Socioeconomic status and sex ratios at birth in contemporary Sweden. No evidence for a TriversWillard effect using longitudinal register data across a wide range of status indicators.

Based on evolutionary theory, Trivers and Willard (TW) predicted that high-status parents display male-biased and low-status parents female-biased parental investment and sex ratios. In previous research we demonstrated that the TW effect does not show in parental gender preferences as expressed in differential fertility stopping (Kolk & Schnettler 2012) and likely also not in parental investment (Schnettler 2010). Research on the TW effect in sex ratios at birth remains mixed, however. Part of the reason for mixed results is that generally, the expected effect is small in human populations. Here we look at how sensitive the TW effect is to different status indicators and what role female employment status plays in mediating the TW effect. We use logistic regression on micro level data to examine the effect of socioeconomic status on the sex ratio at birth. We use individual data on births and individuals from Swedish administrative register data (N≈ 2 000 000 births). A wide arrange of socioeconomic traits are analyzed including income, wealth, occupational class and educational level. Early analysis reveals an ambiguous pattern that is contrary to the Trivers-Williard effect.