Sex-Preference For Children and Fertility Behaviour in Mid-Level Civil Servants of Bangladesh: A Sample Survey of the Participants of 38th ACAD*

Despite some success stories in the family planning sectol; there still exists a number of, factors? in inducing larger family size in the country. One of these factors acting as deterrent to quick fertility reduction is the prevalence of .;trong son-preference in Bangladesh society. The persistence of son-preference and high fertility in Bangladesh is deep-rooted in the structure of the society and its culture. The children become a source of social, economical and political security of the parents. Besides, the desire for children especially for sons as a means of continuation of family line, grown up sons contribute to the family economically and protect the interest and prestige of the family by providing physical labour in times of necessity. Moreovel: the recourse of the family rests with the frown up sons to take up the responsibilities of the family in times of cconomic crises and the death of the husband The incidence of high infant mortality in Bangladesh acts as stimulating desire for additional children both as )nsurancel and 'replacement' mechanism. More strenuous effort would be expected to replace dead sons than dead daughters. This exhibits a sex-selective replacement mechanism in the countly. Although a good number of research studies are available in the countiy to know nature and extent of the sex-preference PI-evailing in rural Bangladesh. unfortunately there is no literature available to know the extent of prevalence in the civil servants. Civil servants being educated and exercising a great deal of power and position are usually believed to have liberal outlook about family norms. They are to a greater extent involved in the policy making process of the government. Any shiji from the well-publicised policy of 'Two-child family of either sex' to 'One-child family' may bring some bearings upon the fertility behaviour of'the couples. How far this policy change will be pragmatic (1s responded by diffizrent sections of the society should be studied first and be preceded to any policy modzfications. The absence of any such research findings about sex-preference prevailing in the civil servants has inspired the researcher to undertake the study. The article elaborated the research findings of a sample survey conducted by the author among the participants of 38th ACAD (Advanced Course on Administration and Development) of BPATC. No statistical test was done to generalise the result but the article will enable a reader understand an overall demographic behaviour of Civil Servants in Bangladesh. I MDS (Joint Secretary to the Government), Bangladesh Public Administration Training Centre, Saver, Dhaka 44 Bangladesh Journal of Public Administration A sample of 43 respondents has been selected rmdomlyfr-om the participants of 38th ACAD and a survey is conducted in the regard. This paper has three chapters. Chapter one deals with introduction, objective, scope, rationale, methodology and limitation of the study. Chapter two deals with the literature review of sex-preference. This chapter mainly concentrates on available literature review of the subject under study. Chapter three deals with the data analysis and findings and conclusion of the study. From data analysis and findings it is observed that even civil servants being members of the upper stratum of the society favour a family size greater than two. The mean desired family size is 2.37. The most favoured combination of .family is one son and one daughter (65%). The second highly satishing combination is two sons and two daughters. Civil servants are more likely to make a balance of sex in the family but whenever the sex-balancing is controlled obviously son-preference dominates. Two sons in combination with one daughter is preferably to one son and two daughters. Although more than half of the respondents (57%) are reluctant about the sex of the first-born child it is observed that about three-fourth of them had their first-born child as son. The striking feature of the sex-preference is that when civil servants are blessed with first child as son only then a greater portion of them showed liberal attitude towards sex-selectivity of additional child. About 55.8% of the respondents believe that absence of desired number of sons in the family contribute as nuinbelone factor contributing to the desire for more children in th family. This is followed by a (second ranked) factor as insecurity of survival of the present siblings. The first factor reflects the strong son-preference prevailing while second jizctor discloses the tendency of the parents to resort to insurance mechanism of fertility behavioul: In fine, it can be safely remarked that civil servants are not totally emancipated ?om the evils of sex-composition of the children and reluctance to secselectivity, these improvements are too volatile to withstand any policy change of the government.

This paper has three chapters. Chapter one deals with introduction, objective, scope, rationale, methodology and limitation of the study. Chapter two deals with the literature review of sex-preference. This chapter mainly concentrates on available literature review of the subject under study. Chapter three deals with the data analysis and findings and conclusion of the study.
From data analysis and findings it is observed that even civil servants being members of the upper stratum of the society favour a family size greater than two. The mean desired family size is 2.37. The most favoured combination of .family is one son and one daughter (65%). The second highly satishing combination is two sons and two daughters. Civil servants are more likely to make a balance of sex in the family but whenever the sex-balancing is controlled obviously son-preference dominates. Two sons in combination with one daughter is preferably to one son and two daughters. Although more than half of the respondents (57%) are reluctant about the sex of the first-born child it is observed that about three-fourth of them had their first-born child as son. The striking feature of the sex-preference is that when civil servants are blessed with first child as son only then a greater portion of them showed liberal attitude towards sex-selectivity of additional child. About 55.8% of the respondents believe that absence of desired number of sons in the family contribute as nuinbel-one factor contributing to the desire for more children in th family. This is followed by a (second ranked) factor as insecurity of survival of the present siblings. The first factor reflects the strong son-preference prevailing while second jizctor discloses the tendency of the parents to resort to insurance mechanism of fertility behavioul: In fine, it can be safely remarked that civil servants are not totally emancipated ?om the evils of sex-composition of the children and reluctance to secselectivity, these improvements are too volatile to withstand any policy change of the government.

Introduction
Bangladesh ranks as the World's eighth and Asia's fifth most populous country with a high density of population in the world -more than 800 people per square kilometer. Since independence it has been trying hard to restrain the population growth. So far it has been remarkably successful in reducing population growth to the level of 1.8 percent annually. Total Fertility Rate (TFR) has fallen from 6.8 in 1975 to 3.1 in 1997. The Contraceptive Prevalence Rate (CPR) has increased from 9% in 1975 to 45% in 1993 and 52% in 1999. Along with this success story, there still remains some shortfalls in the mortality area. The

Mohanimad Haroon Chowdhury
Infant Mortality Rate (IMR) is still more than 60 per thousand live births and under-5 child mortality rate is 11 8 in 1996. Total population as projected with a goal of NRR-1' by the end of 2005 will exceed 132 millions. ~o v e i m e n t has a well propagated family planning policy of "Two child-family of either sex", still struggles with a number of socio-economic deterrents in the society. One of these obstacles is the strong sex-preference prevailing in the country.
There are good reasons why people might want large families. Children bring satisfaction and joy to parents and also bring some practical benefitsespecially in terms of social and economic securities. One factor which could tip the scale towards larger families is strong son-preference in the society. Sex-preference visa-vis the son-preference is associated with the socio-cultural and economic considerations in the society. It is the product of social, economical, psychological and cultural values which affect the reproductive behaviour of the couples either as independent variables or in some cases as intervening variables.

Rationale of the Study
Civil servants usually belong to the upper echelon of the society and are in greater extent involved with the policy making process of the government. "Two children of either sex" is the well-publicised policy of the government and acts as the root inducement to the future family planning in the country. But the present size of the population and the future projection even with a drastic reduction of population growth might compel the policy makers to shifl the family planning programmes from "Two-child family" to "Otie-child family" of either sex. How far this policy shifl will be pragmatic and beneficial still remains to be answered in the context of desired family size and the degree of sex-preference prevailing in different sections of the society. Though there are a good number of research studies available to apprise one about the rural and poor people of the country but unfortunately, no literature is available to h o w the extent of prevalence among the civil I NRR-I means Net Reproduction Rate is one which indicates that a woman will be replaced by a single daughter who will complete her total fertility petiod as the surviving replacement. servants. It is a felt necessity for any interested reader as well as the concerned department of the government dealing with reproductive beheviour to know the prevalence of sex-preference in the upper stratum of the society specially the civil servants. The study will serve as a source of data also.

Objective of the Study
The main objective of this study is to find out the nature and the extent of sex-preference prevailing in the mid-level civil servants in the country. The paper attempts to know : the desired family size of the civil servants; the prevalence of sex-preferences among the civil servants r the impact of sex-preference upon fertility behaviour of the civil servants the opinion of the civil servants about the factors inducing larger family size.

Methodology of the Study
The present study is mainly based on a sample survey among the participant of the 38th ACAD). The study deals with the subject in two partsf the first part deals with the literature review of the sexpreference in the society as a whole and the second part with the sample survey. A sample of 43 respondents has been selected ramdomly from total 87 participants of the running 38th ACAD. A questionnaire (annexure-1) has been developed and has been administered to the sample population by the researcher. The data, thereby collected have been analysed. The detailed report of the study with tables is dealt in chapter-3.

Scope of the Study
The study will reveal the degree of sex preference among the civil servants. In the absence of any such back-ground knowledge about the civil servants, we are usually tempted to disentangle this section from the greater part of the society. Being a member of the educated section of the society, exercising power and prestige, a civil servant is usually believed to have nourished all liberal ideas of small family norms and Mohanit~lad Ifat-0011 (%owdl~rrty is generally accepted as free from sex-discrimination between the children. The study will unearth the opinion of the civil servants regarding their desired family size, the desire for additional children and the preference for son in iniluencing the fertility behaviour. These information about the civil servants will help the government to know the extent of sex-preference, its seve~ity which in turn might help in reshaping the family planning policies and programmes in the near future.

Limitations of the Study
The study mainly focussed on the opinion of the mid-level civil servants regarding their expected family size, their craze from son, and opinion regarding the factors influencing larger family size. The fertility behaviour of the civil servants has been dealt not as'a matter of objectives rather as a concomitant impact of the sex-preference upon reproductive behaviour. Fertility as a concept is usually related to women of child bearing age. However, in this study, this part is mainly covered by the statements made by the spouses just to reflect the present reproductive behavior in the family. Furthermore the study has not tried to find out any causal relationship for the prevalence of sex-preference in the civil servants.

Extent of Son-Preference
The extent of son preference nnd its impact upon the reproductive behaviour of the eligible couples greatly influence the slow pace of fertility reduction in the country. The World Fertility Survey (WFS) for Bangladesh 1975 revealed a strong son preference among Bangladeshi women. With no sons but one daughter 84% of the mothers wanted their next child to be a son while the reverse accounted for qnly 27% of the mothers. A strong son preference has also been observed in Indian society. The ideal sex-composition of children in urban India was found to be two sons and one daughter and the intensity of preference for son was observed + 33.' (S. Lahevi Intensity of sex-preference was measured by an index defined by the excess (or deficit) of ideal (for desired) number of sons over that of daughters expressed, in percentage of the ideal (or desired) number of children. The incidence of high infant mortality in Bangladesh acts as a stimulating desire for additional children in the family. Parents want usual guarantee of normal' life of the children. In order to be insured about the number of surviving children larger family has always been desired by the parents. This is called 'insurance mechanism' of high fertility. On the other hand, in the aftermath of the death of a child in the family, parents try to replace the dead child by an addition of new lace in the family as quickly as possible. This is called 'replacement mechanism' of the fertility behaviour. More strenuous effort would be expected to replace dead sons than daughters which exhibit a sexselective replacement meclianism in Bangladesh (A.J. Sufian and Non E. Johnson 1989).

Continuation of family line and old age security
It is believed that sons keep alive the name and identity of the parents in a matrilineal society of ours. The main objective of the procreation of son is to keep the lineage (bangsha). A greater number of son is considered as a source of strength in the family and an inspiration to maintain kinship. In a study conducted in Kamnagar of Rajshahi, it is observed that soilless parents feel inferiority complex in comparison with parents who have sons. This is simply associated with a feeling of gain in status having a son (P.C Sarkar 1997). Apart from continuation of family line and strengthening family status, parents desire for male offspring mainly for old age care and assistance.

Sex-Preference nd Fertility behaviour as revealed in Bangladesh Fertility Survey 1989
In a study' conducted by the present researcher with the primary raw data of Bangladesh Fertility Survey 1989 revealed a strong sonpreference prevailing in Bangladeshi society. Two surviving sons and one daughter appeared to be the determining combination of children for fertility behaviour. The mean desired family size is 2.93 that is on an average a three-child family is preferred by Bangladesh women.
The desire for additional children is dependent upon the nuuiber of surviving children. Among the women with two surviving children the desire falls down to 50 present from 90 percent as observed in case of The study was conducted as a requirement of Post Graduate Diploma Depree in Population Studies, conferred by IIPS, Mumbai, India, 7 April 1995. women with one surviving child. But the situation appeared to be discriminatory when the preference of son is taken into the picture.
In the hope of getting one son, mothers intended to increase the family size up to four daughters; in such a case even 50% of the mothers refused to stop child bearing. On the other-hand, if one son is born to a family of single daughter about 54% of the currently married women wanted to stop child bearing. Mothers with no sons were less likely to use contraception than mothers with no daughters. With single sex the highest contraception occurred with four surviving sons (42.8%) but it did not happen with four daughters only (26%). An addition of a son to a family of single daughter raised the contraception rate to 38.4%; a further addition of a son raises contraception; rate to a maximum of 46.4% making the most desired family of two sons and one daughter. Even a third son in the family maintained the contraception rate almost same (46.3%). Therefore two to three sons in combination of one daughter was the most favoured sex composition of a family in Bangladesh. Furthermore, when .question of balancing of sex comes up, two sons and two daughters registered 44.3% contraception rate indicating that Bangladeshi mothers were tolerant to a sex balancing among four children also. However, Bangladeshi mothers were ready to reach a balance of four sons and four daughter as the last tolerable limit (Haroon 1995). This has been explained by

Data Analysis and Findings
In this chapter, we have analysed the data and put up the findings under seven sub-sections (1) Demographic characteristics (2)

2
Bangladesh Jomnnl o f Public Admiizistl-(ition Satisfying combination of the sex of the children (3) Opinion regarding present family size (4) Fertility behaviour of the civil servants (5) Sex-preference of the children (6) Sex-preference as revealed from the opinion regarding factors inducing larger family (7) Conclusion.

Demographic characteristics
The total number of respondents are 43. All of them are married. All respondents are deputy secretaries to the Government of Bangladesh. The mean age of civil servants under study is 46.46 years and the mean age of their spouses is 39.96. The mean paritylmean family size is 2.30 i.e on an average the family size of the civil servants is greater than 2. On the otherhand, the mean desired family size of the civil servants is 2.37. The most favoured family size is two-child family (65.1%) but still then about one-third of the civil servants (32.5%) favours more than two-child family. Civil servants prefer neither childless family nor a family greater than 4 ( Table-2 and Table-3).  Civil servants with higher parity of three favoured both two child family (42.8%) and three child family (35.7%). It is to some extent the ellect of the motivational programme of the government recommending a "Two-child family" and to some extent the effect of education and social influence. But which part is in force has not been ascertained in this study. The mean desired family size as influenced by their number of surviving children has been increased to 2.42 from 2.37. It is surprisingly observed that one-child family is the least favoured family size. Only one respondent out of 43 preferred onechild family. On the other hand, exactly one-third of civil servants having a parity of 4; prefer four-child family. This happens because of defensive behaviour of the civil servants with higher parity. One good sign is noted that civil servants with higher parity of more than two prefer smaller family to present size. This can be attributable to the realisation and awareness built in the outlook towards small family norms due to their education standard and social status (

Mean
The most favoured corr~bination of the family is one son and one daughter (65.1%). The second highly satisfying combination is 2 sons and 2 daughters (18.6%). It implies that civil servants are more likely to make a balance of sex in the family. But whenever the question of sex balancing is withheld, then obviously son-preference dominates. Two sons with one daughter is highly preferable to one son and two daughters' family. (Table-5).

2.42
Sex-Preference for Children and Fertility Behaviour ... A Saniple Survey 55 Table-5: Table showing distribution of civil servants according to satisfying combination of children.

Opinion regarding present family size
About half of the respondents regard their present family size is equal to desired family size (51.2%). The rest half of the respondents is bifurcated into almost equally responding groups; 23.2% more than desired and 25.6% less than desired family size respectively.  (1 00%) be equal to expected has a parity of three. This signifies that civil servants want their family size at least equal to two child family. This is explained in Table-'.

Fertility behaviour of the civil .servants
It is surprisingly observed that even one fourth of the civil servants do not practice any contraception (25.6%). 74.4% of the total respondents use contraception but only an insignificant percentage (3.2%) of the acceptors accepted permanent methods. 96.8% of the respondents are currently using temporary contraception. Why almost all civil servants are using temporary methods of contraception was not verified in this study since this is not directly related to the subject under study.  When the respondents who regard their family size to be less than expected family size were asked to opine about the sex-seleclivily of the additional child, it is observed that 45.5% were reluctant about the sex of the additional child. Rather daughter preference is observed as 36.3% of them wanted the next child to be a daughter.

11
(1 00%) present first born child. Although, it is observed that more than half of respondents who are reluctant about the sex of the first born child, 72.7% of them had their present first born child as son. It implies that civil servants tended to hide the actual inducement of son-preference as because they were satisfied with the son already have. Only 27.3% of the group voting for "either sex" had their present first born child as daughter (See Table 11 and fable-12).  were liberal enough to .vote for "either sex". On the other hand, respondents who have their first child as daughters favoured 'either sex' appears to be elusive. Sex-preference is dominating among the civil servants as it has been observed by their opinion when verified with the sex of the present first born child 3.6 Sex preference as revealed from the opinion regarding the factors inducing lai-ge family Respondents when asked to express their opinion regarding the factors that contribute to desire more children, son-preference dominated the picture. Bnnglodesli Journal o f h~b l i c Administration About 55.8% of the civil servant believe thai absence of desired number of sons contribute highly as number one cause for desiring more children. This is followed by insecurity of survival of the present siblings (32.5%). The first factor reflects the strong son preference prevailing in the society while the second factor discloses the tendency of the parents to resort to insurance mechanism. The incidence of high mortality greatly contribute to the insurance mechanism of the family planning. As second most contributing factor i,e absence of desired number of daughters ranks high 29.7%. This information seems to have some spurious character that about 30.2% of the civil servants do not have daughters so as a natural tendency to have daughters in the family they feel that absence of required number of daughter may contribute to increase family size Table 14.. below explains the situation.

Sex of the children Only son
Only daughter Both Total From our data analysis, it has been observed that even the civil servants being the members of the upper stratum of the society favour a family size which is greater than 2. (Mean desired size is 2.37). The motivational policy of the government needs to be continued with more seriousness. Further more, it is revealed that son preference dominates the selection of sex of the children when balancing of sex in the family is withheld or controlled. More striking feature of the sex-preference is that when civil servants are blessed with first child as son only then a greater portion of them showed liberal attitude towards the sex-selectivity of additional child. Still more than half of the civil servants believes that birth of required number of sons in the family contribute to desire for more children in the family.
As compared with the greater part of the society, the similarities are apparent. Two-child family norm is yet to be achieved by the civil servants More striking similarity is found in case of balancing of sex in the family Like greater part by the Bangladeshi society, civil servants also enjoy the sex-balancing situation in the family. With respect to sex-balance in the family, they are ready to tolerate two sons and two daughters making family size equal to 4.
In fine, it can be safely remarked that our civil servants are not totally emancipated from the evils of sex-dominance prevailing in the greater part of the Bangladeshi Society. Although some liberal attitudes have been observed with regard to sex composition and reluctance to sexselectivity; these improvements are too volatile to withstand any policy change of the government. If the government ,in a near future declares a policy of 'one-child family, this will invite complications in the reproductive behaviour. A threshold of economic development along with the improvement in female status and increasing female work participation in the main stream of life can bring positive changes in the attitude of the people even in the upper stratum of the society. Until then, any policy change in the family planning area' affecting fertility behaviour is not recommended. Let us hope for the future when conducive environment bringing positive changes in the