Verified by 
Physically more attractive parents are more likely to have daughters than physically less attractive parents, both in the United States and in the United Kingdom.
In an earlier post, I explain that the logic of the generalized Trivers-Willard hypothesis (gTWH) leads to the prediction that physically more attractive parents are more likely to have daughters than physically less attractive parents.
The gTWH proposes that parents who possess any heritable trait which increases the female reproductive success more than the male reproductive success are more likely to have daughters.
Physical attractiveness, while advantageous for both boys and girls, is even more beneficial for girls than for boys. Men prefer beautiful women for both long-term and short-term mating, whereas women prefer handsome men only for short-term mating (casual affairs and one-night stands), not for long-term mating, for which other traits, such as wealth and status, become more important.
Thus, the gTWH predicts that physically more attractive parents are more likely to have daughters than physically less attractive parents, and, as I explain in the earlier post, the analysis of the National Longitudinal Study of Adolescent Health (Add Health) confirms the prediction.
Both the “sexy son” hypothesis and the good-gene sexual selection theory posit that physically attractive men can increase their reproductive success, not by forming long-term pair-bonded relationships (“marriages”) in which to raise and invest in children, but by having a large number of extrapair copulations with otherwise mated women and cuckolding their mates. So should more attractive parents have more sons instead? Can handsome sons achieve higher reproductive success than beautiful daughters?
Given that the probability of conception per coital act is estimated to be about .03, a man must have 33 extrapair copulation partners (with whom he has sex once each) in order to be able to expect to produce one child (number of potential conception = .99). A man can produce roughly the same number of children with one sexual partner with whom he has regular sex (twice a week) (number of potential conception = .96). It would be very difficult for a man to have more than 30 extra-pair copulation partners in a year, especially in the ancestral environment, where our ancestors lived in a small band of about 150 genetically related individuals (men, women, and children). It would, therefore, be nearly impossible for a physically attractive man to match the reproductive success of a physically attractive woman through only short-term mating. Hence physical attractiveness is more beneficial to girls than to boys.
The analysis of the National Child Development Study (NCDS) in the United Kingdom, which has data on the respondents’ completed fertility at age 47 (virtually all men and women complete their lifetime reproduction by age 45), replicates the earlier findings from the Add Health data in the United States and show that physically more attractive parents are indeed more likely to have daughters than physically less attractive parents.
Physical attractiveness of the NCDS respondents is measured at age 7 by their teachers, who choose up to three adjectives from a highly eclectic list of five to describe the children physically: “attractive,” “unattractive,” “looks underfed,” “abnormal feature,” and “scruffy & dirty.” The child is coded as attractive if it is described at all as “attractive,” and it is coded as “unattractive” if it is described at all as “unattractive.” Then, the sex of the respondent’s first child is measured 40 years later, at age 47.
As you can see in the following graph, British children who are described by their teachers as “attractive” at age 7 are less likely to have a son as their first child 40 years later than those who are not so described. The proportion of sons among the “attractive” NCDS respondents is .50491, whereas the same proportion among everyone else is .52029.
The following graph shows that British children who are described by their teachers as “unattractive” at age 7 are more likely to have a son as their first child 40 years later than those who are not so described. The proportion of sons among the “unattractive” NCDS respondents is .52320, whereas the same proportion among everyone else is .50518.
Multiple binary logistic regression analysis shows that being physically attractive statistically significantly increases the odds of having a daughter as the first child, net of sex, age at first child, education, social class, earnings, height, and weight. Being physically attractive at age 7 increases the odds of having a daughter by 23 percent or decreases the odds of having a son by 19 percent. Similarly, net of the same control variables, being physically unattractive at age 7 decreases the odds of having a daughter by 20 percent or increases the odds of having a son by 25 percent.
The hypothetical average attractive NCDS respondent (who has sample mean values on all of the control variables included in the regression equation) has a probability of having a daughter of p = .50127. In contrast, the hypothetical average unattractive NCDS respondent has a probability of having a daughter of p = .56285. It appears that natural selection does help individual genes to spread, by subtly biasing the offspring sex ratio so that beautiful people, who can benefit from having a daughter, do indeed have slightly more daughters than ugly people, who cannot so benefit.
