Any prospects of fertility recovery in low-fertility societies?

Experience of death of a child may have significant influence on the fertility intention and consequently on the contraceptive behavior of the married women. This paper evaluates the impact of child loss on the fertility and contraception among married women in Bangladesh using the data (N=10146) from Bangladesh Demographic and Health Survey 2007. About 14% of the married women experienced a loss of at least one son while 12.8% lost their daughters during their reproductive life span. Among the currently married women 10% had no children, 19.6% had one child and the rest had two or more children. About 55% of the respondents were currently using contraceptive methods. Poisson regression models revealed that the relative risk of having children for the loss of a son was 1.168, which was 1.178 for the loss of a daughter. Binary logistic regression models suggested that respondents were 0.704 times and 0.872 times significantly less likely to use contraception for the loss of a son and a daughter respectively than their counterparts. This study recommends that reproductive health programs and child immunization programs should be strengthened to reduce the child mortality. Proper counseling is also necessary to minimize the shock due to child loss.

High spousal age difference shows different reproductive outcomes, especially among the married adolescent girls. This is because older husbands use to take reproductive decisions alone and try to compensate for delayed marriage. This study hypothesizes that adolescents with comparatively older husbands tend to get more children and have children earlier than their counterparts. Bangladesh DHS 2007 was used for the in-depth analyses. About 43% of the adolescents were married before the legal age (18 years) at marriage for girls in Bangladesh. About 22.3% of the adolescents had spousal age difference 5 years or less. Almost half (47%) of the married adolescents have their first child within one year of marriage. Poisson regression model suggested that the relative risk of having children among adolescent with spousal age difference 5 years or less was 0.554 compared to the adolescents with spousal age difference more than 15 years. The relative risk increased to 0.730 for spousal age difference between 6-10 years. A binary logistic regression model suggested that adolescents having spousal age difference 11-15 years were 2.45 times significantly more likely to have their first child within one year of their marriage compared to adolescents having spousal age difference 5 years or less.

This paper examines changes over time in the proportion of women's non-numeric responses to questions about desired family size. Such responses have often been interpreted through the lens of classical demographic transition theory, as an indication that reproduction has not yet entered into the "calculus of conscious choice" (Coale 1973:65). Yet non-numeric desired family size has rarely been investigated in a cross-national framework, and never across time. Thus we know little about the processes underlying changes in these responses. This study uses over 15 years of DHS data from 33 countries representing three world regions. Taking a multi-level modeling approach, we use country- and individual-level indicators to examine the factors associated with non-numeric IFS. We then examine how the relationship between non-numeric IFS and individual- and country-level predictors changes over time. Results suggest that education and knowledge of contraception have the most salient associations with non-numeric IFS; with both being negatively associated with this type of response. While the overall effect of education remains consistently strong over time, we find evidence to suggest the the association between non-numeric responses and knowledge of modern contraception decreases over time.