Embed Size (px)
Citation preview
/. biosoc. Sci. (1987) 19, 405-425
CONTINUING FERTILITY TRANSITIONS IN APLURAL SOCIETY: ETHNIC TRENDS AND
DIFFERENTIALS IN PENINSULAR MALAYSIA
LIN LEAN LIM,* GAVIN W. JONES AND CHARLES HIRSCHMAN*
Department of Demography, Australian National University, Canberra
Summary. Fertility in Peninsular Malaysia has declined continuously from thelate 1950s, reaching a TFR of 3735 in 1983. All ethnic groups in Malaysia havecontributed to this modern demographic transition but the rate of change hasbeen most rapid for Chinese and Indians, Malay fertility having reached aplateau in the early 1980s. The effect of age structure, marital patterns andmarital fertility (by parity) on the fertility declines for each ethnic communityare analysed. Continuation of current trends would lead to replacement-levelfertility for Malaysian Chinese and Indians by the year 2000. Malay fertility islikely to continue to decline but at a more moderate pace.
Introduction
Just over two decades ago, the first signs of fertility decline in Asia were reported forTaiwan, Korea, Hong Kong, and Singapore. Since then, almost every country in eastand south-east Asia has started on the path from high to low fertility. The pace ofdecline varies considerably from country to country, and there appear to be multipleexplanations for it (Freedman, 1979; Jones, 1984). As one of the most rapidlydeveloping countries in Asia, Malaysia exemplifies the contemporary pattern of Asiandemographic transitions. This paper charts the continuing fertility transitions inPeninsular Malaysia, both for the total population and for each of the three majorethnic groups, Malays, Chinese, and Indians.
In the years following independence in 1957, Peninsular Malaysia (formerlyMalaya) experienced extraordinarily rapid socioeconomic development. Per caputGDP rose by an average of 6-5% per annum from 1960 to 1970 and by 7-7% perannum from 1970 to 1983, and stood at US $1860 in 1982. Even in the relativelysuccessful group of ASEAN economies (Malaysia, Thailand, Indonesia, Singapore,and the Philippines), Malaysia’s record is exceptional, second only to that ofSingapore. Of equal importance was progress in social development, the spread ofmodern medical services, and major improvements in transportation andcommunications throughout the country. For example, about three-quarters ofwomen aged 15-19 in 1980 went to secondary school, whereas in the previous
Present address: see list of contributors, p. ii.
405
406 L. L. Lim, G. W. Jones and C. Hirschman
generation (those aged 35-39 in 1980) only 15% were able to do so (Malaysia
Department of Statistics, 1983a). The infant mortality rate which stood at 102 in 1947
decreased to 21 in 1982. (Hirschman, 1980; Malaysia Department of Statistics,
1983b). These changes had major consequences on marriage patterns, women’s roles,
and the desirability of large families.Each of the three peoples of Malaysia shared in these social changes and modified
their fertility patterns, though not to the same degree. For the Malay populations, the
most rural of the three communities and the group largely bypassed by the economic
developments of the colonial era, the post-independence changes were the most
dramatic. The government, whose political base was largely in rural Malay areas,
fostered policies to encourage faster educational and economic advance in the Malay
community. Although it is too soon to say how far Malays ’caught up’, they certainly
made major educational strides (Hirschman, 1979; Sidhu & Jones, 1981). Chinese and
Indians, who form almost half of the population of Peninsular Malaysia, with
traditionally a stronger hold in urban areas and non-agricultural employment, also
benefited from the expanding economy. However, the Indian community seems to have
experienced recent deterioration in infant mortality, unemployment levels, and in the
proportion of places in institutions of higher education (Kusnic & DaVanzo, 1982).This study of ethnic variations in fertility levels and trends updates earlier analyses
of trends in Malaysia with recent data from the 1980 census of population (Saw,1967a; Cho, Palmore & Saunders, 1968; Palmore, Chander & Fernandez, 1975;
Hirschman & Fernandez, 1980; Sidhu & Jones, 1981, Jones & Tan, 1985).
Material
The data came from the vital registration system and population censuses of Malaysia.The vital registration system provided information on the number of births in each
year, regularly published by the Malaysian Department of Statistics and (prior to
1963) in the Reports of the Registrar-General on Births, Deaths, Marriages, Divorces,
and Adoptions. The registration of births may have been about 90% complete in the
1950s (Saw, 1964) and virtually complete after the late 1960s (Cho et al., 1968). To the
extent that coverage of birth registration improved over time, the fertility indices
underestimate the magnitude of the fertility decline. The denominators for the
fertility measures are based upon census counts and a consistent series of intercensal
estimates (Hirschman, 1974). The census populations of 1947, 1957, 1970, and 1980
were adjusted for underenumeration (1970 and 1980) and other inconsistencies to
make them as comparable as possible.
Results
Figure 1 presents, for the 30 years from 1947 to 1983, the crude birth rate, the generalfertility rate and the total fertility rate (only available since 1958). The crude birth rate
influenced by age-sex composition as well as by reproductive behaviour was
depressed by the relatively unbalanced age-sex structure of the late 1940s (whichfollowed a period of heavy immigration and the dislocation of the Japaneseoccupation), especially for Chinese and Indians, and its levelling off in the 1970s was
Fertility differentials in Malaysia 407
1945 1955 1965 1975 1985Year
Fig. 1. Trends in indices of fertility. Peninsular Malaysia, 1947-83.
due, in large part, to the rapid increase in the proportion of the most fertile groups.The general and total fertility rates provide more refined measures of underlyingfertility behaviour. There were two distinctly different trends of fertility for the postWorld War II era. From the late 1940s to the mid-1950s, both the crude birth rate andthe general fertility rate rose, although there were wide year to year fluctuations.Then from the late 1950s to the present, there was continuing decline. Whereas thecrude birth rate decline was checked in the 1970s, the general and total fertility ratedeclines showed only a slight pause around 1980. One interpretation of the decade ofpersistently high, perhaps rising, fertility in the late 1940s and early 1950s (Saw,1967b) is that improvements in health and rising levels of income due to the rubberboom of the Korean War period allowed many families to attain high fertility. Exceptfor the massive resettlement of rural people, largely Chinese, into ’new villages’(nucleated settlements, often adjacent to existing towns) during the Malayanemergency (Sandhu, 1964) there was little structural change in the economy or societythat would have motivated families to have few children. However, by the late 1950s,it appears that conditions were right for the beginnings of a downward trend in fertilitylevels.
The three fertility rates for selected years from 1947 to 1983 and changes in them(Tables 1 and 2) for the total population of Peninsular Malaysia and separately forMalays, Chinese, and Indians show that fertility assessed by the crude birth rates washigh, more than 40 births per thousand population, for all three ethnic groups in thelate 1940s. The general and total fertility rates indicate that fertility was considerablyhigher for Chinese and especially for Indians than for Malays. Part of the reason forthe lower Malay fertility may have been higher levels of marital disruption, or lesscomplete registration of Malay births (Smith, 1952; Hirschman & Fernandez, 1980).
By the late 1950s and early 1960s, fertility began to decline for all three ethnic
408 L. L. Urn, G. W. Jones and C. Hirschman
Table 1. Trends in fertility rates by ethnicity for selected years
Rate and year
Crude birth rate
19471958196319701975197819801983
General fertility rate
19471958196319701975197819801983
Total fertility rate
1958196319701975197819801983
Total
43-142-739-232-530-629-630-329-9
188196184146129121122117
6187593248864213388138853735
Malay
41-645-541-434.233-231-633-634-8
169196186152141130136139
5882585150904650428344754538
Chinese
44-338-635-730-526-926-425-222-5
20818517213611310710086
6546591746243638337231342721
Indian
49-145-340-730-829-929-129-827-6
236
239222147127118116
105
721669424798393134813375
3013
communities. The total fertility rates show that the declines were larger and more
consistent for the Chinese and Indian population. In the early 1960s, Chinese fertility
dropped below the Malay level, and Indian fertility did the same by the late 1960s.
The pace of fertility decline of the Chinese and Indian populations was phenomenal.Annual decreases of 3-4% a year were common, and by 1980, Chinese and Indian
total fertility rates were less than half of their 1958 levels. Rapid declines continued
for both groups in the early 1980s, with the Indian rate falling to an average of three
children per woman in 1983, and for the Chinese even lower. These remarkable
figures are comparable to fertility levels in industrial countries around 1960.
By comparison, the Malay fertility decline appears much slower. In a broader
historical context, however, the Malay trend is quite appreciable with a decline from
almost six children per woman in 1958 to about four and one-half in 1983. Althoughthe Malay crude birth rate shows no change from 1970 to the present this is, in part, a
product of changes in age structure. But from 1978 to 1983, there appears to have
been a real increase in Malay fertility (1978 TFR=4283, 1983 TFR=4538).
Fertility differentials in Malaysia
Table 2. Annual changes in fertility rates by ethnicity
Average annual % change
409
Rate and period
Crude birth rate
1947-581958-631963-701970-751975-781978-801980-831970-801978-83
General fertility rate
1947-581958-631963-701970-751975-781978-801980-831970-801978-83
Total fertility rate
1958-631963-701970-751975-781978-801980-831970-801978-83
Total
-0-1-1-7-2-6-1-2-1-1
1-2-0-4-0-70-2
0-4-1-3-3-3-2-4-2-10-4
-1-4-1-8-0-7
-0-8-2-7-2.9-2-7
0-1-1-3-2-3
-0-8
Malay
0-8-1.9-2-7-0-6-1-6
3-11-2
-0-21-9
1-4-1-0-2-8-1-5-2-72-30-7
-1-11-3
-0-1-2-0-1-8-2-72-20-5
-1-31-2
Chinese
-1-2-1-5-2-2-2-5-0-6-2-3-3-7-1.9-3-1
-1-1-1.4
-3-3-3-6-1-8-3-3-4.9-3-0-4-3
-2-0-3.5-4.7-2-5-3-6-4.6-3-8-4.2
Indian
-0-7-2.1-3-9-0-6-0-9
1-2-2-5-0-3-1-1
0-1-1-5-5-7-2.9-2-4-0-9-3-3-2-3-2-3
-0-8-5-1-3-9-4-0-1-5-3-7-3-5-2-8
The challenge of age structure
As noted by Sidhu & Jones (1981) the slackening of immigration of Chinese andIndians in the 1930s and 1940s and changes in composition of immigrants set the scenefor a faster natural increase of the population after World War II. The changingcharacter of the migration flows of the 1930s-fewer males and a substantial influx offemales-went far to balance the sex ratio in the Chinese and Indian communities,conducive to higher reproduction levels. Thus by 1947 the reproductive potential ofthe Chinese and Indian populations was higher than it had ever been in the past,although it was still not quite as high as for the Malays, who had a slightly largerproportion of reproductive-age females (Table 3). High fertility rates prevailingduring the late 1940s and 1950s gave a broad base to the Peninsular Malaysian age
410 L. L. Lyn, G. W. Jones and C. Hirschman
Table 3. Age structure of women by ethnicity
Age groups(years)
All women15-2425-3435-49
15^t9
No. of womenf
Malay women15-2425-3435-49
15-49
No. ofwomenf
Chinese women15-2425-3435-49
15-49
No. of woment
Indian women15-2425-3435-49
15-49
No. of women+Both sexes.
t 1000s.
1947
8-37-17-5
22-9
1120
9-18-07-5
24.6
595
7-65-97-8
21-3
399
7-17-364
20-8
111
% of total
1958
8-8646-6
21-8
1439
9-37-26-8
23-3
.762
8-75-66-5
20-8
513
7-55-66-0
19-0
141
populati
1970
9.96-16-3
22-3
2041
9-86-26-6
22-5
1086
10-16-35-9
224
733
9.95-35-7
20-9
204
ion*
1980
10-87-56-7
25-0
2856
11-07-264
24.7
1562
10-37-87-1
25-2
975
11-77-86-1
25-6
300
1983
10-77-96-9
25-5
3148
10-87-76-6
25-1
1733
1048-17-6
26-0
1066
11-18-664
26-2
328
pyramid, and so were responsible for the large numbers of births during the 1970s and
1980s when these cohorts reached the peak reproductive ages.The rise in the proportion of women of reproductive age in the 1970s (Table 3) can
therefore be traced back to these earlier developments. Actually, Table 3 shows that
during the entire period of declining crude birth rate in Peninsular Malaysia (from1958 onwards), women of reproductive age have constituted an increasing proportionto the total population, implying an increase in birth rates if fertility rates had
remained unchanged. Thus the decline in birth rates was achieved in spite of, rather
than with the aid of, trends in age structure. The trends have differed by ethnic group.For both Chinese and Indians the proportion of reproductive-age women in the total
Fertility differentials in Malaysia 411
population rose steadily and substantially-by 25% and 38%, respectively, between1958 and 1983-thus providing quite an obstacle to be cleared before declines in birthrates were possible. The fact that such declines occurred testifies to the strength of thefertility decline. For Malays, the trends in age structure were rather different.Between 1958 and 1970, a slight fall in the proportion of reproductive-age womensupported the crude birth rate decline, and the subsequent increase (by 12% between1970 and 1983) was not as great as that of either Chinese or Indians.
Along with the ’normalization’ of the age structure in the late 1940s and its effectsin producing a broad-based age pyramid, the sharp decline in fertility among theChinese and Indians after 1958 was a major reason for the rise in the proportion offemales aged 15-49 in the population, because it led to an undercutting at the base ofthe age pyramid. This is a normal characteristic of populations undergoing fertilitydecline, and means that crude birth rates seldom reflect the strength of the fertilitydecline, which is better portrayed by trends in more refined measures of fertility. Eventhe general fertility rate does not always fully portray the strength of the fertilitydecline, if there is a tendency for reproductive-age women to become more heavilyconcentrated in the ages of peak reproduction, but in the present data both it and thetotal fertility rate showed similar trends over the period.
For each ethnic group, there has been a tendency for the age distribution withinthe group of reproductive-age women to grow younger, reflecting the entry into theyounger reproductive ages of the large birth cohorts of the 1950s and early 1960s. Theeffect of this on crude birth rates is hard to determine, because rising age at marriageand increasing use of contraception meant that fertility was increasingly concentratedin the more central reproductive ages.
By the 1990s, the earlier declines in fertility will bring about a decline in theproportion of the total population made up of females in the main reproductive ages.After that point, further declines in fertility will be reflected in a sharper decline in thecrude birth rate and hence the rate of population increase.
Changes in marital patterns
Changes in marriage have played a major role in the fertility decline observed inPeninsular Malaysia in recent decades. Since births outside wedlock have traditionallybeen rare for all ethnic groups, age at marriage and proportions married have haddirect implications for ^fertility. Looking back three and four decades, familystructure, patterns of marriage and marital distribution differed substantially for thedifferent ethnic groups. Arranged marriages were the norm for all three groups, butfor the Chinese they took place at older ages than for the others. Near-universalmarriage characterized the Malays and Indians, and only in the Chinese was there anon-trivial proportion of women who did not marry before the end of theirreproductive years (around 3%). The age gap between spouses was wide, especiallyfor the Indians and to a lesser extent the Malays. In 1957, the difference in median ageat first marriage between men and women was 7-2 years for Indians, 5-0 for Malaysand 3-7 for Chinese. For the Indians, this was partly due to the shortage of potentialspouses in a traditionally male-dominated immigrant community. But the Indian andMalay communities had a tradition of wide age differences between spouses. TheChinese, despite a relative shortage of females, maintained a smaller gap in age of
412 L. L. Lim, G. W. Jones and C. Hirschman
spouses, and a larger proportion of the Chinese men simply remained unmarried
(Jones, 1980).Fertility is affected not only by age at marriage and proportions marrying, but also
by marital dissolution and the extent of remarriage. Here, too, the situation differed
considerably between the three ethnic groups. Divorce was much more common
among the Malays than among the other groups-indeed, from the early post-war
years (and probably earlier) until the mid-1970s, divorce rates among Malays in
Kelantan and Trengganu appear to have been the highest in the world (Jones, 1981).The effect of marital disruption on fertility was tempered by the frequency and
rapidity of remarriage; even so, at the time of the 1947 census, about 8% of
reproductive-age Malay females were currently divorced, compared with below 1% of
Chinese and Indians (Sidhu & Jones, 1981). In terms of reproductive capacity, the
removal of more Malay women from potential reproduction by divorce left their
reproductive potential in the younger and middle childbearing years below that of the
Indians, although at the later childbearing ages it was partially offset by the much
higher proportions widowed among the Indians. Compared to the Malays, the
Chinese had higher proportions never married, much lower proportions divorced, and
similar proportions widowed. The net outcome was that in the older reproductive ages
(30-44) a smaller proportion of Chinese than of Malay and Indian women were
outside the married state at any given time (Sidhu & Jones, 1981).Between 1947 and 1980, age at marriage changed dramatically for females of all
ethnic groups (Table 4). The transition to higher age at marriage for Chinese was
completed earlier, and since 1970 has risen by only a year. For Malays and Indians,
the rise began later, proceeded faster and continued right up to 1980, when the
median ages at first marriage were Malays 22, Indians 23, Chinese 24 years. In 1980,
Malay women on average were marrying 5 years later, and Indian women 6 years
Table 4. Median age at first marriage by ethnicity
Total Malay Chinese Indian
Age at first marriage (years)1947 18-2 16-6 20-2 17-1
1957 19-0 17-1 21-6 17-2
1970 21-4 20-5 23-6 21-0
1974 21-4 23-8 22-0
1980 22-9 22-1 24-1 23-1
Average annual increase (%)1947-57 0-4 0-3 0-7 0-1
1957-70 1-0 1-5 0-7 1-7
1970-74 1-1 0-2 1-2
1974-80 0-7 0-5 0-2 0-8
Not available.
Source: Sidhu & Jones (1981). Figures for 1980 computed from Department of
Statistics (1983) General Report of the 1980 Population Census, Vol. 2.
Fertility differentials in Malaysia 413
Table 5. Percentage distribution of women never married byethnicity
Year and age group(years) ChineseMalayTotal Indian
1957
15-1920-2425-2930-3435-3940-4445-49
63-021-45-62-11-51-51-4
45-99.42-41-20-80-60-6
89-743-211-43-82-72-62.5
46-89.42-51-10-50-50-6
15-49
197015-1920-2425-2930-3435-3940-4445-49
15-49
19-4
83-943-013-85-63-31-9
1-3
32-5
12-4
77-332-48-73-31-91-1
0-7
26-9
31-7
94-059-721-49.55-73-42-4
41-0
11-9
83-037-011-73-92-11-40-9
32-3
198015-1920-24
25-2930-3435-3940-4445-49
15-49
91-553-821-710-15-33-52-8
37-9
89-548-617-27-93-82-21-7
35-8
95-463-228-313-37-65-84-6
41-0
91-254-824-310-65-12.92-2
40-0
later, than had their mothers’ generation in 1957. The proportion never-marriedamong Malay and Indian women aged 20-24 rose from one-tenth to one-half over thisperiod, and the relative change was even greater at ages 25-29 (Table 5).
Such remarkable changes, which were accompanied by an almost complete shiftfrom parent-arranged to self-arranged marriages, meant that vast numbers of Malayand Indian women were still single in their 20s in 1980 (240,000 compared with 19,000in 1957) and had virtually no role models from earlier generations to guide them. Thedeep-seated social changes underlying these trends were no doubt influenced byexpansion of education, especially for girls, rising prosperity and wider employmentopportunities for women (many of them only available through migration) and acommunication revolution (Jones, 1981; Lee, 1982).
414 L. L. Lim, G. W. Jones and C. Hirschman
At the same time, rates of widowhood declined due to lowered mortality arid a
narrowing of age differences between spouses, and rates of divorce for Malays
declined sharply in the 1950s and early 1960s, a decline that resulted from many of the
same sources as did the rise in age at marriage (Jones, 1980, 1981). The rise in age at
marriage itself was undoubtedly a major reason for declining divorce rates. Couples
were more mature at the time of marriage and, with the tendency towards self-
arranged marriages, had a greater stake in the success of the marriage.
Proportions currently married therefore fell sharply at the younger reproductive
ages concurrently with their steady rise at the older reproductive ages. On balance,
however, the rise in age at marriage was sharp enough to outweigh the other
influences, and therefore the proportion of all Peninsular Malaysian females of
reproductive age who were currently married fell from 72% in 1957 to 62% in 1970
and to 59% in 1980. This decline differed considerably by ethnic group and was
greatest for Indians-from 80% in 1957 to 56% in 1980, most of the decline occurring
in the 1957-70 period.
Desired number of children and practice of contraception
The key factors influencing fertility of married women are the number of children
they desire and their control of fertility, either through deliberate practice of
contraception or through the fertility-inhibiting effects of breast-feeding. Women who
marry later often find that, for a variety of reasons, they do not reach the number of
children they say they want.
In surveys conducted over the past 20 years, desired family size among women in
the reproductive ages in Peninsular Malaysia has been higher for the Malays than for
the other ethnic groups, higher in rural than in urban areas and higher in women with
less education and lower income. Part of the ethnic differentials in the 1966-67 West
Malaysian Survey can be explained by the lower educational levels of the Malays and
their greater concentration in rural areas, for when these factors are held constant,
desired family size among the Malays did not always exceed that among the other
ethnic groups. Similar multivariate analysis has not yet been performed for the most
recent survey-the 1983 Survey on Health and Family Planning in Johore and Perak
(National Family Planning Board, 1984) but the wide differences recorded in mean
desired family size (Malays 5-0, Chinese 3-95, Indians 3-3) are unlikely to be fully
explained by such socioeconomic differences, particularly in view of the narrowing of
educational differences in recent years and the increased urbanization of the Malays.
Certainly, the Ethnicity and Fertility Survey conducted in 1980 found that the higher
desired family size for Malays survived a range of socioeconomic controls, introduced
singly or in multivariate analysis (Rahman, 1983). In recent years Malay women,
other things being equal, have wanted more children than Chinese or Indian women.
As far as trends are concerned, although the cross-tabulations provided by the
published reports of the different surveys are rarely directly comparable, it appearsthat average desired number of children has declined over the past two decades for all
ethnic groups. The decline has clearly been greatest for the Indians, and least for the
Malays, and indeed both the 1980 Ethnicity and Fertility Survey and the 1983 Johore-
Perak Survey record desired numbers of children for Malays that show no decline-
indeed some rise-over figures in the 1974 national survey.
Fertility differentials in Malaysia 415Malays breast-feed for longer than Chinese or Indians. Over time, however,
breast-feeding has tended to decline for all groups, thus exerting upward pressure onfertility (DaVanzo & Haaga, 1982; Chander et al., 1977). This has been more thanoffset by the sharp increase in the practice of contraception, which according toavailable survey data has risen from 9% of married women of reproductive age in1966-67 to 16% in 1970, 35% in 1974 and 59% in 1983 (Johore and Perak only). In1974, Chinese women were twice as likely to be practising contraception as wereMalay women, with Indian usage being a little lower than for Chinese. In the 1983survey, the Chinese-Malay differential had narrowed somewhat, but this appears tobe mainly due to better recording of folk methods (used mainly by Malays) in 1983.As far as modern methods are concerned, the condom and sterilization were scarcelyused by Malay couples, but heavily used by Chinese and Indians, whereas Malays andChinese relied on the pill much more than Indians. The Indian pattern, then, showeda substantial proportion of contraception oriented to stopping childbearing, ratherthan spacing births, whereas stopping/spacing motivations for the other groups areless clear.
Age patterns of fertilityComparable shifts in age-specific fertility patterns accompanied the changes in the
marital structure of the female population. Both age patterns of fertility and theextent and timing of the decline among different age groups of women differ distinctlyby ethnicity (Figs 2 and 3).
At the onset of the fertility decline in the late 1950s, the Malay and Indianschedules of fertility were markedly younger than that for the Chinese, due largely toyounger average age at marriage. In the 15-19 age group, the fertility rate for Malayswas more than four times and for Indians more than five times that of the Chinese, inspite of the Chinese total fertility rate being higher than the Malay by more than a halfchild per woman. Since younger childbearing tends to produce a higher rate ofpopulation growth, due to the shorter mean length of generation (Coale & Tye, 1961)the higher Chinese total fertility rate in 1958 did not necessarily imply higherpopulation growth.
Figure 2 shows that in 1958 Chinese women had considerably lower fertility in thetwo youngest age groups (15-19 and 20-24) than the other ethnic groups, but afterreaching a peak in the late twenties, Chinese age-specific fertility rates above age 30years were far higher than those for the Indians or Malays. In contrast, the Malay andIndian age patterns of fertility reached a peak in the early twenties and declinedthereafter rather more sharply than did the Chinese. Malay fertility was quite low forwomen above age 30. This reflected the influence both of the higher proportion ofMalay women not currently married at the older ages due to high divorce rates, andthe lower marital fertility rates of older women, irrespective of whether they had beenmarried only once or several times (Palmore & Marzuki, 1969; Hirschman &Fernandez, 1980).
The patterns of fertility decline in the years following 1958 reflect the influence ofthese initial fertility schedules (Fig. 3). As was to be expected, the fall in fertility ratesfor the Chinese was negligible for the 15-19 age group. Fertility decline for theChinese started at the older ages and the rates of decline for women aged 35 years and
L. L. Lwi, G. W. Jones and C. Hirschman416
400-
15-19 20-24 25-29 30-34 35-39 40-44 45-49 15-19 20-24 25-29 30-34 35-39 40-44 45-49
Age (Years) Age (Years)
Fig. 2. Age-specific fertility rates (ASFR) in 1958, 1970, 1980 and 1983, for each ethnic group.
above accelerated over the period. Taking the two periods of 1958-70 and 1970-83,
Chinese fertility decline has been remarkably even across all age groups, except the
youngest and oldest where rates were already very low. This suggests a general fall in
fertility within marriage, with women stopping childbearing once they achieved their
417Fertility differentials in Malaysia
400n
15-)9 20-24 25-29 30-34 35-39 40-44 45.49 15-19 20-24 25-29 30-34 35-39 40-44 45-49
Age (Years)Age (Years)
195819701983
15-19 20-24 25-29 30-34 35-39 40-44 45-49
Age (Years)
Fig. 3. Age-specific fertility rates (ASFR) for Malays, Chinese and Indians, by year.
desired family size. As the decline was especially sharp for women in the 25-34 agegroups, where the fertility potential is highest, it tended to have a large effect on thecompleted size of Chinese families. Even for the 20-24 age group, where fertility wasalready quite low due to delayed marriage, the fertility rate fell by more than half over
418 L. L. Lim, G. W. Jones and C. Hirschman
the period. The total fertility rate for the Chinese fell from 6-5 in 1958 to 2-7 in
1983.The Indian age pattern of fertility decline began at the youngest ages where
fertility was highest and then proceeded to cover all ages. A striking feature is the
dramatic decline in childbearing for the 15-19 year age group, matched only by a
slightly smaller decline for the 20-24 age group between 1958 and 1970, reflectingmainly the sharp rise in age at marriage during this period. The age-specific fertilityrate of Indian women aged 15-19 fell from 222 per thousand in 1958 to only 69 by the
year 1970. For Indian women at ages 25-29 and 30-34, fertility was fairly steady up to
the mid-1960s, even increasing slightly for the 30-34 age group, but then declined
consistently. The distinct shift in the main childbearing years for Indian women fromtheir early to late twenties as a result of rising age at marriage, and a sharp drop in
childbearing by the age of 40 meant that the reduction in completed family size was
greatest for the Indians, by more than four births per woman between 1958 and 1983.The picture of Malay age-specific fertility over time is decidedly more complex.
The decline in fertility rates for Malay women has been confined almost entirely to the
15-19 and 20-24 age categories. This heavy concentration of the decline among the
youngest age groups reflects the effect of increasing age at marriage, which led to a
shift in the peak childbearing years from 20-24 in the late 1950s to 25-29 by the late1970s. At older ages, the period trends have been uneven. At ages 30-34, there was an
actual increase in fertility in the 1960s with a decline only beginning in the 1970s. This
age pattern is consistent with the evidence that while Malay women were marryinglater, their desired family size changed little, so it was simply a matter of postponingchildbearing to older ages. The sharp decline in divorce rates among Malays duringthis period, by tending to inflate the proportion currently married, may also have
contributed to the increase in fertility for those in their thirties. Taking the entire
period 1958-83, Malay age-specific fertility rates for all groups aged 30 and above
were in fact higher in 1983 than in 1958 largely because of the increases in marital
fertility for all age groups above 25 years in the 1978-83 period.These ethnic differences in the age patterns of fertility and the fertility transition
by age groups account for the changes in the shape and shifts of the curves for the
years depicted in Fig. 2. Comparison of the four figures shows:(1) the change in the relative position of the Malay fertility schedule from the
lowest in 1958 to the highest by the 1980s; the Malay age-specific fertility rates
for every 5-year age group above 25 rose from lowest to highest place amongthe ethnic groups;
(2) an earlier peak in childbearing among the Malays and Indians from 1958 to
1970 and a subsequent shift to older ages so that by the 1980s, the ages 25-29
represent the prime childbearing years for all three ethnic groups;(3) that throughout the entire period of fertility transition, Chinese fertility for the
two youngest age groups remained below the levels for the other two ethnic
groups; as age at first marriage rose and teenage marriages became uncommonfor all three ethnic groups, differences in age-specific fertility rates for those
aged 15-19 narrowed considerably; fertility of the 20-24 age group, however,
consistently remained lowest for the Chinese;(4) that from 1958 to 1970, the sharp decline at the older ages brought the Chinese
Fertility differentials in Malaysia 419
fertility schedule below that of Malays aged 35 and over, although still higherthan that for the Indians for all ages above 25; the higher age-specific fertilityrates for the Chinese compared to the Indians at the older ages may be due tothe Chinese rates having been pushed up by the older pattern of childbearingor the Indian rates having been pushed down by the higher incidence ofwidowhood (due to the historically wide age gap between spouses and alsohigher mortality levels); by the 1980s, however, the Chinese and Indianfertility schedules had more or less merged above the peak years ofchildbearing;
(5) that the fertility curves for the Chinese and Indians above the primechildbearing years have become more concave, conforming to the expectedshape of controlled fertility curves (Knodel, 1977); the concave shape of thefertility schedules above age 25 implies sharp drops in fertility once desiredfamily size is reached, so that childbearing rarely continues above age 40 and isalmost unheard of at ages 45 and above; the shape of the fertility curve forolder Malay women, on the other hand, has changed relatively little.
In the fertility transition over 25 years, then, although the Chinese decline startedfrom the older ages and the Indian from the youngest ages, both widened to cover allages, suggesting that rising age at marriage and family limitation within marriagecombined to produce the drop in fertility to levels as low as those prevailing in manywestern countries-as recently as the 1960s. The more moderate Malay decline, on theother hand, confined mainly to the younger ages below 25, reflects the importance oftrends in marital structure. The more complex Malay patterns for the middle andolder ages (falling then rising fertility over time) appear to reflect the interplaybetween changes in the timing of births consequent on sharp increases in age atmarriage, and changes in completed fertility in the midstream of the Malay fertilitytransition.
Change in crude birth rates: 1958 to 1970 and 1970 to 1980
It is possible to study the relative effects of changing age-sex composition, maritalstructure and marital fertility on the fertility declines for each ethnic group byanalysing the change in crude birth rates between two dates. Unfortunately, becauseof the unavailability of census or nationwide survey data after 1980, it is not possibleto do this for the period" of greatest interest between 1978 and 1983 when Malayfertility rates rose. The method of analysis adopted here is that of Retherford & Cho(1974) which separates the contribution to changes in the crude birth rate of changesin (i) age-sex composition of the population, (ii) marital structure as indicated by theproportion of women currently married (and therefore exposed to the risk ofchildbearing), and (iii) marital age-specific fertility rates.
Marital fertility can be further analysed by birth order or parity. Total marital’fertility is a sum of lower order (first, second, and third) and higher order (fourth orhigher) births. A rise in age at marriage and delay in the onset of childbearing maydefer lower order births to the older age categories, and perhaps even contribute to atemporary rise in overall marital fertility.
Tables 6 and 7 show the contributions of these three components to changes in thecrude birth rates of the three ethnic groups for two periods, 1958-70 and 1970-80.
420 L. L. L;’m, G. W. Jones and C. Hirschman
Table 6. Contributions to change in the crude birth rate (CBR) of the Malay, Chineseand Indian populations, 1958-70, Peninsular Malaysia
CBR
Age group(years) Age
Proportionmarried ASMF* Total 1958 1970
Malay15-1920-2425-2930-3435-3940-4445-49
15-49
Chinese15-1920-2425-29
30-3435-3940-4445-49
15-49
0-87-0-73-1-96-0-50-0-630-040-02
-2-88
0-291-410-441-370-31
-0.18-0-09
3-54
-5-10-3-33-0-460-060-150-100-05
-8-35
-0-97-3-21-1-18-0-33-0-020-060-02
-5-63
-0-19-0-56-0-250-580-280-210-00
0-08
0-060-26
-1-38-1-90-1.78-1-01-0-25
-5-99
-4-43-4-61-2-670-14
-0-190-350-08
-11-33
-0-62-1-53-2-12-0-86
-1-49-1-13-0-32
-8-08
45-5 34-2
38-6 30-5
Indian15-1920-2425-2930-3435-3940-4445-49
15-49
2-911-55
-1-680-55
-0-04-0-110-00
3-18
-7-04-4-73-0-82-0-12
0-050-060-03
-12-57
-0-21-0-33-2-10-1-42-0-88-0-10-0-07
-5-11
-4-34-3-51-4-61-0-99-0-87-0-15-0-04
-14-51 45-3 30-8
Age-specific marital fertility.
In Table 7 the contributions to changes between 1970 and 1980 by order of births are
also shown. The contribution of each component change is expressed in terms of the
points of the crude birth rate. The row figures indicate the effect for each age group,while the column totals show the sum effect of each factor. The grand total is
equivalent to the total change in the crude birth rate over the period.Table 6 shows that for the period 1958-70 the strong decline in the Malay crude
birth rate by more than 11 points was due mainly to age structure (almost 3 points)and marital structure (8V2 points), the latter especially in the two youngest age groups.Marital fertility barely changed during the period, since the slight declines below age25 were offset by slight increases for the older ages.
Fertility differentials in Malaysia 421
Table 7. Contributions to change in the crude birth rate (CBR) of the Malay, Chinese
Agegroup(years)
Malay15-1920-2425-2930-3435-3940-4445-49
Total
Chinese
15-1920-2425-29
30-3435-39
40-4445-49
Total
Indian15-1920-2425-29
30-3435-3940-4445-49
Total
and Ir
Age
0-182-072-420-12
-0-010-06
-0-03
4-80
-0-030-432-151-240-320-180-03
4-33
0-193-44
4-501-09
-0-270-14
0-02
9-09
idian popi
Proportionmarried
-2-39-2-53-0-72-0-220-010-030-01
-5-81
-0-31-0-69-0-80-0-22-0-010-0100
-2-03
-2-13-3-70
1-23-0-35-0-06-0-01
-00
-7-47
jlations,
Total
0-55-0-030-41
-0-22-0-110-01
-0-21
0-41
0-21-1.23
1-52-2-12-1-86-0-86-0-18
-7-56
0-65-0-090-25
-1-38-1-33-0-55-0-12
-2-58
1970-80,
Age-spe
&2
0-531-161-710-430-04
-0-01-0-02
3-84
0-22-0-19
1-030-420-060-0100
1-56
0-671-661-770-610-03-0000
4-74
Peninsul
cine mariti
Pa
3
-0-01-0-520-530-410-06
-0.01-0-01
0-43
-0-02-0-49-0-280-180-040000
-0-58
-0-07-0-410-620-280-030000
0-46
ar Malay
il fertility
rity
4
-00-0-47-0-310-310.0600
-0-02
-0-42
-0-01-0-37-0-72-0-13-0-01
00-0-01
-1-24
-0-01-0-730-010-08
-0-04
-0-0100
-0-70
sia
5+
0-03-0-20-1-52-1-38-0-260-04
-0-15
-3-44
0-01-0-18-1-55-2-60-1.94-0-87-0-16
-7-30
0-05-0-61-2.17-2-35-1-35-0-54-0-10
-7-08
Total
-1-66-0-492-12
-0-33-0.110-10
-0-23
-0-60
-0-14-1.49-0-17-Ml-1-55-0-67-0-15
-5-26
-1-29-0-35
3-51-0-65-1-67-0-42-0-10
-0-96
CBR-1970, 1980: Malay-34.2, 33-6; Chinese-30-5, 25-2; Indian-30-8, 29-8.
There was a very different pattern for the 1970-80 period (Table 7) with the Malaycrude birth rate falling by only half a point. While further although slower decline inthe percentages of women married at the younger reproductive ages continued toreinforce the downward effect of marital status trends on birth rates, the age structureeffects began to move in the opposite direction, notably as a result of the entry of largecohorts born in the late 1940s and 1950s into their prime childbearing years. Theoverall marital fertility component showed no net reduction in the 1970-80 period
422 L. L. Lim, G. W. Jones and C. Hirschman
(Table 7), but there were actually dynamic changes in the timing and quantity of
births. First, second and third births went up sharply for Malay women but were offset
by equally large reductions in higher order births. The increase in lower order births,
especially among those in their twenties and early thirties, is almost certainly a
reflection of increasing age at marriage and delayed family formation.
It was the marital fertility by parity of women in their prime childbearing years,25-29 and 30-34, that was crucial in determining the recent changes in Malay fertility.While the marital fertility component tended to increase fertility slightly in the 25-29
age group and reduce it very slightly in the 30-34 age group, for both these age groupsthere was an increase in first to third order births and a reduction of higher order
births. Even for age groups 35 and above, there was actually a reduction of fifth and
higher births. Table 7 strongly suggests a shift in the timing of creation of small and
medium size families and a decline in the incidence of large families among Malays,and that the halt in the downward transition is little more than a partial ’making-up’ of
births by Malay women who are marrying later.The opposite was true for the Chinese. The decline in the Chinese crude birth rate
was achieved in spite of age structure changes which were making for higher fertility.In the first period (Table 6), however, the age structure effects were more than offset
by changing marital status distribution, and it was the decline in marital fertility that
was the primary influence in lowering the Chinese crude birth rate. By the second
period, 1970-80 (Table 7), the effect of declining marital fertility was even more
pronounced. Had it not been for the substantial decline in marital fertility, there
would have been no fall in the Chinese crude birth rate since the lowered proportionmarried was not able to offset completely the greater tendency of age structure
changes to raise the birth rate. That the fall in marital fertility among the Chinese
involved an actual limitation of family size is evident (Table 7). There was a dramatic
drop in fifth and higher order births which alone would have reduced the crude birth
rate by 7-3 points. Even third and fourth order births fell in the 1970-80 period. The
parity analysis shows a trend towards a pattern of fertility characteristic of a
population with a high degree of control over fertility.Like the Chinese, the Indians experienced the combined negative effects on
fertility of changes in proportions married and marital fertility, which more than offset
the positive effect of the age structure (Tables 6 and 7). Unlike the Chinese
experience, however, it was primarily the sharp drop in the numbers marrying at agesbelow their early twenties that lowered Indian fertility. Particularly in the first period1958-70, the effect of changing marital structure contributed so substantially to
fertility decline that the crude birth rate showed an overall decline of 14-5 points. Inthe 1970s, however, the number of Indian women entering their childbearing yearswas so large that their positive effect on fertility could not be completely offset by the
increase in age at first marriage, and it was only because of the additional modest
decline in marital fertility that the crude birth rate fell slightly over the decade.
Although the decline in marital fertility was smaller in the second than in the first
period, the parity analysis (Table 7) shows that, again like the Chinese, family size
limitation was occurring. Fourth, and particularly fifth and higher order births
declined over the 1970s. The increase in first, second and third order births was clearlydue to the postponement of childbearing because of the higher age at first marriage.
Fertility differentials in Malaysia 423
Future fertility in Peninsular Malaysia
Linear extrapolation of recent trends in fertility for Chinese and Indians point towardsthe attainment of replacement-level fertility (total fertility rate of 2-1) in the nearfuture, in about 1995 by Chinese and 2000 by Indians. Experience in Singaporesuggests that the decline will not necessarily then cease; fertility levels have fallen wellbelow replacement level in Singapore (total fertility rate of 1-5) and have shown nosign of rising again over the past decade. But marital fertility rates for Indians havenot fallen very much in recent times (only enough to lower the birth rate by 2-6 pointsbetween 1970 and 1980), and the continuing decline in Indian fertility has been duemainly to the continuing rise in age at marriage and in proportions unmarried. If thesetrends slacken, Indian fertility will fall rapidly to replacement level only if maritalfertility rates decline more rapidly than they have in recent years. However, the signsare right: marital fertility rates have fallen in the 1970s at parity four, and sharply atparities five and above; and in recent surveys, family size desires expressed by Indianwomen have been lower than those of Chinese, even though actual fertility levels forChinese are lower than those for Indians.
The future of Malay fertility is much more uncertain, since the evidence points indifferent directions. The rapid rise in age at marriage has been the major factor in thedecline of Malay fertility. With a female median age at marriage of 22 in 1980 andsignificant numbers of unmarried women in their late twenties and early thirties, therewill certainly be demographic pressures for a continued downward trend in Malaylevels of reproduction. But the rise in period fertility from 1978 to 1983, theemergence of a high fertility region in the east coast states, and the persistence of highfertility norms suggest that the pace of the Malay fertility transition may slacken. Theparity analysis from 1970 to 1980 suggests a decline in large Malay families-especiallyof five or more children. However, the relative cost of children, particularlyeducational cost, may be less of a constraint for Malay than for Chinese and Indianparents (Hirschman, 1986). It seems that Malay fertility transitions will continue, butat a somewhat slower pace, and the fertility differential between the Malays and theother ethnic communities may continue to widen for some time.
As regards recent policy changes, in September 1982 the Prime Minister statedthat Malaysia should aim- for a population of 70 million (almost a five-fold increaseover the present population), and this aim has now been formalized by its inclusion inMalaysia’s development plan. Child relief tax allowances have been increased forevery subsequent child instead of the earlier downward scale, and maternity benefitshave been extended up to the fifth child. The Prime Minister exhorted Malaysians to’go for five’, and other politicians have sung the praises of early marriages and largefamilies.
What effect all this will have (Jones & Lim, 1985) depends on the extent to whichpeople are swayed by appeals to the national interest and by politicians’ claims thatlarge families are good for them. In this most personal of all decisions, it would besurprising if many people put the national interest over private concerns; unlessMalaysians are different from the French (exhorted by successive governments tohave more children, to little apparent effect) it will take substantial economicincentives to put very much upward pressure on fertility. The incentives offered thus
424 L. L. Lwi, G. W. Jones and C. Hirschman
far in Malaysia are symbolic rather than economically meaningful, and it is unlikelythat substantial subsidies for childbearing will be introduced.
It may be that, for some Malays, appeals to ethnic solidarity will be effective,
particularly if these appeals can be linked to religious revivalism. But to encourageethnic rivalry would be a dangerous policy indeed, and though there are suspicions in
non-Malay quarters that change in the relative size of ethnic groups is the ultimate
object of the 70 million policy, this has not been apparent, nor is it likely to become
so, in public statements by political leaders.
References
CHANDER, R., PALAN, V. 1., Aziz, N. L. & TAN, B. A. (1977) Malaysia Fertility and Family Survey
1974, p. 162. Department of Statistics and National Family Planning Board, Kuala Lumpur.CHO, L. J., PALMORE, J. & SAUNDERS, L. (1968) Recent fertility trends in West Malaysia.
Demography, S, 732.
COALE, A. J. & TYE, C. Y. (1961) The significance of age patterns of fertility in high fertility
populations. Milbank mend Fund Q. Bull. 39, 631.
DAVANZO, J. & HAAGA, J. (1982) Anatomy of a fertility decline: Peninsular Malaysia, 1950-1976.
Popul. Stud. 36, 373.FREEDMAN, R. (1979) Theories of fertility decline: a reappraisal. Social Forces. 58, 1.
HIRSCHMAN, C. (1974) Estimates of the Intercensal Population, 1957-1970, by Sex, Community and
Age Group: Peninsular Malaysia. Research Paper No. 9. Department of Statistics, Kuala Lumpur.HIRSCHMAN, C. (1979) Political independence and educational opportunity in Peninsular Malaysia.
Social. Educ. 52, 67.HIRSCHMAN, C. (1980) Demographic trends in Peninsular Malaysia, 1947-75. Popul. Dev. Rev. 6,
117.
HIRSCHMAN, C. (1986) The recent rise in Malay fertility: a new trend or a temporary lull in a fertility
transition? Demography, 23, 161.HIRSCHMAN, C. & FERNANDEZ, D. (1980) The decline of fertility in Peninsular Malaysia. Genus. 36,
93.JONES, G. W. (1980) Trends in marriage and divorce in Peninsular Malaysia. Popul. Stud. 34, 279.
JONES, G. W. (1981) Malay marriage and divorce in Peninsular Malaysia: three decades of change.
Popul. Dev. Rev. 7, 261.JONES, G. (1984) Demographic Transition in Asia, Chap. 1. Maruzen, Singapore.
JONES, G. W. & LIM, L. L. (1985) Scenarios for future population growth in Malaysia. Kajian
Malaysia, 3, 1.
JONES, G. W. & TAN, P. C. (1985) Recent and prospective population trends in Malaysia. J. S. E.
Asian Stud. 16, 262.KNODEL, J. (1977) Age patterns of fertility and the fertility transition: evidence from Europe and
Asia. Popul. Stud. 31, 219.
KUSNIC, M. M. & DAVANZO, J. (1982) Who are the poor in Malaysia? The sensitivity of poverty
profiles to definition of income. Popul. Dev. Rev. 8 (Suppl.), 17.
LEE, K. H. (1982) Age at first marriage in Malaysia. J. Marr. & Fam. 44, 785.
MALAYSIA DEPARTMENT OF STATISTICS (1983a) 1980 Population and Housing Census of Malaysia:
General Report of the Population Census, Vol. 2. Department of Statistics, Kuala Lumpur.MALAYSIA DEPARTMENT OF STATISTICS (1983b) Vital Statistics Peninsular Malaysia, 1982, p. 508.
^"partment of Statistics, Kuala Lumpur.
Fertility differentials in Malaysia 425
NATIONAL FAMILY PLANNING BOARD (1984) Survey on Health and Family Planning in Johore andPerak, Malaysia, Studies on Health and Family Planning in ASEAN Countries. NFPB, KualaLumpur.
PALMORE, J., CHANDER, R. & FERNANDEZ, D. (1975) The demographic situation in Malaysia. In:Population and Development in Southeast Asia. Edited by J. Kanter & L. McCaffrey. Heath,Lexington.
PALMORE, J. A. & MARZUKI, A. B. (1969) Marriage patterns and cumulative fertility in WestMalaysia, 1966-67. Demography, 6, 394.
RAHMAN, A. A. (1983) Family Size Preference and Use of Contraception in Peninsular Malaysia.MA thesis, Australian National University, Canberra.
RETHERFORD, R. & CHO, L. J. (1974) Comparative analysis of recent fertility trends in East Asia. In:International Union for the Scientific Study of Population, International Population Conference,Liege, Vol. 2, p. 163. IUSSP, Liege.
SANDHU, K. S. (1964) Emergency resettlement in Malaya. J. trap. Geogr. 18, 157.SAW, S. H. (1964) A note on the under-registration of births in Malaya during the intercensal period
1947-1957. Popul. Stud. 18, 35.SAW, S. H. (1967a) Fertility differentials in early postwar Malaya. Demography, 4, 641.SAW, S. H. (1967b) A note on the fertility levels in Malaya during 1947-1957. Malayan econ. Rev 12
117.SIDHU, M. S. & JONES, G. W. (1981) Population Dynamics in a Plural Society: Peninsular Malaysia,
Chap. 11. University of Malaya Co-operative Bookshop, Kuala Lumpur.SMITH, T. E. (1952) Population Growth in Malaya: an Analysis of Recent Trends. Oxford University
Press, London.
ReceivedTith August 1986
