LATE CHOSON PROPOSALS 365
of higher ratios of production on the most fertile land in Korea that would have
yielded a total product six times greater than Europe, or about 45 bushels/acre,
the average production figures that Hishimoto and Hoon K. Lee discovered just
about matched the productivity of the least fertile land, undoubtedly because
the vast majority of land in Korea was at the lowest level of fertility.
Crude though these calculations may be, they call into question much of the
recent Korean literature about improvements in agricultural productivity. There
is little question that the extension of irrigation, transplantation, and double crop-
ping in the south must have improved average yields over what they were at the
beginning of the Choson dynasty, but Hishimoto held Korean production stan-
dards at the end of the dynasty in relatively low esteem and took pride in the
accomplishments of the Japanese colonial administration in raising acreage pro-
ductivity in only eight years (by 1917) to 20 bushels/acre, a figure just slightly
less than the extrapolated estimate for Japanese acreage production around 1880.^29
Pre-Meiji Japan would provide a better comparison for Korean rice produc-
tivity than with the wheat culture of pre-modern Europe, especially since so
much has been written about the economic advances made in Tokugawa Japan
prior to the Meiji Restoration in 1868. Unfortunately, the figures for agricul-
tural productivity have been mired in dispute since James Nakamura rejected
government statistics, claiming they had underreported actual production. He
proposed an extremely high productivity rate of 84.6 bushels per hectare of paddy
or 34.5 bushels/acre for the years 1878-82, a figure 2.2 times greater than Hishi-
moto's estimate for Korea in 1910.^30 Nakamura was attempting to argue that
the pre-modern expansion of agricultural productivity in Japan actually fueled
the industrial development of Japan in the Meiji period by mobilizing savings
for investment in industry. The other two, more modest, estimates for rice yields
in Japan in this period were 25. I bushels/acre (61.4 bushels/ha) and 26.9
bushels/acre (66 bushels/ha), and the last estimate is the conclusion of a more
recent study designed to modify Nakamura's claims." In summary, Hishimoto's
figure of 15.55 bushels/acre for Korean productivity about 1910 was about 58
percent of this most recent revised figure for Japanese productivity for 188o,
and it is possible that the average yield on Korean paddy may have been still
lower in that year. For that matter the estimate of yield in the Kamakura period
in Japan (1191-1333) was 22.5 bushels/acre, still a higher percentage than
Korean production in 1910.32
On the other hand, productivity in Korea in 1910 appears to have about the
same as China in 1400 in the early Ming dynasty. Although estimating the aver-
age yield of rice (unhulled) per acre for the pre-twentieth century era is quite
difficult because the figures either represent yields on the best quality land, or
the quality is not specified. Francesca Bray found citations of productivity for
the period from 1050 to 1700 indicating a range from 3.1 to 21.6 bushels/acre.^33
Medium quality land in Shanghai in 151 I yielded 9.27 bushels/acre, and the
yield in good years in Canton in the late seventeenth century was I 1.8
bushels/acre.^34 If Dwight Perkins's calculations of average yields for China are