A new analysis suggests that the number of "missing women" worldwide—females absent from populations due to sex-selective abortion or neglect—may be significantly underestimated because standard demographic models ignore how a mother's chronic stress affects the sex of her baby at birth. For decades, demographers assumed the natural sex ratio at birth is nearly constant across all human populations: roughly 105 boys for every 100 girls. A separate biological literature, however, shows that maternal stress—from poverty, war, or discrimination—can cause male fetuses to be lost at higher rates, which should push sex ratios below 105. The contradiction has never been resolved. This project argues that previous studies failed to account for unmeasured, stable differences between mothers. By developing a new statistical method that controls for these hidden factors, the researcher has preliminary evidence that maternal stress does shift sex ratios after all. If confirmed across multiple countries, this work would overturn a 50-year scientific consensus. It would also produce more accurate country-level estimates of missing women—a figure currently used by the UN and NGOs to track gender-based discrimination. The research is primarily fundamental social science, but its findings could directly inform policies on maternal health, prenatal care, and gender equality monitoring.
View original technical description
Explaining variation in sex-ratios at birth (SRBs) has been a fascination of social scientists since the 1600s. It is also a priority for policy makers, as changes in SRBs often indicate sex-selective abortion and discrimination against women and girls. Over the past 50 years, a general consensus among demographers has emerged, that sex ratios at birth are mostly constant at 105 male births per 100 females across groups, with only small variations by race or ethnicity. Recently however, a new biological literature suggests that maternal condition may induce male conceptions to be lost at higher rates. These two literatures are in direct contradiction to each other, as the biological literature suggests that chronic indicators of maternal stress should be associated with heterogeneous SRBs, which is generally not found in large population datasets. This project proposes a novel explanation for this contradiction. Specifically, the PI documents that all previous analyses fail to properly control for time-invariant heterogeneity in maternal condition, and develops a new methodology to do so. Once implemented, the PI demonstrates that this new methodology confirms that maternal condition changes of SRBs, challenging the scientific consensus on the constancy of SRBs and reconciling the contradictory findings of these literatures. This project has three main goals. In Project 1, the PI seeks to investigate these preliminary findings across various countries and settings, thereby establishing a new scientific consensus on the effect of socioeconomic status on SRBs. Project 2 utilizes the new methodology to refine global estimates of the number of missing women by providing new country-level estimates. In Project 3, we seek to reconcile various long-standing puzzles in the literature on SRBs, socioeconomic factors, and maternal stress, including changing sex ratios during and after wars, economic downturns, and across racial groups.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know