d 1895 was due only to a very
small extent to defections. The introduction of the linotype decreased
the opportunity for employment in the trade, and the gradual shrinkage
in the amount of German printing done in the United States due to the
falling off in German immigration was accentuated by the depression.

While the two unions having the most highly developed beneficiary
systems thus show an ability to retain members during periods of
depression, it would be absurd to assume that this result is solely the
effect of the establishment of the benefits. The Cigar Makers' Union in
1892 would undoubtedly have held its membership better than it did in
1872 even if it had developed no benefits. It is interesting in this
connection to note that while in the depression of 1873-1878 the
membership of the Typographical Union fell from 9799 to 4260, a loss of
forty per cent., and the number of local unions decreased from 105 to
60, in the great depression of 1893-1897 the membership fell from 31,379
in 1894 to 28,096 in 1897, a loss of only ten per cent. Part even of
this small loss was due to the withdrawal of the pressmen and
bookbinders from the organization. It thus appears that the
Typographical Union with a death benefit of sixty-five dollars and a
home for the aged held its membership almost as well as the Cigar Makers
with their much more highly developed beneficiary system. The change in
the power of the Typographical Union to retain its membership was
obviously due not so much to the establishment of beneficiary features
as to the greater support which it gave its members in collective
bargaining.

A comparison of the effect of the depression of 1893-1897 on the
Typographical Union and on the Brotherhood of Carpenters makes the point
still clearer. In 1893 when the depression set in the per capita
expenditure of the Typographical Union for beneficiary features was
$1.50, while that of the Carpenters was $1.40. The death benefit in the
Carpenters' union was graded in such a way as to offer an additional
incentive to retain membership. The two unions were, as far as the
development of benefits is concerned, on about the same plane. As has
been noted above, the Printers lost almost none of their members. The
Carpenters lost from 1893 to 1895 over half of their membership. The
following table shows the membership of the Carpenters by years from
1890 to 1900:

1890....53,769 1894....33,917 1898....31,508
1891