Regional Variations and Determinants of Fertility in Bangladesh.

Abstract

This study examines the regional variations and determinants of fertility in Bangladesh. Information from several in-depth surveys conducted in four rural areas during 1982-86 is utilized in this study. The four study areas are divided into two regions: Sirajgonj and Abhoynagar. The results suggest that the Sirajgonj region has higher fertility levels than the Abhoynagar region and that the socioeconomic characteristics of the sample households vary between the regions. For example, the literacy rate is much higher in the Abhoynagar than in the Sirajgonj region. Furthermore, the religious composition of the study population also varies between the regions: approximately one-fourth of the study population was non-Muslim in Abhoynagar region, compared to only 3 percent in Sirajgonj region. The estimated natural fertility was approximately the same in each area. Breastfeeding was universal, and the duration of breastfeeding was over 2 years in each study area. The duration of postpartum abstinence was substantially shorter than postpartum amenorrhea, and the two religious groups observed different postpartum abstinence periods. The contraceptive use rate was low, and traditional contraceptive methods were practiced by one-third of all acceptors in the Sirajgonj region. In the Abhoynagar region, the contraceptive use rate was moderate and female sterilization contributed approximately 40 percent of the contraceptive use rate. In both regions, non-Muslims were practicing contraception at a higher rate than Muslims. Application of Bongaarts' proximate determinants framework suggests that the three principal proximate determinants do not entirely explain the observed levels of fertility in the Abhoynagar region. However, the discrepancies become smaller with the assumption of a higher but unreported induced abortion rate in that region. An extension of the proximate determinants framework at the individual level suggests that factors affecting fertility differ between regions. In addition, fertility in the three years after the 1983 in-depth survey is largely determined by whether the respondents were practicing contraception at the time of the in-depth survey. The factors influencing contraceptive use and fertility are investigated by path analysis. The results suggest that the differences in socioeconomic status, religion, and family planning program inputs explain regional variations in contraceptive use and fertility in the study areas.