What can we do with the profit function? 
In the short run, if profits haave to he adjusted upwards or 
downwards becaus e there is a boom in or a slump in investment, 
and the saving mainly comes out of profits, then the utilisation 
will vary and all the other quantities will stay constant, 
these other quantities being rather rigid in the short run. 
In the long run we cannot take the constancy of these 
factors for granted. First , they will chagge in the course of 
technical ( h,h Q ) and economic ( Tf ,w ) develppment. u econd, 
they may haveeto change, if the utilisation is not any more a 
passive factor, and therefore the adjustment of savings and profits 
can not rely on the fluctuations in utilisation any more. 
This second caae is relevant for my theory: I assume 
that firms are in principle concerned with establishing in the long 
run, that is on the average over boom and slump, a normal or 
desired degree of utilisation. This is done by varying the pressuce 
of competition in such a way that either more or less hxekssx 
capaaity is eliminated in a given year. The variation in competitive 
pressure involves a ahift, or change in xhajsax the parameters, 
of the profit function. Thus, in the long run, an increase in 
profits and savings may be brought about by an increase of the 
-fr* — hw 
mark-up —2 rather than an increase in utilisation, 
7T 
and mutatis mutandis the same is true for a decrease in profits. 
In this sense the long run distribution pattern will change aas 
a result of a change in the speed of accumulation (rate of investment). 
In connection with this we have now to eexplain a technical 
point. If the increase in prdducktivity affeet^variable and 
overhead cost in the same proportion ( if h and ho ^.s reduced in 
the same porportion) then a corresponding increase in w/ir 
may compensate it so that the parameters of equation (lo) rremain 
unchanged. In the al&erhative case/of a non-ppoportionate movement 
- for example, if the direct cost decrea:se and the overhead cost 
are const a nt or even increase, a compensating movement of the 
wage-price ratiowill not restore the original values of the 
parameters: the shape of the function will not he the same as 
before, and the amount of compensating movement rquired to keep 
the rate of profit unchanged will depend on the existing degree of 
utilisation.