South Asia with more than 1.3 billion people represents a vast diversity of cultures and socio-economic conditions. Economic status, kinship systems, traditions and religious values, all contribute to a strong overall son preference in South Asia. At a recent international symposium on sex preference (UNFPA, 1994), the countries and areas represented were grouped as follows, taking into account the relationships between fertility decline, son preference and sex ratio at birth:

o Rapid fertility decline, strong son preference, abnormal sex ratio at birth: China and Taiwan Province of China, and Republic of Korea;

o Rapid fertility decline, no son preference, and normal sex ratio at birth: Indonesia, Sri Lanka and Thailand; and

o Slow fertility decline, strong son preference and normal sex ratio at birth: Bangladesh, India and Pakistan.

In this article, we shall review, therefore, the factors responsible for strong son preference in Bangladesh, India and Pakistan, three South Asian countries with a combined population of about 1.2 billion (ESCAP, 1995), and examine the possible reasons for the absence of widespread societal preference for sons in Sri Lanka, a country of 18.2 million people.

_____________

* The author of this article is Director, Population Division, Ministry of Health and Social Services, Colombo, Sri Lanka.

The common indices of sex preference used in various studies are sex ratio at birth, sex differentials in infant and child mortality, and preferences obtained from survey responses.

Abnormal sex ratios at birth have been found to be due to sex-selective abortions. The increasing use of pre-natal ultrasound and amniocentesis procedures, which make selective abortion possible by revealing the sex of a fetus, contribute to high sex ratios at birth (Population Information Program, 1994). Although many South Asian countries have banned pre-natal gender tests to prevent sex selective abortions, illegal tests are available and more female fetuses are aborted than males (Patel, 1989); Benjamin, 1991; Nandan, 1993).

In South Asia, sex preference is mainly manifested in the form of excessive mortality of female children. The excessive mortality of female children relative to males is found to be due to the discrimination against females in the allocation of food and health care within the household. In many South Asia societies, sons are the parents only source of security in old age. This is particularly so where women have little economic independence or cannot inherit property. Son preference is also strong when daughters are more expensive to marry off than sons owing to the dowry system; in addition, women have few opportunities to earn income and invest household resources in female children (Cain, 1984). The two papers on India presented at the aforementioned symposium on son preference showed the impact of son preference on child survival in India. A significant finding was that risks to daughters increased with more older female children in the household. Similar hazards to son survival were not observed. The daughter mortality differentials between northern and southern India were attributed to variations in women's status and the strength of sex preference (Visaria, 1994; Das Gupta, 1994).

Similarly, data from Bangladesh showed a higher risk of survival for girls having older sisters than that of boys who have older brothers. The study showed that the complex interaction of areal variations in service availability, attitudes, preferences, practices, and social and economic status effects the determined outcomes (Bairagi, 1994).

In Pakistan, strong son preference has been noted because sons are regarded as economic assets and old-age security (Khan and Serageldin, 1977; Ali, 1989). More recent data from Pakistan further confirms the continued desire for sons (NIPS, 1992). The Pakistan Demographic and Health Survey of 1990/91 showed that, of the women with no children, about one-third desired to have a son, while the preference for having a daughter was negligible. Among those who had two daughters and no son, almost all (93 per cent) wanted their next child to be a son. However, the incidence of wider neglect of female children or preferential treatment for male children was not evident from the data (Karim, 1994).

In South Asia, cultural factors such as kinship systems and religious traditions also tend to value males more highly than females. In India and Bangladesh for example, traditional patrilineal kinship systems require women to marry out of their families of origin and after marriage not provide financial or even emotional support to their parents (Greenhalgh, 1991). In the Hindu tradition, only sons can pray for and release the souls of dead parents, and only males can perform birth, death and marriage rituals (Benjamin, 1991).

It has been observed that females are more likely than males to die in early childhood particularly in South Asia owing to poor nutrition and health care. Visaria (1967), in a study of inter-State differences in sex ratios between 1901 and 1961, showed that throughout the period sex ratios were persistently higher in the northern Indian States and lower in the southern ones. He argued that mortality differentials by sex were mainly responsible for these differences.

Dyson and Moore (1983) also found that sex differentials in child mortality are much higher in the northern than in southern States of India. They established that the main reason for the relatively high sex ratios in the northern part of the country is higher female mortality; they attributed this phenomenon to the discrimination against females in access to food and medical care. They also related the differentials observed in the north and south to variations in kinship systems and female autonomy.

In a study of 11 villages in Ludhiana District of Punjab, Das Gupta (1987) found that sex differentials by birth order are far stronger than those by socio-economic status. Women's education was associated with reduced child mortality but stronger discrimination against higher birth order girls. She noted that the strong underlying preference for sons appears to be the outcome of women's "structural marginalization" in that culture, as a result of being of low value to their parents.

D'Souza and Chen (1980), using data from Matlab Thana, Comilla District in Bangladesh, found higher female than male mortality from shortly after birth through the child-bearing ages. The most marked differences were found in the age group 1-4 years where female mortality exceeded male mortality by as much as 50 per cent. Son preference in terms of parental care, feeding patterns, intra-family food distribution and treatment of illness has been cited as a possible cause of childhood mortality differences by sex.

Chen and others (1981), using a height-for-age (stunting) indicator as a measure of chronic deprivation among a survey of 882 boys and girls in Matlab, Bangladesh, demonstrated that a higher percentage of girls are severely or moderately malnourished than boys. The authors also found that boys were seen two-thirds more often that girls at the diarrhoea treatment centre in Bangladesh, even though the centre provides free ambulance transport and treatment.

To what extent does son preference have an effect on fertility and contraceptive use? Das (1987) analyzing Indian data in Gujarat State noted that the effect of son preference on overall fertility is significant and concluded that future fertility might be reduced if gender preference could be realized. Bairage and Langesten (1986), using data from Bangladesh, observed that in the study area, although son preference is very strong, more than 98 per cent of women desire to have at least one daughter. However, women with a higher proportion of sons are less likely to want more children and are more likely to practise contraception. Chowdhury and others (1990) found that, in two given areas of Bangladesh with a strong son preference, its effect on fertility was stronger in the area with relatively high contraceptive prevalence. Thus, the authors arribute the absence of fertility effect on son preference in Pakistan to the low contraceptive prevalence in that country. In another study on Bangladesh on the effects of family sex composition on fertility preferences and behaviour during the period 1977-1988, Chowdhury and others (1993) found that the sex composition of living children was systematically related to fertility preferences and behaviour, with a higher number of sons at each family size associated with a higher percentage of women wanting no more children, a higher percentage currently using contraception and lower subsequent fertility.

In a study of the scheduled caste population in the Indian State of Assam, Nath and others (1994) found that couples having two surviving sons are less likely to have a third child than those without a surviving son and those with only one surviving son.

In a recent study in Matlab, Bangladesh, Rahman and others (1992) indicate that son preference can have a strong effect on contraceptive use and fertility. In the study area, which had a contraceptive prevalence of about 50 per cent and an average of four children per couple, the researchers calculated that eliminating the preference for sons would increase contraceptive use by 10 per cent and continuation rates by 15 per cent. Such increases would likely avert nearly one birth for every two couples.

In a study of the effect of sex preference on contraceptive

use and fertility in rural South India, Rajaretnam and Deshpande (1994) found that couples overall prefer families with at least one son and one daughter, but in areas where contraceptive prevalence rates are high, most couples have two sons with or without a daughter before they initiate contraceptive use. In low-prevalence areas, couples often have two sons and one daughter before starting to practise family planning. In the absence of sex preference, the authors noted that contraceptive prevalence rates could be expected to increase by about 12 per cent in the high prevalence areas and by about 25 per cent in low prevalence areas. In both areas, the levels of marital fertility can be expected to decline by about 20 per cent from current levels.

What could be done to reduce the desire for sons in South Asia? As was evident from the studies in India (Das Gupta, 1994) and will be seen from Sri Lankan data, the status of women has an important influence on son preference in South Asia. Increasing the economic opportunities for women and raising the value of women's labour would increase the likelihood of parents regarding their daughters as economic assets rather than as liabilities. Increasing the opportunities of education for female children may increase their income-earning potential and thereby raise their economic value to their parents (Bourne and Walker, 1991).

Better access to food and medical care in general would enable parents not to discriminate against female children in the allocation of household resources. Also, better opportunities for old-age security would minimize the urge to have sons (Cain, 1984). Above all, far-reaching changes in the cultural and economic status of women would enable women to resolve the conflict between the achievement of their smaller family size preferences and sex preferences (table 1).

Region/State	Percentage of couples protected by family planning	Female labour force participation rate	Percentage of females literate	Index of son preference
	1979	1971	1971	1978
South
Kerala	28.8	13	54.3	17.2
Tamil Nadu	28.4	15	26.9	11.5
Andhra Pradesh	26.5	24	15.7	8.9
Karnataka	22.4	14	20.9	11.2
North
Gujarat	20.1	10	24.7	20.8
Rajasthan	13.0	8	8.5	n.a
Uttar Pradesh	11.5	7	10.7	25.0
Madhya Pradesh	20.9	19	10.9	21.9
Punjab	25.0	1	25.9	31.3
East
Bihar	12.2	9	8.7	24.3
West Bengal	21.2	4	22.4	18.4
Orissa	24.4	7	13.9	15.7
All India	22.1	12	18.7	20.2

Note: A son preference index of zero would imply equal preference for sons and daughters.

Source: Adapted from table 5 from Dyson and Moore (1983).

In Sri Lanka, preference for a balanced number from each sex as well as a preference for sons has been observed, although not consistently strongly (Pullum, 1980; Arnold, 1992). It has been shown that improvements in the status of women during the past two decades have weakened the preference for sons in Sri Lanka (Abeykoon, 1994).

Sex ratios at birth by parity do not indicate any sex preference in Sri Lanka (table 2). It can be seen from table 3 that the mean desired family size between two-son and two-daughter families is negligible. The interpretation of this finding is that respondents tend to prefer balanced families. The mean is only 0.19 less for balanced families than for the two imbalanced types grouped together, which indicates that parents tend to give greater value to the small family norm than to the sex of the child. However, when Pullum (1980) examined the 1975 data on the desire for more children among non-pregnant women with two children, it was found that 62 per cent of those women with two girls wanted more children compared with 59 per cent of the women with two boys who wanted more, which indicates a slight son preference. A similar observation was made by Arnold (1992) using data from the 1987 Demographic and Health Survey.

Source: Computed from data of the Registrar General's Department.

Sex composition	Mean	Standard deviation
2 boys	2.79	0.82
1 boy, 1 girl	2.60	0.79
2 girls	2.78	0.86

Source: Pullum (1980) op. cit.

Discrimination against girls has been widely reported in countries of South Asia where son preference is strong. In Sri Lanka too, persistent higher female mortality at ages 1-4 and 5-9 years prior to 1962 has been attributed to greater parental care and favoured treatment with regard to food and medical attention for male children (Nadarajah, 1983). However, recent data do not indicate any evidence of such practices in Sri Lanka (table 4).

	Male	Female	Ratio (male/female)
Mortality
Infant mortality rate	40	25	1.60
Neonatal mortality rate	26	15	1.73
Post-neonatal mortality rate	10	10	1.00
Child mortality rate	10	10	1.00
Nutrition
Stunted (%)	29	26	1.12
Underweight (%)	38	37	1.03
Wasted (%)	12	12	1.00

Note: Data exclude the northern and eastern provinces.

Source: Arnold (1992) op.cit., tables 4 and 6.

Why is Sri Lanka an outlier in South Asia with regard to sex preference? Compared with other South Asian countries, the status of women in Sri Lanka is found to be more advanced. The many social welfare programmes carried out during the post-independence decades did create many favourable conditions which promoted greater participation of women in the development process. These include (a) rapid expansion of literacy and educational attainment of women, (b) improved life expectancy and decline in fertility and (c) wider participation of women in formal and informal economic activities.

Female literacy, which was only 8.5 per cent at the turn of the present century, rose to 83.2 per cent in 1981. The difference in the male and female rates of literacy, which was 33.5 percentage points in 1901, declined to 7.9 percentage points in 1981. With regard to school attendance in 1981, the percentage of those aged 5-14 years attending school was equally high for both sexes at 84 per cent. In the age group 15-19 years, 42 per cent of females were attending school compared with only 29 per cent for males. Nearly 60 per cent of ever-married women in the age group 15-49 years had an education beyond the primary level in 1987.

Female expectation of life at birth, which remained lower than that of males in 1946 at 41.6 years, increased to 72.1 years in 1981 surpassing male life expectancy by 4.4 years; female life expectancy in 1995 is 75 years and that of males 70 years (ESCAP, 1995). The maternal mortality rate, which was around 20 per thousand live births in the mid-1930s, declined to 0.5 per thousand in 1985. Female age at marriage increased from 20.9 years in 1953 to 25.5 years in 1993. The contraceptive prevalence rate rose from a level of 32.0 per cent in 1975 to 66.1 per cent in 1993. The rise in the educational attainment of females has been the single most important socio-economic factor that has contributed to fertility decline in Sri Lanka by influencing the age at marriage and contraceptive use. The total fertility rate (TFR) has declined from an average of 5.1 children per woman during the period 1952-1954 to 2.3 during the period 1988-1993.

The economic participation of women in the modern sector has shown a marked increase in recent years. Female participation in manufacturing industries has increased visibly; in 1986, about 45 per cent of the total employed in this sector were women. This increase has been largely in urban industry. Women have been employed in increasing numbers in export-oriented modern industry. The employment of females in these industries has increased employment opportunities for women and has also given them some degree of economic independence and personal freedom. The increased participation of women in the modern sector has also improved their social mobility. While the initial impetus to fertility decline came about through rising aspirations of females resulting from the expansion of educational opportunities and attainment, in more recent decades, the upward social mobility of females brought about by the wider availability of economic opportunities and participation in the modern economic sectors have also contributed to higher contraceptive use and fertility decline in Sri Lanka.

Within the family, Sri Lankan women are less vulnerable to discrimination and oppression than their counterparts in other South Asian countries. The extreme situation of male dominance such as dowry deaths and widow immolation are absent.

From the foregoing discussion it is evident that the wide-spread son preference in Bangladesh, India and Pakistan is manifested in the form of post-natal discrimination against the girl child. This is in contrast to the situation in East Asia where pre-natal sex discrimination prevails (UNFPA, 1994). It is also clear that the absence of widespread societal preference for sons in Sri Lanka is due to the relatively high status of women in that society.

The strong overall son preference in a population of 1.2 billion people in South Asia has varied demographic, social, economic and health implications for the region.

The demographic impact of son preference appears to be closely associated with family size norms, availability of contraceptive services and sex-selection technologies. The imbalance in sex ratios of children under five years of age resulting from the discrimination against female children may lead to female sex imbalances in the marriageable ages in the future. If females become scarce, the situation may improve the status of women in the long term. However, it may also contribute to an increase in sex-related crimes and violence as well as homosexual activities (Park and Cho, 1995).

The decline of family size norms and the availability of sex-selection technologies among subgroups of the population in South Asia in the course of its demographic transition may contribute to the widening of the social gap between males and females in the future, as children of smaller families who are likely to be predominantly males may be advantaged in the allocation of household resources for education, nutrition and health care. Also, with poor medical technologies and facilities available for induced abortions, it is likely that maternal mortality and morbidity rates may increase in South Asia with declining family size norms in the future. Therefore, the possible long-term improvement of the position of women resulting from the projected shortage of females in the reproductive ages may be offset due to the above implications.

The gradual erosion of the widespread societal preference for sons in the countries with a large population in South Asia may be brought about, as has taken place in Sri Lanka, by raising the economic and social value of the girl child through education. In a modernizing society such as in the Republic of Korea, where sex preference and status of women are relatively high, Hong (1994) found a clear negative relationship between boy preference and the educational attainment of the mother. Providing easy access to family planning services would also relieve women from the burden of excessive child-bearing and release them to participate in productive economic activities.

There is no doubt that State intervention in primary education and health care including family planning by ensuring widespread accessibility to quality services, would enable parents to attach a higher value to the education and health care of their children irrespective of sex.

Over a period of time, it is likely that wider societal acceptance of education, health care and family planning will occur owing to the demonstration effect of the benefits of these services to families and the community at large. This would contribute to the elimination of post-natal sex discrimination and enable couples to resolve the conflict between the achievement of small family norms and sex preferences.

Abeykoon, A.T.P.L. (1994). "Is sex preference an obstacle to reaching replacement fertility in Sri Lanka?". Paper presented at International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, November.

Ahmed, Nilufer R. (1981). "Family size and sex preference among women in rural Bangladesh", Studies in Family Planning, 12(3).

Ali, Syed Nubashir (1989). "Does Son Preference Matter?", Journal of Biosocial Science, 21(4).

Arnold, Fred (1992). "Sex preference and its demographic and health implications", International Family Planning Perspectives, 18(3).

Bairagi, Radheshyam and Ray L. Langsten (1986). "Sex preference for children and its implications for fertility in rural Bangladesh", Studies in Family Planning, 17(6).

Bairagi, Radheshyam (1994). "Excess female child mortality: its levels, trends and differentials in rural Bangladesh". Paper presented at International Symposium on Issues related to Sex Preference for Children in Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, November.

Benjamin, J. (1991). "Socio-religious status of girl child in India", In: Devasia, L. and Devasia, V.V. (eds.) Girl Child in India (New Delhi: Ashish Publishing).

Bourne, K.L. and G.M. Walker (1991). "The differential effect of mothers' education on mortality of boys and girls in India", Population Studies, 45(2).

Cain, M. (1984). Women's Status and Fertility in Developing Countries: Son Preference and Economic Security, Population and Development Series No. 7, (Washington, D.C.: World Bank).

Chen, Lincoln C., Emdadul Huq and Stan D'Souza (1981). "Sex bias in the family allocation of food and health care in rural Bangladesh", Population and Development Review, 7(1).

Chowdhury, A.I., Radheshyam Bairagi and Michael A. Koinig (1993). "Effects of family sex composition on fertility preference and behaviour in rural Bangladesh", Journal of Biosocial Science, 25(4).

Chowdhury, Mridul K. and Radheshyam Bairagi (1990). "Son preference and fertility in Bangladesh", Population and Development Review, 16(4).

Das Gupta, Monica (1987). "Selective discrimination against female children in rural Punjab, India", Population and Development Review, 13(1).

__________ (1994). "Discrimination against female children in India". Paper presented at International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, November.

Das, Narayan (1987). "Sex preference and fertility behavior: a study of recent Indian data", Demography, 24(4).

D'Souza, Stan and Lincoln C. Chen (1980). "Sex differentials in mortality in rural Bangladesh", Population and Development Review, 6(2).

Dyson, Tim and Mick Moore (1983). "On kinship structure, female autonomy and demographic behavior in India", Population and Development Review, 9(1).

ESCAP (1995). "1995 ESCAP Popuulation Data Sheet" (Bangkok: Economic and Social Commission for Asia and the Pacific).

Greenhalgh, S. (1991). Women in the Informal Enterprise: Empowerment or Exploitation? (New York: Population Council).

Hong, Moon Sik (1994). "Boy preference and imbalance in sex ratio in Korea". Paper presented at International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, November.

Karim, Mehtab S. (1994). "Sex preference in Pakistan". Paper presented at the International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, November.

Khan, M.A. and I. Sirageldin (1977). "Son preference and the demand for additional children in Pakistan", Demography, 14(4).

Nadarajah, T. (1983). "The transition from higher female to higher male mortality in Sri Lanka", Population and Development Review, 9(2).

Nandan, G. (1993). "India to ban prenatal sex determination", British Medical Journal, vol. 306.

Nath, Dilip C. and Kenneth C. Land (1994). "Sex preference and third birth intervals in a traditional Indian society", Journal of Biosocial Science, 26(3).

NIPS (1992). Pakistan Demographic and Health Survey 1990-1991 (Islamabad: National Institute of Population Studies).

Patel, Vibhuti (1989). "Sex-determination and Sex-preselection test in India: modern techniques for femicide", Bulletin of Concerned Asian Scholars, 21:2-10.

Park, Chai Bin and Nam-Hoon Cho (1995). "Consequences of son preference in a low-fertility society: imbalance of sex ratio at birth in Korea", Population and Development Review, 21(1).

Population Information Program (1994). "Opportunities for women through reproductive choice", Population Reports, Series M, No. 12.

Pullum, Thomas W. (1980). "Illustrative analysis: fertility preferences in Sri Lanka", WFS Scientific Reports, No. 9.

Rajaretnam, T. and R.V. Deshpande (1994). "The effect of sex preference on contraceptive use and fertility in rural South India", International Family Planning Perspectives, 20(3).

Rahman, M., J. Akbar, J.F. Phillips and S. Becker (1992). "Contraceptive use in Matlab Bangladesh: the role of gender preference", Studies in Family Planning, 23(4).

UNFPA (1994). Proceedings of the International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, 21-24 November.

Visaria, Leela (1994). "Deficit of women, son preference and demographic transition in India". Paper presented at International Symposium on Issues Related to Sex Preference for Children in the Rapidly Changing Demographic Dynamics in Asia, Seoul, Republic of Korea, November.

Visaria, Pravin M. (1969). The Sex Ratio of the Population of India, Census of India 1961, Vol. 1, Monograph No. 10.

For further information on this material please contact: loftus.unescap@un.org