Working Paper No. 2006 | 09
Trivers-Willard at Birth and One Year: Evidence from U.S. Natality Data 1983-2001
by Douglas Almond (Economics) and Lena Edlund (Economics)
Trivers and Willard [1973] (henceforth TW) hypothesized that evolution would favor deviations from the population sex ratio to vary with maternal condition: mothers in good condition would have more sons and mothers in poor condition would have more daughters. The relevance of TW for the human population is still debated, as are the proximate causes at play. The authors analyze linked birth-death micro data for the U.S. covering 48 million births and 310 thousand deaths for 1983-2001. They find that: (i) the sex ratio varies with maternal marital status, age and education, consistent with the TW hypothesis; (ii) infant deaths reinforce the TW pattern found for the sex ratio at birth; and (iii) this tendency is particularly strong for the post-neonatal mortality (28-365 days from birth). While inherent male weakness may be one mechanism through which a TW effect is obtained, we also find that mothers older than 35 years are more likely to have a girl die than younger mothers. This pattern is noteworthy since advanced maternal age is a biological risk factor, suggesting a role for socioeconomic factors in excess male infant mortality.
Download 
