Changing Pattern of Fertility in Bangladesh

Abstract

In this study, an attempt has been made to assess the factors associated with changing pattern of fertility in Bangladesh using nationally representative data from Bangladesh Demographic and Health Survey (BOHS), 2007. Multivariate technique named Logistic regression analysis has been used to find out the effects of the selected demographic and socio-economic factors on fertility pattern. Geometric distribution, Beta Geometric distribution and also multivariate technique named Cox's proportional hazard regression analysis have been used to identify the relationship between fertility, conception wait and fecundability. Fertility is still high in Bangladesh, though it has been declining over time. A major cause of declining fertility has been the steady increase in contraceptive use over the last 32 years; another major cause of declining fertility has been the steady increase the age at marriage. Current contraceptive prevalence rate (CPR) is 56% in 2007 BOHS (Mitra et, al., May 2009). The result of Logistic regression analysis shows that several socio-economic and demographic factors significantly affect on fertility. These are age at first marriage, current age of respondent, place of residence, religion, region, respondent's educational level, partner's educational level, work status of women, partner's occupation, contraceptive use, spousal age difference, marital duration, wealth index, body mass index, mass media contact, partner's age etc. From the result of logistic regression analysis, we observed that lower age at marriage giving higher fertility on the other hand higher educated women giving lower fertility. From place of residence we observed that fertility is higher in rural areas. There are several reasons, these include may be the rural women are less educated than urban women; rural women have less media connection etc. Regional difference reveals that fertility is higher in Rajshahi and lower in Sylhet Division. Barisal, Chittagong, Dhaka and Khulna division have intermediate levels of fertility. Religion has affect on fertility behavior through Muslims and Non-Muslims. The analysis shows that fertility among Muslims is higher as compared with Non-Muslims in each age group. Work status of women suggests that labor force participation may be consequence of lower fertility than non-working counterpart. Women who are involved with any service are not dependent on men (husbands), both socially and mentally have their own rights and absence of dependence, men cannot forcibly use women to increase their fertility. This has resulted in lower fertility. Fecundability is regarded as one of the important proximate parameters of fertility performance of the married women. Due to the complex nature of fecundability, we have attempted in this study to estimate mean fecundability from the first conception interval, which is not associated with postpartum infecundability. The first conception intervals have been estimated indirectly by utilizing the data. Since the cohort of women is not homogenous in regards to reproductive performance, we have attempted to estimate the mean recognizable effective fecundability by fitting the Pearson Type-I beta geometric model with parameters a and b to the observed distribution of first conception delay in addition to geometric distribution. In our analysis, we have estimated the parameters by the method of moments. The purpose of the present study, to estimate the mean conception delay, mean and corresponding variance of fecundability and levels, trends and differentials of fecundability of the Bangladeshi women. The mean conception delay of the Bangladeshi women has been found 23.88 months after their first marriage and the mean fecundability is 0.042, which is estimated by geometric distribution. The theoretical arithmetic and harmonic mean fecundabilities are found 0.045 and 0.042 respectively by fitting Beta geometric distribution. This study reveals that the women with higher education have lower mean conception delay and higher mean fecundability. We have also seen from this study that age at first marriage has negative relation with conception wait and positive relation with fecundabihity. It is observed that conception wait is decreasing and level of fecundability is increasing with the increasing age at first marriage whatever be the marital duration. Moreover, the fecundability decreases with the increasing marital duration whatever be the ages at first marriage. This indicates that the more the age at first marriage the higher the fecundability level and less the conception wait and vice-versa. Furthermore, the more the marital duration the less the fecundability and higher the conception wait and vice-versa. We get the significant regression coefficient between age at first marriage and conception wait as -1.849, which reflects that with the increase o 1.85 months. The trend analysis shows that conception wait is lower consequently fecundability is higher in the recent past than at some distant point of time. The multivariate analysis through the Cox's proportional Hazard Regression model shows that the respondent age at first marriage, current age of respondent, place of residence, religion, region, respondent's educational level, partner's educational level, work status of women, partner's occupation, contraceptive use, spousal age difference, marital duration, wealth index, body mass index, mass media contact and partner's age are found to have statistically significant association with the marriage to first conception wait. From the result of multivariate analysis we conclude that the associated factors which affect fertility those factors also affect fecundability with the same direction. f age at first marriage by one year, conception wait tends to decrease By 1.85 months. The trend analysis shows that conception wait is lower consequently fecundability is higher in the recent past than at some distant point of time. The multivariate analysis through the Cox's proportional Hazard Regression model shows that the respondent age at first marriage, current age of respondent, place of residence, religion, region, respondent's educational level, partner's educational level, work status of women, partner's occupation, contraceptive use, spousal age difference, marital duration, wealth index, body mass index, mass media contact and partner's age are found to have statistically significant association with the marriage to first conception wait.