Sex composition and its impact on future childbearing: a longitudinal study from urban Uttar Pradesh

BACKGROUND The sex composition of existing children has been shown to influence childbearing decision-making and behaviors of women and couples. One aspect of this influence is the preference for sons. In India, where son preference is deeply entrenched, research has normally focused on rural areas using cross-sectional data. However, urban areas in India are rapidly changing, with profound implications for childbearing patterns. Yet, evidence on the effect of the sex composition of current children on subsequent childbearing intentions and behavior in urban areas is scant. In this study, we analyze the impact of sex composition of children on subsequent (1) parity progression, (2) contraceptive use, and (3) desire for another child. METHODS We analyze prospective data from women over a four year period in urban Uttar Pradesh using discrete-time event history logistic regression models to analyze parity progression from the first to second parity, second to third parity, and third to fourth parity. We also use logistic regression models to analyze contraceptive use and desire for another child. RESULTS Relative to women with no daughters, women with no sons had significantly higher odds of progressing to the next birth (parity 1 - aOR: 1.31; CI: 1.04-1.66; parity 2 - aOR: 4.65; CI: 3.11-6.93; parity 3 - aOR:3.45; CI: 1.83-6.52), as well as reduced odds of using contraception (parity 2 - aOR:.58; CI: .44-.76; parity 3 - aOR: .58; CI: .35-.98). Relative to women with two or more sons, women with two or more daughters had significantly higher odds of wanting to have another child (parity 1 - aOR: 1.33; CI: 1.06-1.67; parity 2 - aOR: 3.96; CI: 2.45-6.41; parity 3-4.89; CI: 2.22-10.77). CONCLUSIONS Our study demonstrates the pervasiveness of son preference in urban areas of Uttar Pradesh. We discuss these findings for future programmatic strategies to mitigate son preference in urban settings.