Parental Investment, Status, and Child Gender: Some Evidence for the Trivers–Willard Hypothesis from a Survey Experiment

Abstract

This study critically evaluates and empirically tests the Trivers–Willard (TW) hypothesis, which proposes a relationship between parental socioeconomic status and sex: Parents with higher status are expected to be more likely to have male offspring and to preferentially invest in male offspring, whereas parents with lower status are expected to be more likely to have female offspring and to preferentially invest in daughters. Although the TW hypothesis has been explored in terms of offspring sex ratio and parental investment, findings in modern developed societies generally show null results, with notable exceptions in the domain of parental investment in their children’s education. Previous studies have often not explicitly addressed the potential underlying mechanisms of the TW effect. This includes the authors of the original hypothesis (Trivers and Willard 1973), who discussed some potential mechanisms but ultimately left the question of mechanisms unanswered. Building on Matthews’s (2011) proposition to explore psychological underpinnings, this paper posits that the TW effect, if present, may be rooted in general parental preferences. To investigate this, a factorial survey experiment was designed to measure respondents’ preferences in parental investment while minimizing social desirability bias. The study specifically examines the extent to which respondents’ assessments of favorability and fairness in various parental investment scenarios depend on child characteristics believed to influence differential parental behavior. The findings reveal patterns that are somewhat in line with the TW hypothesis but are minor and lack statistical significance. The article concludes by proposing three future research directions aimed at further unraveling the intricacies of the TW effect.