Transient cultural influences on infant mortality: Fire-Horse daughters in Japan

Parental investment theory suggests that the quality and quantity of parental care depends, in part, on assessments of whether offspring will survive and yield grandchildren. Consistent with this theory, we hypothesize that parental perception that a birth cohort will have low reproductive success coincides with higher than expected infant mortality in the cohort. We test this hypothesis in industrialized Japan in 1966 when cultural aversion to females born in the astrological year of the Fire-Horse may have jeopardized the life of female infants.

We applied time-series methods to cohort infant mortality data for Japan, from 1947 to 1976, to test whether female infant mortality in 1966 rose above levels expected from history, male infant mortality, and fertility. Methods control for the secular decline in infant mortality as well as other temporal patterns that could induce spurious associations.

Findings support the hypothesis in that female infant mortality rises by 1.1 deaths per 1,000 live births above expected levels (coefficient = 0.0011; standard error = 0.0005; P = 0.03). The result indicates an excess of 721 female infant deaths statistically attributable to the Fire-Horse year. Findings remain robust to control for male infant mortality and the secular decline in mortality over the test period.

The discovery of a predictable, acute increase in female infant mortality during the Fire-Horse year supports the relevance of parental investment theory to developed countries. Results should encourage further research on the health sequelae of abrupt, population-level shifts in culture.