistics by Sadie (1970). In addition, the Human Sciences Research Council (HSRC)
conducted a number of small-scale surveys from the 1960s to the 1980s, the results of which were used
to estimate fertility levels (Mostert and Malherbe, 1974; Lötter and van Tonder, 1976; Van Tonder,
1985). Probably the most comprehensive source of information on fertility patterns before the 1998
SADHS was a large DHS-type survey conducted in 1987-1989 (Mostert, 1990). This collection of vital
statistics, analytical work and surveys provide reasonably accurate pointers to historical fertility trends
(see Figure 3.3) for different population groups.
The data in Figure 3.3 show that fertility among whites started to decline during the 19th century and
then remained at a level of between 3 and 4 for more than six decades before declining below 3 towards
the late 1960s. Thereafter the decline continued to below the replacement level. In 1990 the TFR was 1.7.
The fertility of the Asian segment of the population started its decline in the 1950s, largely due to
increasing age at marriage and the use of contraception. By 1990, the fertility of this population (TFR of
2.3) was nearing the replacement level. The fertility of the coloured population gradually increased in the
two decades before the 1960s. Thereafter, the introduction of modern contraception heralded a decline
which continues to the present. The fertility of the African population was at a high level during the
middle of this century (a TFR of nearly 7). After 1960, African fertility started to decline, slowly at first,
but subsequently gained momentum. By the mid-1980s, African fertility was about 4.5, and by 1990 it
had declined to an estimated level of around 4.0.
36
The fertility of the white population in South Africa mirrors fertility trends in the more developed world,
where the transition from high to low fertility has been completed. The Asian and the coloured population
have almost completed their fertility transition. The most interesting fertility transition taking place in
South Africa at present is that of the African population. Besides the fact that Africans constitute the
majority of the South African population, and therefore determine overall fertility levels, their fertility
behaviour may also predict the path of future fertility trends in other African countries. Therefore it is
useful to examine trends in African fertility by comparing the results of the 1998 SADHS with a large-
scale DHS-type survey conducted in the period 1987-1989 (Mostert 1990). In both surveys similar
methodologies were utilised. The age-specific fertility rates (ASFRs) for the three years preceding the
surveys are shown in Figure 3.4. Whereas the calculated total fertility rate for the African population was
4.6 in the first survey, it was 3.1 in 1998.
Source: DHS-type survey 1987-1989 and SADHS 1998
Fertility trends can also be examined by looking at rates reported for previous times in the birth histories
of women interviewed in the SADHS. Such rates were calculated for four-year periods prior to the survey
and are presented in Table 3.3. They also show a decline in fertility over the past 12-15 years, although
a much gradual one than implied by comparing external data sources.
Table 3.3 Trends in fertility
Age-specific fertility rates for four-year periods preceding the survey, South Africa
1998
__________________________________________________________________
Age group 0-3 4-7 8-11 12-15 16-19
__________________________________________________________________
15-19 78 93 90 116 104
20-24 136 156 170 196 190
25-29 138 159 162 183 183
30-34 108 134 128 153 [143]
35-39 72 89 99 [122] -
40-44 30 42 [57] - -
45-49 10 [12] - - -
__________________________________________________________________
Note: Age-specific fertility rates per 1,000 women. Estimates enclosed in brackets
are truncated.
37
3.5 Children Ever Born and Living
The distribution of women by the number of children ever born to them is shown in Table 3.4 for all
women and currently married women. In addition, the table gives the mean number of children ever born
to women in each five-year age group as well as the mean number of living children. It is clear that most
women have had at least one birth by age 35. Women in their late thirties have given birth to an average
of 3.2 children. The results also show that 5 percent of women in the age group 45-49 have never given
birth, while a quarter of women in this age group have given birth to 6 children or more. In the age group
45-49 the mean number of children ever born is 4.0; of whom 3.6 were still living.
Differences in the findings for currently married women compared to all women are mainly found among
the younger women. For example, nearly 80 percent of married women aged 20-24 have given birth,
compared to 60 percent of all women in the same age category. Nearly 4 percent of currently married
women aged 40 and older have never given birth, which can be regarded as a rough measure of primary
infertility in the population, since voluntary childnessness is uncommon in most African societies.
38
Table 3.4 Children ever born and living
Percent distribution of all women and currently married women by number of children ever born and mean number ever born and living children, by age of woman, South Africa
1998
Number of children ever born (CEB)
Number Mean Mean no.
Age of no. of of living
group 0 1 2 3 4 5 6 7 8 9 10+ Total women CEB children
ALL WOMEN
15-19 86.8 12.6 0.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 2,249 0.14 0.13
20-24 41.7 40.8 14.4 2.9 0.2 0.0 0.0 0.0 0.0 0.0 0.0 100.0 2,075 0.79 0.74
25-29 18.4 33.2 29.3 12.7 4.4 1.4 0.5 0.0 0.0 0.0 0.0 100.0 1,857 1.58 1.50
30-34 6.8 1 8.3 31.2 21.0 13.9 4.7 2.5 1.1 0.2 0.2 0.0 100.0 1,654 2.49 2.34
35-39 4.9 11.6 22.7 21.5 16.2 10.7 6.9 3.2 1.5 0.5 0.4 100.0 1,636 3.23 2.99
40-44 4.7 11.0 19.8 19.7 15.1 12.5 9.2 4.0 1.8 0.9 1.4 100.0 1,294 3.52 3.18
45-49 5.2 9.9 14.7 15.7 17.2 12.5 9.6 5.9 3.7 1.8 3.6 100.0 970 4.03 3.57
Total 29.5 21.1 18.2 12.0 8.0 4.8 3.2 1.5 0.7 0.3 0.5 100.0 11,735 1.94 1.79
CURRENTLY MARRIED WOMEN
15-19 63.4 33.0 2.1 1.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 73 0.42 0.40
20-24 20.4 43.4 27.6 7.9 0.7 0.0 0.0 0.0 0.0 0.0 0.0 100.0 465 1.25 1.17
25-29 11.6 25.6 34.9 18.2 6.2 2.5 1.0 0.0 0.1 0.0 0.1 100.0 900 1.94 1.83
30-34 3.7 12.5 31.3 23.2 17.6 5.8 3.8 1.7 0.3 0.3 0.0 100.0 1,008 2.83 2.66
35-39 3.5 8.0 22.7 21.8 17.7 11.9 7.5 3.9 2.0 0.5 0.5 100.0 1,114 3.47 3.25
40-44 3.6 8.8 19.3 19.7 15.7 13.2 10.0 4.5 2.2 1.1 1.8 100.0 865 3.74 3.37
45-49 4.1 6.8 14.5 14.5 18.3 14.5 9.8 6.4 4.3 2.2 4.6 100.0 652 4.34 3.89
Total 7.5 15.6 25.1 18.6 13.6 8.3 5.5 2.8 1.4 0.6 1.0 100.0 5,077 2.98 2.75
39
3.6 Birth Intervals
The birth interval is a notable factor in the survival of infants. A baby born soon after a previous child
is at an increased risk of poor health. As a general rule, births with an interval of less than 24 months are
classified as high-risk births. Table 3.5 presents the percent distribution of births in the five years before
the survey by interval since the previous birth, according to selected background characteristics.
Fourteen percent of second and higher order births take place within 24 months of the previous birth.
More than 20 percent of births to women with a tertiary qualification take place less than 24 months after
the previous birth, a higher proportion than found in any other educational category. Among white women
short intervals are even more common, as 30 percent of births take place within 24 months of the previous
birth.
Table 3.5 shows that 53 percent of second and third births occurred at an interval of more than 48 months.
This suggests that many South African women have an early first birth, many while still teenagers, but
postpone the birth of a second child for a considerable period of time.
The median birth interval was 47 months, a very long interval when compared to other sub-Saharan
countries. Urban women have substantially longer birth intervals than rural women (54 months as
opposed to 43 months). A significant difference in birth interval was also found when comparing the
survival status of the previous baby. Where the previous baby has died, the interval is 17 months less than
when the previous sibling survived.
40
Table 3.5 Birth intervals
Percent distribution of births in the five years preceding the survey by number of months since previous birth, according
to selected background characteristics, South Africa 1998.
Number of months since previous birth Median
Characteristic Total no. of Number
months of
since births
7 - 17 18 - 23 24 - 35 36 - 47 48+ previous
birth
Age of mother * 100.0 28.4 15
15-19 * * * *
20.7 38.6 100.0 40.7 1,312
20-29 7.3 11.0 22.4 1,586
17.5 16.8 55.1 100.0 53.0
30-39 4.0 6.6 60.5 416
4.7 15.3 15.6 58.8 100.0
40+ 5.7
Birth order 52.9 100.0 49.7 1,997
2-3 4.8 7.9 17.2 17.2
18.4 46.5 100.0 45.6 1,082
4-6 6.4 8.4 20.4 251
31.7 24.8 25.9 100.0 36.1
7+ 8.3 9.3
Sex of previous baby
Male 18.4 49.3 100.0 47.1 1,745
Female 5.9 8.3 18.2 1,585
20.6 17.9 48.3 100.0 47.0
Survival of 5.3 7.9
previous baby
Living 18.6 49.9 100.0 47.9 3,122
Dead 4.4 7.8 19.2 208
21.2 10.8 31.8 100.0 30.6
22.9 13.2
Residence
Urban 56.5 100.0 53.7 1,557
Rural 4.3 6.8 15.3 17.1
19.0 42.1 100.0 42.9 1,773
6.7 9.3 22.8
Province
Western Cape 61.0 100.0 58.5 251
Eastern Cape 4.1 6.6 11.3 16.9
17.6 39.8 100.0 40.3 519
Northern Cape 7.9 10.6 24.1 64
13.5 17.4 60.8 100.0 53.9
Free State 4.2 4.2 54.9 169
4.0 17.2 16.5 58.5 100.0
KwaZulu-Natal 3.8 100.0 45.0 749
7.6 9.0 19.7 18.8 45.0
North West 55.7 100.0 52.4 212
Gauteng 3.7 6.4 17.3 16.9
17.5 58.0 100.0 53.7 614
Mpumalanga 2.9 7.4 14.2 247
18.7 19.2 48.2 100.0 46.7
Northern 5.4 8.5 41.5 504
8.5 26.7 19.8 39.1 100.0
5.8
Education
No education 42.1 100.0 43.5 404
Sub A-Std 3 6.6 6.8 23.0 21.4
20.6 42.4 100.0 42.2 561
Std 4-Std 5 4.9 8.5 23.6 594
17.3 16.3 53.6 100.0 51.2
Std 6-Std 9 4.9 7.8 47.0 1,239
8.9 19.2 17.5 49.1 100.0
Std 10 5.3 100.0 58.3 372
5.0 6.5 14.6 14.0 59.9
Higher 42.4 100.0 41.3 159
11.1 9.2 14.7 22.6
Population group
African 48.8 100.0 47.1 2,795
African urban 5.3 8.0 20.1 17.9
15.6 59.9 100.0 55.9 1,130
African rural 3.0 5.9 15.6 1,665
23.2 19.4 41.2 100.0 42.4
Coloured 6.8 9.4 58.2 290
5.7 11.0 15.7 64.7 100.0
White 2.9 100.0 35.4 149
13.6 16.6 20.6 21.6 27.6
Asian 33.5 100.0 39.1 71
8.1 9.2 16.4 32.8
Total 48.8 100.0 47.1 3,330
5.6 8.1 19.3 18.2
Note: An asterisk indicates a figure is based on fewer than 25 respondents and has been suppressed.
3.7 Age at First Birth
Table 3.6 shows the percent distribution of women by the age at the birth of their first child and the
median age at first birth. The median age at first birth was approximately 21 years for most age cohorts.
Forty-eight percent of women in the age group 30-34 interviewed by the SADHS had given birth before
turning 20. However, there are indications that this trend is changing. For instance, 40 percent of
respondents aged 20-24 and 25-29 had given birth before reaching age 20, eight percentage points lower
than that of the older cohort of women. Teenage childbearing is discussed further in Chapter 9.
41
Table 3.6 Age at first birth
Percent distribution of women 15-49 years by age at first birth, according to current age, South Africa 1998
Women Age at first birth Number Median
Current with no Total of age at
age births <15 15-17 18-19 20-21 22-24 25+ women first
birth
15-19 86.8 0.6 8.8 3.8 NA NA NA 100.0 2,249 a
20-24 41.7 1.6 18.7 19.8 13.2 5.0 NA 100.0 2,075 a
25-29 18.4 2.9 16.9 20.6 17.8 15.6 7.8 100.0 1,857 20.9
30-34 6.8 3.6 21.0 23.4 16.6 14.3 14.3 100.0 1,654 20.2
35-39 4.9 2.8 19.0 20.6 18.7 19.1 14.9 100.0 1,636 20.7
40-44 4.7 3.1 15.2 21.2 20.9 17.6 17.2 100.0 1,294 21.0
45-49 5.2 2.7 15.3 23.2 16.8 19.3 17.4 100.0 970 21.0
Note: NA = Not applicable.
a = Half or more of women in these age groups had not given birth before entering the age group, making the
calculation of a median age at birth unfeasible.
Table 3.7 shows the median age at first birth for the age cohorts 25-29 to 45-49 by selected background
characteristics. No median age at first birth is provided for the age group 15-24 since a substantial
proportion of women in this age cohort had not yet given birth at the time of the survey. The results show
that median age at first birth is higher for women in urban areas than for women in rural areas. Similarly,
age at first birth increases with higher levels of education. For example, in the age cohort 30-34, women
without any formal education have their first birth around age 19, compared with age 26 for women with
secondary or higher education.
There is considerable regional variation in age at first birth. Age at first birth is lowest in Mpumalanga
(19.5) and highest in the Western Cape (21.8 years). Age at first birth also varies by population group.
Whites have the highest age at first birth (23.8), followed by Asians (22.3) and coloureds (21.2). African
women recorded the lowest age at first birth (20.3 years).
Table 3.7 Median age at first birth by background characteristics
Median age at first birth among women 25-49, by current age and selected background
characteristics, South Africa 1998
_______________________________________________________________________________
Current age
____________________________________________ Women
Background age
characteristic 25-29 30-34 35-39 40-44 45-49
_______________________________________________________________________________ 25-49
Residence
Urban 21.9 20.9 21.2 21.3 21.0 21.3
Non-urban 20.0 19.5 20.0 20.6 21.0 20.1
Province
Western Cape 23.4 21.4 21.9 21.2 21.9 21.8
Eastern Cape 20.8 20.5 20.9 21.5 21.9 21.1
Northern Cape 20.7 20.5 20.8 21.9 20.0 20.8
Free State 21.0 20.1 20.5 21.4 20.6 20.7
KwaZulu-Natal 20.6 20.5 20.7 21.1 22.2 20.9
North West 20.8 20.1 21.3 20.9 19.7 20.7
Gauteng 21.7 20.2 21.3 21.0 20.6 21.0
Mpumalanga 19.7 19.0 20.0 19.3 19.2 19.5
Northern 19.6 19.2 19.3 20.2 20.8 19.7
Education
No education 19.5 18.6 19.8 20.3 20.3 19.8
Sub A - Std 3 19.5 18.9 19.8 20.2 20.8 19.8
Std 4 - Std 5 19.0 19.3 19.9 20.2 20.0 19.7
Std 6 - Std 9 20.5 20.1 20.5 20.9 20.6 20.5
Std 10 23.2 22.1 22.6 23.2 24.2 22.7
Higher a 25.8 24.5 24.9 24.8 24.9
Population group
African 20.6 19.8 20.3 20.6 20.5 20.3
Afr. urban 21.2 20.1 20.6 20.8 19.9 20.6
Afr. non-urban 19.9 19.4 19.9 20.5 21.1 20.0
Coloured 22.0 21.0 21.5 20.8 20.5 21.2
White 24.4 24.7 24.0 23.5 22.8 23.8
Asian 24.2 21.7 20.9 22.2 22.6 22.3
Total 20.9 20.2 20.7 21.0 21.0 20.8
_______________________________________________________________________________
Note: The medians for cohorts 15-19 and 20-24 could not be determined because half of the women
had not had a birth before reaching the lowest age of the age group.
a = Omitted because less than 50 percent in the age group had given birth by age 20
42