Economic and Political Issues of Fertility Transition
in the Arab World—Answers and Open Questions1
Youssef Courbage
Institut National D'Etudes Demographiques
INTRODUCTION
Fertility in the Arab world2 has now (19953) fallen below 4 children
per woman. Among women educated to secondary level in some countries
(Morocco, Tunisia, Lebanon)—a significant population segment—it has
fallen below replacement level.4 Arab fertility is now lower than in the
three regions of subSaharan Africa—Eastern, Central and Western (6.40 in
all three)—but still higher than in other regions of the South, the Far East
(2.80), SouthEast Asia (3.70), the Indian subcontinent (3.70) and Latin
America (2.90). Nonetheless, it is an undeniable trend, for the region has,
rightly or wrongly, a reputation for opposition to change—here in the
shape of an unbridled birthrate. Up to the 1970s Arab demographics seemed
isolationist, a bastion of natural fertilities and resistance to family change.
The paradigm of demographic transition—a falling birthrate accompanied
by educational, economic and social transformations—seemed to lose all
credibility in the Southern Mediterranean region. Significantly declining in
fant mortality had but an imperceptible effect on fertility. Apart from tiny
Lebanon, where small parity had long been common among some sections
of the population, high fertility was the norm everywhere.
The 1960s present an unremarkable picture. What was to prove a fun
damental divide between the Maghreb and Middle East, on either side of
Libya, had not yet emerged, giving some credence to the generalized cul
Please address correspondence to Y. Courbage, INED, 133 Bd Davout, 75020 Paris, 14
France; email: courbage@ined.fr.
Population and Environment: A Journal of Interdisciplinary Studies
Volume 20, Number 4, March 1999
© 1999 Human Sciences Press, Inc. 353
,354
POPULATION AND ENVIRONMENT
turalist inferences about religion—in the present instance the Muslim reli
gion (97% of Arabs)—and fertility. D. Kirk5 concluded briefly and axioma
tically: (1) Islamic fertility is universally high (2) it displays no significant
downwards trend (3) it is higher than that of neighbouring populations of
other religious beliefs. It was a perception which left a lasting imprint on
analyses of the region's demography, for even as late as 1987 "iconoclas
tic" experts in world transitions, j. Cleland and Ch. Wilson, were still pro
pounding a simplified vision of regional realities.6 At the same time, how
ever, Ph. Fargues disproved the received wisdom by showing that Arab
fertility had declined. But there were doubts about its longterm sustain
ability: "On the threshold of the second stage of their demographic transi
tion, Arab population seem to be hesitant."7 Ten years on, "hesitant" may
not be the right word. The odd blips in the decline are merely the trees
hiding the forest of fundamental change. When the question was revisited8
four years ago (April 1993), Arab fertility still stood at 5 children. Today, a
combination of real changes and improved measuring techniques has
brought the figure down to 4, betokening the substantial progress made.
This is no place to add yet a further review of this fertility transition to
other recent ones.9 I shall not consider the pattern of its progression in
detail but simply single out a few determinants, the critical phases, before
going on to rank the reasons: some are irrefutable, others less clearcut, less
quantitative, perhaps more promising. This is the path of discovery which
must be more deeply explored.
SOURCES: UNSUITED TO STUDYING A VOLATILE PROCESS
Ideally, annual registered births would have been tallied with census
returns. In fact, that cannot be done. While all Arab countries do have
registration records, they are used only for death and birth certificates for
administrative purposes. Recent registration records available for public
use exist only in a small minority of countries: Tunisia, Algeria, Egypt,
Kuwait.10 Registration records and census returns are in competition with
internationally funded and supervised surveys. But these are of uncertain
reliability, and registration records offers a more reliable picture of varia
tions over time and locality. In Egypt, it was shown that demographic time
series and economic series (average earnings, emigrant remittances, build
ing permits) improve analysis by factoring back in the economic dimen
sion." In Palestine, a traditional use—by birth histories—smooths out fer
tility oscillations and masks the effect of the Intifada." The usefulness of
census returns is limited by the failure to publish data by individual year of
, 355
YOUSSEF COURBAGE
age, which would allow annual fertility to be reconstructed. This relative
paucity of statistical data leads to the use and abuse of surveys, despite
their patent defects: inaccurate inquiry, limited representativeness, aggrega
tion of multiple (up to 10) years to obtain relevant indicators. Preconcep
tions about the nature of fertility transition result in stereotyped and infi
nitely reproducible questionnaires. Reliance on what is quantifiable results
in less easily identifiable factors being sidelined. The United Nations pub
lishes biannual backcasted series to accompany their population projec
tions. These are not "sources" in the strict sense. They are often approxi
mate, nominal reconstructions updated with a certain lag; by that token,
they mask the real oscillatory movements.
TRANSITION IS NOT AN UNINTERRUPTED PROCESS
Fertility survey birth histories and the UN series often suggest that tran
sition is a smooth process. In fact it is not. The transition from 7 children in
the 1960s to 4 today has not been a linear process. In Morocco, for exam
ple, fertility did not decline between 1962 and 1972 from 7.15 to 6.89
(UN smoothing), but actually increased to 7.40, while Egypt's first fertility
transition between 1965 and 1970 was followed by a bidecennial stagna
tion with annual fluctuations. Between 1970 and 1980, indeed, fertility
increased significantly from 5.30 to 6.15, fell, then rose again to 6.31 in
1987. Even in Tunisia, regarded as a textbook case, fertility decline stalled
briefly in the early 1990s, suggesting that 3 children was the basic irreduc
ible minimum. Syria's birthrate fluctuated by approximately 6% between
1975 and 1985 before a confirmed decline set in.
The current picture is one of striking contrasts. Fertility was almost
identical in all countries in the pretransitional stage. The fractional coun
trytocountry rate variations (Lebanon excepted) were too slight to reflect
real differences. Today, the differences exceed the similarities. Table 1
shows the fertility differentials ranging from 2.31 in Lebanon to 7.13 in
Yemen, and the percentages of fertility transition remaining to be achieved
if fertility transition halts at the replacement level of 2.1. It is eight tenths
complete in the Maghreb, twothirds in the Nile Valley, just over half in the
Middle East and only onefifth in the Arab Peninsula. Table 2 shows the
kaleidoscope of current situations compared to the homogeneous picture
of the 1960s. Extreme and unvarying concentration has given way to a
patchwork: virtually all the boxes are filled from Lebanon to Yemen.
The poles of fertility around a high and relatively unvarying average
lied in the 1960s with Lebanon at one end and Algeria at the other.
