21 
3. In the above ( point A ) we have tacitly - and 
fictitiously - assumed that the labour input in machinery 
is distributed oirer the whole life-time of the machine. 
In reality it will occur in one lump before the increase 
in producivity takes place and labour is displaced: 
The compensation, or rather overcompensation, preceeds 
the displacement. Afterwards, over the whole life of the 
machine, there will be no more compensation, but only 
displacement. 
This explains the expectation of Kalecki that innovations 
will start an umward trend, though this may be reversed 
unless new further innovations follow, 
ad B. 
Compensation undff’this heading depends basically on the 
distribution of the additional (potential) real income. 
If the share of wages is at least as great as in the 
average income of the economy, that is if wages increase 
in pace with productivity, then there is a ggod chance 
for compensation. 
But it is not only necessary that the workers participate 
in the result of the productivity increasem there must also 
be additional investment in order to realise the expansion 
of output required for compensation. 
To illustrate the compensation problem we can make use of 
the department scheme. We assume the productivity has 
increased in department II only. The investment which has 
caused it took place in the past and has already exhausted 
its stimulating effect so that we do not have to consider it 
in the following. We assume that wages per man in department 
II increase in step with the increase in productivity. 
The wage bill there would thus remain unchanged if