On Polygyny and Sex Ratio at Birth: An Evaluation of Whiting's Study

Abstract

A rebuttal is offered to an article by Whiting (1993) who reported that polygynously married women are more likely to produce girls than their monogamous counterparts. He supported his case with data drawn from seven cultural groups in Kenya showing lower sex ratios among the children of polygynously married mothers than among the children of monogamously married mothers. The results were attributed to two possible mechanisms: polygynously married women time their intercourse more closely to the day of ovulation when in conception girls are favored or have less frequent intercourse than monogamously married women. There is corollary evidence that high coital rates increase the sex ratio. While the individual samples used in Whitings analysis ranged from 110 Kipsigis to 1340 Kikuyu children the overall effect was significant (monogamous 0.53 vs. polygynous 0.47). One of Whitings samples was drawn from the Kipsigis a Nilotic Kenyan group on the basis of data collected by Robert Daniels on the Itembe Settlement Scheme. The writers own Kipsigis sample in 1989 from Abosi failed to support Whitings result. Polygynously married women had a birth sex ratio of 0.521 and monogamously married mothers one of 0.526 compared with Itembes 0.458 and 0.534. This discrepancy may have been due to the common pattern in the Abosi area of the coresidence of co-wives in a single homestead. This suggests that the pattern Whiting observed in seven Kenyan populations may be contingent on co-wives maintaining separate and distant residences. In the Abosi sample children born to a mans older polygynously married wives showed lower sex ratios 0.475 than those born to his youngest wives or single wife (0.533). The analyses presented here suggest that more precise hypotheses pertaining to co-wives residence patterns relative age and conformity to customary ideals can be formulated that may explain why polygynous status is associated with low sex ratios in some populations and not others.