zyxwvut
zyxwvut
zyxwv
Journal of International Development: Vol. 7, No. I , 25-45 (1995)
THE START OF THE SUB-SAHARAN
FERTILITY TRANSITION: SOME
ANSWERS AND MANY QUESTIONS
R. LESTHAEGHE
Interface Demography, Vrije Universiteit, Brussels
and
C . JOLLY
Population Committee, National Acudemy of Sciences, Washington DC
Abstract: Factors associated with socioeconomic development such as increased edu-
zyxwv
cation and urbanization were initially associated in sub-Saharan Africa with an
increase in fertility. This was largely due to their eroding effects on durations of
breastfeeding and postpartum abstinence, and in some regions also to lowered levels
of infecundity. During the 1980’s the second phase of the transition emerged in areas
located in Eastern and Southern Africa. This phase is characterized by increasing
proportions of users of modern contraception and by a decrease in desired family size.
Such a decisive turn around did, however, not occur in situations with expanding
economies, but rather during the crises of the last decade. The theory of the “crisis-lcd
fertility transition’ can help to explain this outcome, but it also has its limitations. If
the crisis deepens and hits the health and education sectors, negative effects emerge as
crucial props such as female schooling levels and family planning provisions may suffer
severe set-backs.
1 INTRODUCTION
From at least the 1950s onward population growth rates in sub-Saharan Africa have
steadily increased. Around 1950, t h e rate was about 2.1 per cent p.a.; it rose t o 2.7
per cent in 1970, and t h e estimate for 1990 is 3.2 per cent. A s shown in Table 1,
zyxwvutsrq
none of t h e 39 continental sub-Saharan nations h a d a growth rate in 1950 in excess
of 3.0 p e r cent, whereas there were nine such countries in 1970, a n d n o less than 23
in 1990. During the 1980s, Kenya and Rwanda reached t h e 4.0 per cent level.
These growth rates were the result of the combined effects of t h e following
factors:
zyxwvut
1. A steady decline in mortality: life expectancy a t birth rose for sub-Saharan Africa
zyxwvuts
as a whole from 36 years in 1950 to 43 years in 1970 a n d 51 years in 1990. In
Kenya, Botswana a n d Zimbabwe, the present life expectancy is close to 60
The author’s institutions are not responsible for the content of this paper.
CCC 0954- 1748/95/010025-21
@1994 by John Wiley & Sons, Ltd.
,26 zyxwvutsrq
zyxwvutsrqpo
zyxwvuts
R. Lesthaeghe and C. Jolly
Table 1. Evolution of population growth rates (r) (per
thousand population) in 39 sub-Saharan countries, 1950-90.
zyxwvutsrqp
1950 1970 1990
zyxwvutsrqp
Growth rate ( r )
zyxwvut
35% 0 1 8
30-34.9%, 0 8 15
25-29 .9%9 6 12 14
20-24.9”/, 16 14 2
15-19.9%o 10 3 0
SlS%O 7 1 0
Total 39 39 39
Level of 10th country 23.6%~ 29.6 33.3
20th country (median) 20.6 25.8 31.1
30th country 16.2 22.5 28.2
Source: Computed from Willems and Tabutin (1992) (island populations
excluded).
Table 2 . Evolution of period total fertility rates (TFRs) in 39
sub-Saharan countries, 1950-90.
1950 1970 1990
TFR 38.0 0 2 1
7.0-7 .9 6 8 9
6.0-6.9 23 24 22
5.0-5.9 9 4 7
< 5.0 1 1 0
39 39 39
Level of 10th country 6.85 7.03 7.00
20th country (median) 6.59 6.50 6.55
30th country 5.98 6.24 6.29
Source: Computed from Willems and Tabutin (1902) (island populations
excluded).
zyxwvutsrqp
years. Infant mortality rates declined for the continent from about 190 per
thousand live births in 1950 to about 105 in 1990, with levels below 70 in
Botswana, Kenya, Zimbabwe and the Cape Verde Islands. Mortality rates
below age 5 in these countries are also below 100 per thousand.
2. A sustained increase in fertility: as shown in Table 2, only 6 of the 39 countries
considered had a period total fertility rate (TFR) above 7.0 children in 1950, but
’
this number rose to 10 in 1970 and 1990. Kenya and Rwanda exceeded the 8.0
children level in 1970 and 1980, and in Rwanda the figure for 1990 is still 8.1.
’ The period total fertility rate (TFR) is thc sum of thc prcvailing age-specific fertility rates between the
ages of 15 to 50. The TFR indicates the sizc of the offspring of a fictitious cohort at agc 50 given the
fcrtility of a particular year or period. I t is distinct from the cohort total fertility ratc, which rncasures the
number of offspring at agc 50 in a rcal gcneration of women.
, zyxw The Start of the Sub-Saharan Fertility Transition 27
zyxw
zyxwvutsrq
Simultaneously, several low-fertility countries with TFR values below 6.0 in
1950 joined the middle group (TFR between 6.0 and 6.9), largely because of a
reduction in sterility levels. This applies to Cameroon, with a rise from 5.7 in
1950 to 6.9 in 1990, the Central African Republic (from 5.5 to 6.2), Gabon
(from 4.1 to 5.2), Guinea Bissau (from 5.1 to 5.8) and Equatorial Guinea (from
5.5 to 5.9).
3 . A rejuvenation of the age structure: the median age fell in East Africa from 18.1
years to 16.4 between 1950 and 1990, and a similar trend was witnessed for
Central Africa with a decline from 19.4 to 17.2, and Western Africa from 17.9 to
16.6 (UN, 1992). The combination of rising life expectancy and such a very
young age composition produced a rapid decline in the crude death rates: the
figure for sub-Saharan Africa fell from 28 per thousand in 1950 to 22 per
thousand in 1970 and to 16 per thousand in 1990. Crude birth rates remained
almost stable in view of the increase in TFR and a small reduction in the
proportion of women of reproductive age. In fact, the birth rate for sub-Saharan
Africa was 49 per thousand in 1950 and it is estimated to be about 47 per
thousand in 1990.
This illustrates that the continent has to reduce fertility to restore the balance in
the long term. The prospects for a swift decline in fertility were, however, not
propitious until the 1980s. According to the threshold hypothesis of the World
Bank (1986), African nations had simply not reached the levels of socioeconomic
development beyond which a sustained fertility decline could start, whereas other
authors (e.g. Caldwell and Caldwell 1987, 1990) argued that cultural forces also
militated against sub-Saharan Africa following the Asian example. Still others (e.g.
Lesthaeghe, 1989) argued that many Asian countries had a strong family planning
programme as a component of nation-building, whereas sub-Saharan countries
were late in formulating such intentions and did not have the means of organizational
frameworks to implement them. Yet, from the latest rounds of demographic
surveys, it became apparent that several regions had declining fertility and that this
had been achieved as a result of increased use of modern methods of contraception.
Hence, the research agenda has changed quite suddenly in the wake of these
findings: it appeared that the socioeconomic, culture and organizational barriers
could, at least in a few settings, be overcome.
A more complete account of the evolution of sub-Saharan fertility will be offered
in the next sections. First, we shall document the so-called ‘two-stage transition’
and evaluate the impact of socioeconomic development factors on the countervailing
forces of shrinking traditional birth spacing versus increasing contraceptive use.
Second, we shall assess the role of female education in these processes. Third, the
issue of the ‘quality-quantity trade-off‘ will be discussed in the light of Boserup’s
(1985) script of a ‘crisis-led’ transition and, finally, some major caveats will be
introduced that deal with the reduced investment in key sectors such as health and
schooling.
2 THE TWO-PHASE TRANSITION
It is well known that sub-Saharan African populations had a very strong tradition of
child spacing based on prolonged lactation and concomitant post-partum absti-
