Kerry MacQuarrie

Methodological Innovation in Studying Abortion in Developing Countries: A ‘‘Narrative’’ Quantitative Survey in Madhya Pradesh, India

This article describes the development and implementation of a mixed methods data collection method designed to provide high-quality data on the circumstances surrounding abortion in Madhya Pradesh, India. Data collection took place between 2000 and 2002, beginning with a qualitative phase and culminating in a large-scale, representative survey. The survey itself melded a unique narrative interviewing technique with quantitative survey techniques and collected information on 11,341 individual pregnancies from 2,444 women, with a 97% response rate. Abortion rates calculated using these data are found to be roughly five times higher than the National Family Health Survey-2, a comparable sample using more traditional interviewing techniques, suggesting that this approach reduces the underreporting of abortion while providing the contextual information often lacking in survey data.

TWO SONS AND A DAUGHTER: SEX COMPOSITION AND WOMEN'S REPRODUCTIVE BEHAVIOUR IN MADHYA PRADESH, INDIA

This article examines how the sex composition of womens current children at the start of a pregnancy interval influences both fertility desires and the full range of reproductive actions women may take to realize them, including temporary contraception, abortion and sterilization, in Madhya Pradesh, India, where popular notions of ideal family size and sex composition are dominated by son preference. The analysis is conducted using a dataset of 9127 individual pregnancy intervals from a 2002 statewide representative survey of 2444 women aged 15-39 with at least one child. The results indicate that womens preferences go beyond a singular preference for male children, with the preferred composition of children being two boys and one girl. Women with this composition are 90% less likely to report having wanted another pregnancy (OR 0.097, p < 0.01) relative to those with two girls. These preferences have significant implications for reproductive actions. While sex composition has no statistically significant effect on the use of temporary contraception, those with the preferred sex composition are twice as likely to attempt abortion (OR 2.436, p < 0.01) and twelve times more likely to be sterilized (OR 12.297, p < 0.01) relative to those with two girls only.