including cotton, rape seeds, indigo and silk (Hanley and Yamamura, 1977;
Yasuba, 1986).
There is no evidence (as with the case of Mughal India discussed above) to
suggest that feudal dues increased after the middle of the Tokugawa period to
siphon off higher yields to benefit landlords (Yasuba, 1986). The share of GNP
received by the elite was probably declining (Hanley, 1983) and population
was more or less constant over the nineteenth century. Together these factors
imply that increased rice production would raise the net income of the peas-
antry. There was also considerable diversification of employment and output in
the rural non-farm economy with increased production of cotton cloth and
improvements in sericulture in rural areas, especially after the 1830s (Yasuba,
1986). It is reasonable to suppose that standards of living were higher in 1868
than some research has allowed for but the further step to argue that they were
higher than in Western Europe is difficult to sustain.
Some scholars have concluded that the standard of living in Japan in 1850
was indeed higher than in most of Western Europe (Hanley, 1983). These
scholars illustrate a common problem with the use of indirect evidence and
specifically here confuse the relation between income and nutrition. Hanley
points to various pieces of evidence that indicate ordinary Japanese people had
a fairly good diet. In the late 1870s, he notes, most were eating a diet based on
a mixture of unpolished or crudely polished rice with barley, buckwheat and
various millets,which is more nutritious than polished rice. This may be
evidence for good nutrition but it is not evidence for high incomes; indeed it is
common for standards of nutrition to worsen with rising incomes (see Chapter
1),and the incidence of beriberi rose towards the end of the nineteenth century
in Japan as machine-milled rice (a product of a more urban-industrial econ-
omy) was introduced. Wider welfare indicators, such as the fact that buildings
were usually no more than one and a half stories high giving a population
density much below the largest cities in Europe and well-maintained (by local
government) streets, bridges and water supply systems, simply describe a
well-managed pre-industrial city, and are not evidence of rising incomes.
Although Japan (and possibly China) had a high quality of living in the nine-
teenth century,the evidence is consistent with this occurring at lower levels of
material income than in much of Western Europe. Rising incomes in eigh-
teenth- and nineteenth-century Britain were accompanied by the growth of
fetid slums and worsening health conditions among the urban population or
what the English Romantic poet William Blake termed ‘the dark satanic mills’.
Pomeranz states that life expectancy in the late pre-industrial era in China
was almost 40. This, he argues, was broadly equivalent to that in England and
indicates that average incomes must likewise be similar. There are very signif-
icant conceptual and empirical problems with this claim. These figures are for
persons who survived to at least six months. An estimate of life expectancy for
China in the 1920s that includes infant mortality gives an average of only
around 20 years. By comparison life expectancy at birth in England for
1800–10 was 44.8 (Brenner et al.,2008). Other estimates of life expectancy in
The Great Divergence since 1750 153