Coital rates and sex ratios in the South Pacific.

Abstract

A number of anthropologists have recently cited the papers of William H. James to support questionable hypotheses. Their argument is based on 2 premises: 1) sex ratio (proportion male at birth) is positively related to parental coital rate at the time of conception, and 2) in their independent sets of Micronesian data, there was no evidence of a decline in sex ratio with parental age or parity. Underwood infers that her data constitute a serious challenge to the assumed universality of at least certain aspects of Euramerican patterns of marital sexual behavior. Accordingly, she hypothesizes a Micronesian pattern of marital sexuality in which coital rates do not decline appreciably with duration of marriage. If coital rate is a determinant of sex ratio, then it must be a weak one. This conclusion follows from the facts that though coital rates roughly halve across the 1st year of marriage, yet the concomitant decline in sex ratio of conceptions (in the 1st and 12th months of marriage) is only from about 0.522 to about 0.513. It is clear that the decline in sex ratio with maternal age (or with paternal age, birth order, or duration of marriage) in large human populations is very small indeed. For instance, there were 21 million births in England and Wales 1938-68. The sex ratio of those born to married women 20-24 and 40-44 years were 0.5152 +or- 0.0002, and 0.5129 +or- .0006, respectively. Standard power analysis suggests that to stand 8 chances in 10 of detecting a difference of this magnitude in 2 equal-sized samples at the 0.05 level (one-way), the samples would each have to number more than 500,000. So the failure of Brewis and of Underwood to detect such an effect in their comparatively minute samples can scarcely be regarded as grounds for the inference that the effect does not exist in Micronesia.