Re-examining regional disparities in stature during industrializing Japan: Evidence from the 1878-1917 birth cohorts

This study re-examines secular height trends among 20-year-old Japanese males from the 1878-1917 birth cohorts using conscription examination records. The analysis compares urban centers, such as Tokyo and Osaka, with prefectures exhibiting higher initial average heights (e.g., Iwate and Tottori, ≥ 158.0 cm) and those with lower statures (e.g., Niigata, Saitama, and Kagawa, 155.5-155.7 cm). During the early stages of the Matsukata Deflation (1882-1883), despite a sharp decrease in the number of participants, height gains compared with the previous year (Δheight_it) were observed: +0.7 cm in Kagawa, +0.4 cm in Saitama, and +0.3 cm in Tokyo. Similar trends, a reduction in participants alongside a rise in average height, occurred during the panic of 1890-1891. Furthermore, significant increase in height followed the 1879 cholera epidemic in Tokyo (+1.8 cm) and Osaka (+0.4 cm). By the 1910s, growth in these urban centers substantially exceeded the national average, despite high infant mortality rates. Conversely, the consistently shorter stature in Saitama was attributed to harsh labor conditions in wet-rice cultivation and sericulture. The first-difference regression coefficient of Δheight_it for rural-born conscripts on the log-transformed difference in the number of participants, Δln(n_it), is -0.4740 (p < 0.05) for the 1878-1898 birth cohorts, whereas it reverses to 0.9122 (p < 0.05) for the 1899-1917 birth cohorts. This reversal indicates that selection effect operated strongly in the first half of the period, while scarring effects became more dominant in the second half.