«Suggested citation: Panday, S., Makiwane, M., Ranchod, C., & Letsoalo, T. (2009). Teenage pregnancy in South Africa - with a speciﬁc focus on ...»
Africa, the Caribbean and Asia have a different historical context and overall fertility including adolescent fertility rates is mostly higher than that in developed countries. Given a largely agrarian, rural society in Africa, in some cases children play a crucial role in sustaining the agrarian livelihood and therefore demographic and fertility transition lags behind. In addition, socio-economic status, particularly education, plays a crucial role in reducing fertility (Bongaarts, 2002). Analysis of the determinants of age speciﬁc fertility rates in 24 countries in Africa showed that declines in fertility have stalled in countries where there has been little educational and economic progress (Shapiro & Gebreselassie, 2007). However, countries that have increased the number of women with secondary or higher education have experienced declines in fertility (Shapiro & Gebreselassie, 2007).
0 30 60 90 120 150 Source: United Nations, 2008 National Fertility history in South Africa South Africa has the lowest Total Fertility Rate (TFR) in sub-Saharan Africa with rates comparable to middleincome countries in the developed world (Moultrie & Timeaus, 2003). The fertility decline in SA has run a long
course of almost half a century. Earlier work on fertility in SA documented from the late nineteenth century up to the early 1970s was mostly based on census data (see Gouws, 1987; Sadie, 1978; Simkins & van Heyningen, 1989). From the 1980s, direct estimates were derived from demographic and health surveys, conducted throughout SA. These surveys were modelled on World Fertility Surveys (WFS). Surveys were conducted separately for different administrative areas of SA, with slight variations in content and timing. Results of these surveys were mainly used to estimate fertility levels.
The fertility level of the White population recorded in the early thirties was 3.3. This ﬂuctuated over a narrow range until the early seventies, when it continued to decline to the present estimates of 1.7, considered below replacement level. Fertility levels of the Indian and Coloured population were recorded as 7.3 and 6.3 respectively in the period 1936-1941 (the Indian fertility level was the highest in SA during this period). Fertility declined much faster amongst the Indian population from 7.3 to the present estimate of 2.1 (Grobbelaar, 1984;
Mostert, 1991; Sadie, 1973).
A signiﬁcant decline in fertility among the Coloured and Black population began much later in the early sixties.
In the time period until recently, fertility declined more rapidly among the Coloured population group than among the Black group. Major questions remain about the fertility transition of the Black population. The timing of the onset of the fertility decline is unclear. It has been dated to the ﬁrst half of the 1960s without any strong empirical evidence. We know that the fertility data of the Black population which was available in the 1960s could not have been strong enough to permit a rigorous estimation of trends. New data has been produced in the last twenty years that improves our knowledge of recent trends in fertility at the national level. These include the data from series of surveys by the HSRC from 1987 to 1989, called the South African Demographic and Health Surveys (SADHS), and covered the formerly White SA and the homelands (Mostert, 1991), the Living Standard Survey of 1993 and the results of the annual October Household Surveys since 1993.
Estimates from the Living Standard Surveys (LSS) and the Central Statistics Services (CSS) paint a general picture of declining fertility in recent times. Based on the CSS ﬁgures for 1994, national fertility declined from
4.3 to 2.9 which represents a 32.6% change within a period of less than ten years (1982-1992). Most of the decline in fertility during this period occurred among the Black population.
Figure 2: Overall fertility trends in South Africa
Teenage fertility in relation to overall fertility In most traditional societies of the world, marriage is early and universal. In addition, marriage is associated with the onset of reproductive life. However, in SA, marriage is not universal and when it does happen it is usually delayed. The average age of marriage in SA is estimated at 29 years, and about half of all women in the reproductive ages do not marry (Kalule-Sabiti, Palamuleni, Makiwane & Amoateng, 2007). While early marriage is rare, this does not deter childbearing among adolescents. Data from the 2003 Status of the Youth Survey (Richter et al., 2005) showed that the median age of sexual debut is between ages 16 and 17. By age 20, half of young people have given birth to a child.
The relationship between marriage and fertility is tenuous. While fertility in SA is common outside marriage, non-marriage generally leads to higher fertility. This is reﬂected by the common occurrence of early and unplanned births to unmarried women, which is followed much later by a more controlled childbearing trajectory (Makiwane, 1998).
Generally, overall fertility trends in SA and most developing countries is uni-directional - from high to low fertility. While over time, teenage fertility in SA is also declining, albeit at a slower pace than overall fertility, evidence is that this decline is interrupted from time to time. Available evidence, shown below, indicates that teenage fertility is subject to ﬂuctuations that are uncommon to overall fertility. The two case studies illustrate this point clearly.
Case A: Trends in fertility in the Eastern Cape
The ﬁrst case is based on analysis of data collected in 1993 by the Department of Welfare and Pension in the Eastern Cape for a study titled ‘Fertility and family size preferences in the Eastern Cape: A study of the Transkei sub-region’. The study based its sampling frame on the 1991 census ﬁgures. A representative sample of 2 290 women between the ages of 15 and 49 was interviewed. The questionnaire included sections on household characteristics, demographic information and questions on reproductive health.
Fertility trends were analysed from the data set using Fertility Exposure Analysis (FEA) by Hobcraft and Little (1984). With FEA, estimates of fertility change can be derived from different surveys provided that the data are available and survey methodologies are sufﬁciently similar (see Pullum, Casterline & Shah, 1987). The FEA method enables one to track fertility trends from the birth histories of women interviewed.
From the FEA analysis, Table 1 below shows age speciﬁc fertility rates and total fertility rates in ﬁve year age groups for the period 1980-1994 for women classiﬁed by marital status.
In the late eighties, fertility declined among women in their thirties and forties and the reverse was the case for women below thirty, where a slight increase was recorded.
The study also shows a considerable reduction in total fertility in the ten year period. A more complex picture is evident for women in their teenage years. Generally, overall teenage fertility rose in the early eighties from 40 births per 1000 women to about 46 births per 1000 women and it declined again to about 42 births per 1000 women in the early nineties. A different picture is evident for women who never married. For the same period, fertility of teenage women who never married consistently increased from 22 to 24 and 39 births per 1000 women respectively. This occurred in spite of a massive overall decline in the fertility of never married women during the same period.
Case B: South African fertility trends, 1980-2007 Figure 3: Trends in teenage fertility, 1980-2007
Source: Moultrie & Timaeus, 2003; Statistics South Africa, 2008; Udjo, 2003 Figure 3 tracks national trends in teenage fertility for the period 1980-2001. The estimates for 1980, 1985 and 1990 were reported in Udjo (2003). These estimates were derived from 1970 census, birth registrations, and the 1995, 1997 and 1998 October Household Surveys. The estimates for 1996 and 2001 were derived from 1996 and 2001 census, discussed in Moultrie and Timaeus (2002). The estimates for 2007 were derived from the Community Survey (Statistics South Africa, 2008) The composite estimates for the period indicate that there was a decline in teenage fertility throughout the eighties, followed by a spike in the early nineties. This spike coincided with the political transition taking place in the country at the time. The above trends were veriﬁed below from a demographic surveillance site - Africa Centre for Health and Population Studies in rural KwaZulu-Natal. The ﬁgure below conﬁrms a spike in fertility in the early nineties, followed by a sustained decline in teenage fertility until 2005.
40 20 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Source: Moultrie & McGrath, 2007 Various recent data sources also indicate that teenage fertility in SA has declined over the past decade. Using Census data and the 1998 SADHS, Moultrie and McGrath (2007) demonstrated a 10% decrease in teenage fertility between 1996 (78 per 1000) and 2001 (65 per 1000). A further decline in teenage fertility (54 per 1000) is noted in the 2007 Community Survey (Statistics South Africa, 2008).
Table 2: Teenage fertility rates (per 1000 women) by population group, 1996-2001
Source: Moultrie & McGrath, 2007 The 2003 RHRU survey reported that 15% of teenage women aged 15-19 had ever been pregnant. This is comparable with the results of the 1998 SADHS that showed that 13.2% of women aged 15-19 years had given birth or 16.4% had ever been pregnant (DOH, MRC & Measure DHS, 2002). The 2003 SADHS reported that pregnancy amongst 15-19 year olds declined from 16.4% in 1998 to 12% by 2003 (DOH, MRC & OrcMacro, 2007). However, problems with ﬁeldwork in 2003, limit the quality and hence the reliability of the 2003 data.
Lessons from analysis of trends in fertility
Teenage fertility has been declining, however, at a slower pace than overall fertility. In addition, the two case studies show that teenage fertility is subject to ﬂuctuations generated by the socio-political environment, more so than overall fertility that follows the classic trends predicted by the demographic transition from high to low
fertility. As a result of these ﬂuctuations, the share of overall births attributable to young people has risen over time. The main lesson is that strategies that seek to reduce teenage fertility should be contextualized to the socio-political environment.
Differentials Data from the 1998 SADHS shows that fertility levels differ by age, location, educational level and population group (see Table 3).
Table 3: Adolescent pregnancy and motherhood, South Africa, 1998
Note: An asterisk indicates that a ﬁgure is based on fewer than 25 unweighted cases and has been suppressed.
Source: South African Demographic and Health Survey, 1998
Age National and international data show that fertility increases with age. In the US it is estimated that 822,000 pregnancies occurred in the 15–19 year age group in 2000; of these approximately two thirds were among 18– 19 year olds (The Alan Guttmacher Institute, 2004). Studies conducted in Africa also conﬁrm this relationship.
For example, teenage pregnancy rates in Kenya double from 17% at ages 15-16 to 34% at ages 17-18 (Were, 2007).
As shown in Table 2, fertility in SA rises from a low of 2% among 15 year olds and peaks at 30.2% among 19 year olds (DOH, MRC & Measure DHS, 2002). Data from the 2003 RHRU survey also shows that teenagers aged 17-19 account for 93% of teenage fertility (see ﬁgure 5 below) (Harrison, 2008b).
Figure 5: Pregnancy among 15-19 year olds by age, 2003
Source: Harrison, 2008b This ﬁnding is in keeping with the proportional increase in sexual activity with age. While only 10.1% of adolescents aged 15 are sexually active, by 19 years of age, 60.6% are sexually active (Shisana et al., 2005).
Harrison (2008b) explains that the chance of falling pregnant increases with age. At the age of 15 the ratio of the number of girls reporting sexual activity to the number who fall pregnant is 13:1. This decreases to 7:1 at age 16 and to 3:1 from 17 years of age onwards. This may be related to increased levels of unprotected sex at older ages Lower levels of induced and spontaneous abortion among older adolescents may also contribute to higher pregnancy rates in this group (Darroch et al., 2001, Dryburgh, 2003). For example, data from France, Great Britain and Sweden show that the abortion ratio is substantially higher among adolescents aged 15-17 than that among those aged 18-19 (Darroch et al., 2001). Scholars have also attributed the higher pregnancy rates among older adolescents’ to biological and social changes. Biological changes such as physical maturity and higher hormone levels may result in greater sexual attractiveness and desire. Social changes include greater peer pressure to have sex, changes in perceived norms about sexual and contraceptive behaviours as well as greater freedom and independence that brings with it more sexual opportunities (Kirby, 2002).
Location Where young people reside confers signiﬁcant risk for early pregnancy, STIS and HIV through a number of structural mechanisms – access to health services, concentration of poverty and unemployment and poor educational attainment. The contribution of these factors to early pregnancy is discussed in greater detail under the structural determinants of pregnancy.
Province Free State, Gauteng and North West provinces had lower proportions of teenage fertility while Mpumalanga, Northern Cape, Limpopo and Eastern Cape reported high levels of early pregnancy (see Figure 6 below and Table 2 above).
Figure 6: Percentage of women aged 15-19 who are mothers per province, 1998