132	Perspectives in American History

grew, its industrial sector failed to grow with it and declined in
relative importance. The 1790 directory (the fourth published) is
remarkably full and detailed, even by London standards, but it/
suffers from one obvious blemish: no clerks are designated as sue
though many clerks are undoubtedly concealed among the uniden-
tified males. By contrast, the yellow fever list of 1795, thougn con-
taining far fewer names than either the freemen's registers or the
city directory, probably contains a more reasonable sample of New
York's adult male population (women and children were more
likely to have been out of town), and thus provides the best avail-
able picture of vocational distribution. Because yf its less distorted
measurement of nonhouseholders, including clerks, this 1795 death
list is not only better than the other earlier and longer New York
lists but also probably represents nonhouseholding elements in the
population more clearly than do the Philadelphia data. Neverthe-
less, the sample is not perfect and has to be used with caution.

There is a further problem common to the data of all the towns.
In assigning specific occupations to/one or another of the sectors,
certain difficult and somewhat arbitrary decisions have had to be
made, for some occupations might rightly have fallen in more than
one sector. Care must be taken/to distribute such overlapping crafts
so as not to favor any one sector. Thus, butchers and bakers (two
numerous crafts) have been assigned to the service sector, where
they are usually thought/to belong, though many in those trades
worked not for the loca/market alone but also for the export mar-
ket and belong at leas/pardy in the industrial sector. On the other
hand, furniture makers and shoemakers (cordwainers), another
numerous craft, Mve been placed in the industrial sector, for
they made good/in job lots for external sale, but a good part of
them also worked for local orders and could be put in the service
sector. It is impossible, given present knowledge, to divide the
totals employed in such crafts into fractions working for the local
and external markets; however, it is more than likely that any
overweighting caused by the assignment of certain crafts in to to to
the external market (industrial sector) is balanced by the other
overweighting caused by the assignment of other crafts in toto to