This paper proposes a new explanation for religious differences in fertility in India by incorporating the issue of gender bias into the debate. It reports the results from an econometric investigation of the factors influencing the sex ratio at birth and among currently living children, by religion and by caste, for a sample of over 10,000 women. The investigation paid particular attention to religion and caste by subdividing the sample into Hindu, Muslim and dalit women who had all terminated their fertility. It enquired whether the effect of different variables on the sex ratio varied according to the religion and caste of the women. The econometric analysis found that a husband being literate served to raise the sex ratio ? both at birth and of currently living children ? but that the effect of husbands? literacy was stronger for Muslims and dalits than it was for Hindus. In other words, while the illiteracy of husbands exacerbated ?son preference? (and its obverse, ?daughter aversion?), the preference for sons (and the aversion to daughters) exercised a stronger hold on Hindu families than it did on Muslim and dalit families.
