108 ECONOMIC ESSAYS IN HONOR OF JOHN BATES CLARK 
a correct theoretical explanation of the forces governing the pro- 
cesses of distribution. 
1. It will be noted from the discussion in the three preceding 
sections that we have tacitly assumed that the shares of the 
total product which the factors originally secured were equal, 
and that where only a change in bargaining power had occurred 
that an increase of one percent in the return to one factor meant 
a corresponding decrease of one percent in the return per unit 
of the other factor. But neither of these assumptions need be 
true, and in real life they certainly are not. What modifications in 
them would such other variables necessitate in our theory? Let 
us suppose that labor originally received two-thirds and capital 
but one-third of the total product. Then if, without any change 
in the net effectiveness of industry, labor were to increase its 
return per unit by five percent, its share of the total product 
would then rise to seventy percent; but the share of capital would 
fall to thirty percent, and if we assume that the total product 
would be unaltered, this would mean a fall of ten percent in the 
payment for each unit of capital. Thus, what would be a five 
percent increase in the return for each unit of labor would be 
a decrease of ten percent for each unit of capital. This would, 
of course, cause different movements of the supplies of these 
factors even though their elasticities were to be the same. Thus 
if each of their elasticities were positive and equal to unity, there 
would be an increase of five percent in the quantity of labor and 
a decrease of ten percent in the quantity of capital. This would 
be a stronger force towards restoring the original equilibrium than 
as if the supply of capital had only contracted in the same pro- 
portion by which the supply of labor had expanded. 
If the supply of labor were completely inelastic, while that 
of capital had positive unit elasticity, then an improvement in 
labor’s bargaining power would have similar results. For while 
the supply of labor would not increase, the supply of capital 
would decrease at twice the rate which it would, had the total 
product of industry been originally divided equally between the 
two. In consequence, the final gain of labor would be less than 
it would were a one percent gain for labor to cause a loss of 
only one percent to capital. 
The same results can be traced for all sets of positive elas- 
ticities. The larger is the share of the total product which is