The historical demography of Japan to 1700suffered from recurrent epidemics striking a populace with little or no immunity, killing off
massive numbers, and then burning itself out, only to reappear and wreak the same havoc a gen-
eration later. In McNeill’s words, these outbreaks of pestilence “must have cut repeatedly and
heavily into Japanese population, and held back the economic and cultural development of the
islands in a drastic fashion.”^20
Utilizing the “McNeill thesis,” I uncovered evidence of the massive smallpox epidemic of
735–737. Narrative sources portrayed the plague as truly horrific, a catastrophe unknown in
Japan for several centuries. The Nara court even issued two long orders explaining to the popu-
lace how to care for itself during the outbreak. Most importantly, I used data obtained from
records on rice loans to infer mortality over a wide swath of Japan: 25–33 percent for the three
years. On the basis of this and other evidence, I lent my support to the McNeill thesis. The eighth
century and the figure of six million or so seemed to represent the end of the long growth cycle
that had begun with the Yayoi period. Subsequent archaeological and other evidence has bol-
stered this conclusion.^21
Disease was not the only cause for the eventual halt to population growth by the late eighth
century. Examination of famine records suggested that the commoner populace was chronically
malnourished, suffering from a widespread famine once every three years.^22 Finally, during the
eighth century, the Nara court financed the construction of as many as five massive capitals and
uncounted aristocratic mansions and Buddhist temples. All this building apparently took a large
toll on the lush forests of the Kinai region (Kyoto–Osaka–Nara). Tiles and timbers were recycled
in the late 700s to raise the palaces at Nagaoka and Heian, and tile- makers cut their wood from a
poor secondary forest cover for fuel. Erosion followed.^23
I also made another lasting contribution to Nara demography. The Shōsōin collection con-
tained numerous raw population data that included the sex and age of village inhabitants. Under
the expert guidance of Kobayashi Kazumasa of Kyoto University, I compiled the data into popu-
lation age pyramids by sex. I then submitted them to a comparison to various population pyr-
amids from around the world.^24
Remarkably for census data gathered in the early 700s, four of the comparisons yielded cred-
ible results for fertility, mortality, life expectancy, and infant deaths to age five. One result is
summarized in Table 16.1.
While the achievement was considerable, scholars may question the method. Although the
Nara population pyramids fit most closely into “Model West” containing modern Japanese age
and sex data, the difference between one outcome and another slightly different one was small.
Then, too, I had had to hold both the life expectancy and the growth rate—the two parameters
in the population tables—as unknowns. Finally, the raw data had required some mathematical
smoothing to achieve any result at all.
Yet the story told by these statistics and the others was unsurprising. Both fertility and mor-
tality were very high; the birth rate suggested that women were having children almost all the
Table 16.1 Vital statistics for Mino province in 702
Birth rate: 51.21 persons/1,000
Death rate: 40.21 persons/1,000
Growth rate: 11 persons/1,000
Life expectancy at birth: 27.75 years
Infant mortality to age 5: 53.39 percent
Average age at death over age 5: 41.56
Source: Sho ̄so ̄in monjo, in Dai Nihon Komonjo, vol. 1. Tokyo: To ̄kyo ̄ daigaku shuppankai, 1901.