4. 
Thus the possibilities of increasing the capital-coefficient without 
reducing the profit rate depend on the size of the ruling profit margin, in 
other words, on the distribution of income. (It should immediately be noted that 
this ruling profit margin will be different for different size classes of firms). 
We now turn to the relation of capital intensity and productivity. It 
can be shown that the growth rate of the capital-output ratio equals the growth 
rate of capital per man minus the growth rate of output per man. On p. 33 I used 
somewhat different concepts: Instead of capital per man I used the capital-cost 
ratio, and instead of output per man the sales-cost ratio. If we take the wage as 
constant, the analysis of p. 33 will come to the same as the present analysis, but 
it should be noted that in fact real wages can differ between firms (e.g., of 
different sizes), and especially over time. 
In the light of the engineering data, it seems that the growth in pro 
ductivity will outweigh the growth of capital intensity over a wide range of 
output: In consequence the capital-output ratio will decline over this range. 
This is indeed apparently the course of much of the technical progress. It might 
be that beyond a point the relation will be reversed, the capital growth will out 
weigh the productivity growth, so that the capital-output ratio will increase; 
there will be diminishing (marginal) returns to capital which on p. 33 I still 
regarded as plausible. Today I doubt whether the range of diminishing returns 
of capital-intensification is actually entered in the ordinary course of events; 
such techniques are perhaps not developed at all, and diminishing returns to 
capital-intensification exist only theoretically beyond the point at which 
actually developed methods end. This point is, however, shifted further on in 
the course of technical progress. This avoidance of the range of increasing 
capital-output ratios may result from the above stated influence of high profit 
margins. 
What can be the relevance of this analysis to the question of scale? 
In Chapter III I used it to explain why the highest size groups of corporations 
show signs of diminishing profit rate, for which I could find no technical reasons 
(diseconomies of scale I assumed to be unimportant). But if the capital 
coefficient does not rise with increasing scale, the whole argument falls to the 
ground. There is an even more direct objection against it: Why, one might ask, 
do the big concerns apply such methods of production, if they yield them a smaller 
profit rate than less capital-using methods would? In fact, the analysis of 
Chapter III is much more suitable for explaining why the large firms do not