bourers now employed in the production of calico may be immensely
less in proportion to a given quantity of calico, the total number
required for the millions of yards now used greatly exceeds the number
engaged when the whole work was performed without any aid of
machinery."[175]

Now, turning from the consideration of the particular instance, which
we shall find reason to believe is peculiarly unfortunate when we deal
with the statistics of the cotton industry, it must be observed that
economic theory makes dead against this _a priori_ optimism. Ignoring,
for the sake of convenience, the not improbable result that an economy
of production may, at any rate for a time, swell profits instead of
reducing prices, it will be evident that the whole value of the
argument turns upon the effect of a fall of price in stimulating
increased consumption. Now the problem, how far a given fall in price
will force increased consumption, we have found in our discussion of
monopoly prices to involve extremely intricate knowledge of the
special circumstances of each case, and refined calculations of human
motives. Everything depends upon "elasticity of demand," and we are
certainly not justified in assuming that in a particular industry a
given fall of prices due to machine-production will stimulate so large
an increase of consumption that employment will be given to as many,
or more persons than were formerly employed. On the contrary, if we
apply a similarly graduated fall of prices to two different classes of
goods, we shall observe a widely different effect in the stimulation
of consumption. A reduction of fifty per cent. in the price of one
class of manufactured goods may treble or quadruple the consumption,
while the same reduction in another class may increase the consumption
by only twenty per cent. In the former case it is probable that the
ultimate effect of the machinery which has produced the fall in
expenses of production and in prices will be a considerable increase
in the aggregate demand for labour, while in the latter case there
will be a net displacement. It is therefore impossible to argue _a
priori_ that the ultimate effect of a particular introduction of
machinery must be an increased demand for labour, and that the labour
displaced by the machinery will be directly or indirectly absorbed in
forwarding the increased production caused by machinery. It is alleged
that the use of steam-hammers has displaced nine of the