August 2008 - Volume 2, Issue 4

Demand for Child in Rajshahi City, Bangladesh: A Multivariate Analysis

Authors: Shamima Akter1, Md. Mizanur Rahman2, Mahmudul Hasan3 and Mst. Papia Sultana4 1. Shamima Akter, M. Phil Fellow Department of Population Science and Human Resource Development, University of Rajshahi, Bangladesh E-mail: samimarub@yahoo.com 2. Md. Mizanur Rahman, Lecturer Department of Population Science and Human Resource Development University of Rajshahi, Bangladesh E-mail: mizanur_rub@yahoo.com 3. Mahmudul Hasan, Assistant Professor Department of Population Science and Human Resource Development, University of Rajshahi , Bangladesh E-mail: mahmud_ru@hotmail.com 4. Mst. Papia Sultana2, Lecturer Department of Statistics University of Rajshahi, Bangladesh E-mail: papiasul75@yahoo.com Correspondance: Shamima Akter Department of Population Science and Human Resource Development University of Rajshahi, Rajshahi-6205 Bangladesh E-mail: samimarub@yahoo.com

ABSTRACT The main aim of the present study is to estimate the factors associated with demand of child in Rajshahi City of Bangladesh. In this regard the necessary data was collected in June 2004 under the project of UNFPA entitled "Strengthening the Department of Population Science and Human Resource Development", University of Rajshahi, Bangladesh. The numbers of reproductive aged couples who have at least one child were selected for interview during the study. By using some statistical tools such as contingency analysis with - test and logistic regression model shows that socio-demographic factors significantly associated with the demand of child. The result of the study open that after controlling other factors, the socio-demographic characteristics that statistically and significantly affect the probability of demand of child in the study location like current age of respondent, age at marriage, respondent education, husband occupation, number of male child, number of female child and visit of health worker. Here current age and numbers of male and female children are statistically highly significantly influencing the demand of children. Key Words and Phrases: Demand for Child, Chi-square, Cross-tabulation, Logistic regression model, Odds ratios.

INTRODUCTION

In a developing country like Bangladesh it is almost compulsory that every boys and girls must marry and parent hood is a common feature immediately after marriage. Bachelor hood and those married couple who cannot procreate child are socially disapproved. It is believed that those who do not have any children must have some sort of physical defect, and they treat as sterile couple. Thus social status is also an important reason for demand of children (A. Bhende and Kanitkar, 1997). Demand for a child is very much related to age, and it is also somewhat linked with health, the no. of times she has been giving birth and so on. Also in cases of dead baby, how many it can be, the demand cannot be stopped. Such many reasons may influence the demand (Hans Raj, 2000).

Education may affect fertility through rising age at marriage, giving alternative source of new normative orientations and expansion of vision, increasing women's knowledge and use of birth control as well as providing better opportunity for labour force participation (Ryder, 1967; Janowitz, 1976). Husband's education may have a significant effect on the children ever born to ever married women. Because in our society almost everything depends on the opinion of a husband. Now a days higher educated person are giving importance to the educated female for marriage, as a result a balance is prevailing upon the families, which play a negative role in fertility (Ali M. 2003). In many studies it is observed that education especially secondary or higher level is important for reducing fertility, infant and child mortality, improving the human capital of the population (S. K. Sarkar, 2004, P 146). Also education along with improved family planning services can help women avoid unintended pregnancies and the abortions that sometimes follow them. Therefore, social economic or other changes drastically affect the value or cost of children and reduce fertility rates within education groups.
In developing countries like Bangladesh, the husband's occupation is closely related to fertility. According to Prof. Donald J Bogu's view fertility is influenced by the occupation of the head of the family. The people with good education are likely to check fertility, where as those with manual occupation are likely to have more children. He has related this to income also. According to him where income is low, fertility goes up, but income is not very high, then the fertility is the lowest, but when income considerably increases with that fertility again goes up. According to opinion it is wrong to think that when the family is riches the number of children will be less (Bogue, 1969).

Studies concerning the impact of labour force participation on fertility suggest that working women have lower fertility than their non-working counterparts (Devanzo, 1972; UN, 1973). It also argued that labour force participation would have a decreasing effect on fertility only if it is incompatible with child bearing. The direction is of causality between labour force participation and fertility is not certain, because labour force participation may be a consequence as well as a cause of lower fertility.

Bangladesh has undergone a remarkable demographic transition over the last two decades. The total fertility rate has declined from about 6.3 in 1975 to 3.3 in 1999-2000 (BDHS, 1999-2000). Striking change have been observed were asked how many children they would ideally like to have, the response was an average 4.1 children (Huq and Cleand,1990). The desire for additional children declined noticeably over the past few decades. For example the percentage of women with two children who want no more children has risen from only 48 percent in 1991 to 66 percent in 1999-2000. Also in 1991, 45% of married women with two children wanted to have another child in future in the 1999-2000 BDHS surveys (Mitra et al 2001). Data from the BDHS indicate that after almost a decade long stagnation, the Bangladesh fertility rate has declined slightly to 3.0 per women (BDHS, 2004).

After marriage every couple is to desire have children. It is an inner feeling and source of satisfaction to every married person to have children. Thus in every society it is very much essential to find out demand of children to couple, which influence the whole family structure on the one hand and the society on the other. Therefore the main objective in this paper is to focus the background characteristics of the respondent and also to isolate the factors which are influencing the demand for child.

CONCEPT OF DESIRE AND DEMAND FOR CHILDREN

When parents expect the "baby" with their family decision as well as with the view of their surrounding economic and social constraints is termed as demand for a child (Schultz, 1972).When parents expect to have their own children according to their mental incitement and care for them emotionally ignoring the economic point of view of bearing and rearing them and without equating cost and benefits involved with children may be termed as desire for children.

DATA AND ANALYTICAL METHODS

The data has been collected from a field survey conducted in the urban area of Rajshahi City. Three wards of Rajshahi City were selected as a representative part and collected information from 2250 women by preparing a pre-scheduled questionnaire during the month of June 2004 and conducted through in-depth interviews. Respondents in the study are reproductive aged couples. The Present Study needs those respondents who had at least one child. Again this study was concentrated only on the urban respondents. Thus we got 2064 respondent who were related to this characteristic through purposive sampling method.

Analytical Methodology

In our study we have mainly used the tabular form of data. To see the association between demand for a child and other cofactors we used the most common contingency analysis with - test procedure. We have also used the logistic regression model to identify the impact of some most influential variables on demand for a child.

- Test
- test is mainly used to test the hypothesis which specifics the nature of one or more distribution. We know the mathematical form of the distribution; hypothesis regarding the sample that has been drowning from the distribution is tested by - statistic. For testing hypothesis, we used to compare observed set frequencies with a corresponding set of frequencies that are expected under null hypothesis. The test statistic is defined as,

Where
O_i= observed frequencies, i = 1, 2, 3 ............., k
E_i= expected frequencies, i = 1, 2, 3 ............., k
n = which is distributed as with (k-p) degree of freedom, where p is the number of independent restrictions imposed for the calculation of the set expected frequencies.

Logistic Regression Model

An interesting method that does not require any distributional assumptions concerning explanatory variables is Cox's linear logistic regression model (1970). That logistic regression model can be used not only to identify risk factors but also to predict the probability of success. The general logistic model express a qualitative dependent variable as a function of several independent variables, both qualitative and quantitative (Fox, 1984). The logistic regression model allows a categorical variable (dichotomous or polytomous variable) as dependent variable. Let Y is a dichotomous dependent variable, which takes values 0 and 1. i.e.

Also let there is a collection of k independent variables, which will be denoted by the vector, and be a (k+1) 1 vector of unknown parameters.
For simplification, we can use the quantity the probability that the event occurs conditional on the value of X. Hence
and

Hence

The central part of logistic regression in a transformation of is known as logit transformation, which is defined in terms of , is as follows:

which is the logit of the multiple logistic regression models.

RESULTS AND DISCUSSION

There are several factors which are responsible in determining demand for a child to couple. Some basic characteristics of the study population that are treated as background variables in our study have been incorporated in Table 1.

Age is an important characteristic of a respondent. Demand for a child is very much related with age and affected by the age variation of ever-married women. From Table 1 we observe that in our study area the highest number of respondent is aged less than 25 years, which is 27.2% and the lowest for ages 40 and above which is 13.4%.

We observed that age of the respondent is highly significantly associated with demand of child. Demand of child is highest position in age group less 25, medium position in age group 25-30 years and it is rapidly decreasing in the age groups 30-40 years and 35-40 years. It also shows that top lowest demand of child in the age group 40 years and above (Figure 1).

Age at the marriage is one of the important factors in demography as it is directly related to fertility in many societies. In Bangladesh the minimum age at first marriage is 18 but in our study area early marriage is most frequent. We observe that 55.3% women's first age at marriage is less than 15 years, 37.6% women's age at first marriage is between 15-20 years, and the most vulnerable sight is that only 7.1% women's age at first marriage is above 20 years.

Fertility may be affected and influenced by the preaching of religions. Those religions which do not put any bar on the number of marriage and children are likely to have more fertility than others. Because more the number of wives normally, there are, more the children. Islam is the predominant religion in Bangladesh as well as in our study area. From Table 1 we see that in this area 98.3% are Muslims and only 1.7% are non-Muslims.

Education is one of the most important factors of fertility. From Table 1 it shows that, in this area maximum respondent are with education level secondary (39.5%), but incase of their husband maximum is higher educated (29.8%). Females with education level illiterate, primary and higher it is 19.9%, 26.3% and 14.3% respectively. Incase of their husband the percentage for illiterate, primary and secondary education level is 14.8%, 19.4%, 36.0% and 29.8% respectively.
In many studies it is observed that demand for children is higher among non-working or house wives as compared to working women (S. K. Sarkar). From Table 1 it is observed that around 93.4% women are house keepers in this area, only 4.7% engaged in service and another 1.9% engaged in other occupation.

It is observed that maximum husbands of the respondent in this area are engaged in service (48.1%), 24.3% engaged in farming and 27.7% engaged in other category (Table1).

Total number of living children as well as number of male and female children of couples can influence the demand for a child and their aspiration to have another child.

From Table 1 it is observed that 19.1% have only one male child, and incase of two, three and more than four children their percentage is 32.1%, 43.6% and 5.2%. In case of female children their percentage is 23.1, 30.6, 36.5 and 9.7 respectively. The Figure 2 shows that few percent of the respondent have one child (both male and female), top most percent of the respondent have three children and lowest percent have four and more children. This means that maximum respondent prefer two-three children, if they satisfy getting according to their preference the demand of children tremendously decreases.

In a less developed country like Bangladesh, health workers are very important person for giving advice and increase knowledge about new methods of health as well as reproductive health. They visit every house and make aware specially women and their babies health as well. From table 1 it shows that in our study area 52% respondents' house regularly visited by health worker, 19.8% irregularly visited and 28.2% houses were out of visit by the health worker.

Access to mass media always plays a vital role to make awareness and consciousness of women. In Bangladesh television is a popular media, which has reached in maximum house of both in rural and urban area. From table 1 we see that in our study area 77.4% respondent watches television and 22.6% do not.

Chi-square Test

Pearson Chi-square test procedure for testing significance among demand for children and socio-demographic variables in Table 2 are as follows-

To see the association between demand for child and various background variables like age, religion, occupation, age at marriage, education, number of male and female children of couple, visit of health worker and watches television a well known procedure Pearson chi-square test is used. From Table 2 it is found that current age of respondent, age at marriage, education of respondent, number of male children, number of female children and visit of health worker is highly significant impact with the variable demand for children. While the variables religion, education of husband, occupation of respondent and her husband and watches television shows insignificant result, that is they have no impact on demand for child to couple.

Factors associated with demand for a child: A Logistic Regression Analysis
Logistic regression analysis shows that the significant variables are the respondents' current age, age at marriage, respondents' level of education, husbands' occupation, number of living male and female children and frequency of visits by health workers. The remaining explanatory variables, namely religion, husbands' education, respondents' occupation and watch television do not seem to have significant independent effects on the demand of child.

* Significance at p<.01, ** Significance at p<.05, *** Significance at p<.10

From Table 3 it shows that age of the respondent is highly significantly related with demand for children to couple. Considering age group <25 years as a reference category we see that, demand for children is less likely as their age increases. All the co-efficient calculated for each age group is negative in sign. The result indicates that the women of age group (25-29), (30-34), (35-39) & 40+ have (1-0.300)*100=70.0%, (1-0.219)*100=78.10%, (1-0.169) =83.1% and (1-0.112)*100=88.80% lower risk to demand for additional children as compared to age groups <25 years.

From Table 3 we observe that the regression coefficient of females and their husband for different levels of education are not statistically significant even except secondary level education for the female and they have positive effect. It is a general convention that demand for a child reduces as level of education raises up (S K Sarkar, 2004). Data from the world fertility surveys and the demographic and health surveys in Bangladesh and worldwide confirm the positive effect of education on reproductive behavior (Schultz, 1994; World Bank, 1994).

Occupation both for females and their husband plays an important role determining demand for a child. From Table 3 we see that in our study area occupation of respondent shows insignificant results but for their husband it shows significant result. The regression co-efficient of husband occupation shows that the risk of demand for children is 73.50% and 74.50% lower for service and other category of occupation as compared to the husband whose current occupation is farming (Table-3).

From Table 3 we see that number of male and female children of couple shows highly significant result to wants more children that is demand for child is very much related with these variables. Considering number of male child for one as a reference category we see that the risk is 43.40%, 78.40% and 89.30% lower for the couple whose number of male children is 2, 3, & 4+ respectively. On the other hand by assuming number of female child for one as a reference category it shows that the risk for demand of more child is 58.2% & 52.70% lower for those couple whose number of present female children is 2 & 3. But for the couple whose number of female children is 4 and above the risk is 1.363 times higher. This may because; aspiration of having a male child for a couple is a common scenario.

From Table 3 it is observed that visit of health worker has significant impact on demand for a child. Considering irregular visit of health worker as a reference category we see that it is 1.59 times higher, where health worker do not at all visit couple's house and it is 22.90% lower where health workers visited regularly.

CONCLUDING REMARKS

The study of demand of children has received increasing attention in developing countries because of its identifying capacity of segment of population having high and low fertility. The constitutional law of Bangladesh strictly prohibits early marriage. However in our study area a large proportion of the respondent is less than 25 years of age and most of them get their first marriage before 15 years. Major respondent in this area are illiterate and engaged in household activities only. Demand of child is largely depend on age and age at marriage of women. The result also indicates that 43.6 percent and 36.5 percent of the respondent in the study area have three male and female children respectively.

The results of the inferential analysis show that all the explanatory variables are statistically and significantly affect the probability of demand for child except religion, husband education level, respondent occupation and watches television of women in the study location. The sex and number of children is also an important factor for demand of children. The couples who have female children continue to go on giving birth to children till a boy is born. In many cases the same is true for the couple who have female children. Visits of health worker can change a couple's opinion to have a child because they increase awareness of good health both about child and their mother.

The results of the study Suggests that in order to accelerate the process of fertility decline in the country greater priority is to development in the social sector including enhancement of women's status, especially through increased female education and employment opportunities and improved access to mass media. In order to increase age at marriage the legislating law of marriage should maintain strictly. The government of Bangladesh should be aimed at operational zing the delivery of the wider reproductive health service package in rural and urban areas of the country.

ACKNOWLEDGEMENT

The author would like to special thanks the funding support from UNFPA to strengthen the department of Population Science and Human Resource Development, University of Rajshahi, Bangladesh. I also thanks to Professor Abul Basher Mian, Director of Institute of Bangladesh Studies, University of Rajshahi, Bangladesh for his necessary directions.

REFERENCES

Ali M. Arshed (2001): "Fertility Patterns and Differentials in Bangladesh"; M. Sc Thesis, Dept. of Statistics, University of Rajshahi; Bangladesh.

Asha, A. Bhende and Tara Kanitkar (1997): "Principles of Population Studies", Himalaya publishing house.

B. B. S.(1994): "Bangladesh Population Census Report, 1991. National Volume, Dhaka".

B. B. S.(2001): "Bangladesh Population Census Report, 2001. National Volume, Dhaka".

Bogue, D. J. (1969): "Principles of Demography". New York: John Wiley and Sons, Inc.

Cleland, J, James Phillips, Sajeda Amin G.M. Kamal (1993): "Fertility Decline in Post-Independence Bangladesh: Evidence from recent Demographic Surveys". Population Council Division Working paper No.54,1993.

Government of Bangladesh (GOB) (1994): "Country Report : Bangladesh International Conference on Population and Development, Cairo, September 5-13,1994. Dhaka: Government of Bangladesh".

Hans Raj (2000): "Fundamentals of Demography" With Special Reference to India, Population Studies, Surjeet Publication, India.

Huq. N and J. Cleland (1990): "Bangladesh Fertility Survey". NIPORT, Azimpur, Dhaka, Bangladesh.

Janowitz, B.S (1976): "An Analysis of the Impact of Education on Family Size". Demography, 13(2).

Mitra,S.N. and Associate (2001): "Bangladesh Demographic and Health Survey 1999-2000". Institute of Population Research and Training (NIPORT), Dhaka, Bangladesh.

Mitra,S.N. and Associate (2005), "Bangladesh Demographic and Health Survey
(BDHS), 2002-2003". National Institute of Population Research and Training (NIPORT), Dhaka, Bangladesh.

Ryder, N. B. (1967): "The Character of Modern Fertility". In W. Peterson (ed) Readings in Population. The Mac Millan Company, N. Y.

Sarkar S. K.(2004): "Demand for a child in Bangladesh: A Multivariate Statistical Analysis". Unpublished Ph. D. Thesis, Department of Statistics, University of Rajshahi, Bangladesh.

Schultz, T. W. (1972): "The value of Children: An Economic perspective", Journal of political economy, Vol. 81, No. 2, 1973.

United Nations (1973): "The Determinant and Consequences of Population Trends", vol.1. Population Studies No. 50. New york.

World Bank (1995): "Bangladesh: From Stabilization to growth. Washington, D. C. The World Bank.