17 
criterion of size, as Hr. Johnstone^ proposed, -the 
automated plants would count as small (or lower medium) 
sized, and output per man would accordingly have a maximum 
at this level. I cannot find anything enlightening in this 
use of language, hut it does in itself, of course, not 
contradict my picture of the relations of technical progress 
and scale, which requires in any case a description of the 
plant in terms of oeeond variables (man, output, capital). 
v 7 ; ■) t ' t t.c a/ lo 
A point on which I would like to correct myself again: 
Just as I was wrong to play down diseconomies I was too 
unfavourable on the chances of the small firms, although 
my remarks on JLts shrinking degree of independence were 
borne out by the developments since. 
One has to take into account the emergence and vast growth 
of new industries mostly composed of small firms (new types f f 
(6- 
service industries of various trends, like the "American 
laundry” etc.). Further, the tendency for large industrial 
concerns in some industries to subcontract work has been 
widespread. In retail trade and service industries a kind 
of polarization seems to take place in advanced industrial 
countries: On the one hand the bulk of essentials is traded 
in self-service shops which are best carried on on a not too 
tb. Johnstone, Statistical Cost Analysis, London i960