purposes instability can only mean that the system is unable to 
cope with them adequately. 
A necessary condition for macroeconomic change - trend or cycle - 
is that the disturbances and the corresponding reactions in the 
small do not offset each other so as to permit stability of the 
whole system but rather tend to go to a large extent in the same 
direction. They are unlike a self regulating system either because 
the individual movements reinforce each other (imitation) or 
because they all respond to a common signal,for example, the rate 
of interest. 
In my opinion trend and cycle are to be treated from the point of 
view of the following "research program" (in the sense of 
Lakatos): The macroeconomic system is like a machine which works 
up and transforms the disturbances which are fed into it from the 
outside in the course of time. This approach is dictated by a 
dominating practical interest in economic policy: We want to know 
how the system reacts to various measures or events and how the 
daily or monthly movements of the Konjunktur (state or tendency of 
trade) have come about in each concrete case. This preoccupation 
with the practical aims of the theory are the reason why I can not 
be convinced by Goodwins insistence on a non-linear treatment of 
the cycle. The response of the system to disturbances is, I hope, 
adequately dealt with by linear approximations. This is true also 
for the long term development which is only the result of an 
accumulation of short term changes. 
This leads to the question of the unity of trend and 
cycle.Unfortunately very often independent and separate theories 
have been produced for the one and the other even by those authors 
who intended to arrive at a unified theory. The failure stems 
probably from the mathematical formulation which is so much easier 
if you have one equation for each,independent of each other. But 
in reality the trend component and the cyclical component are 
determined at the same time and are parts of the same process, 
separated only artificially by statistical or analytic 
prodedures.The unity of trend and cycle has been proclaimed most 
forcefully by Goodwin (1982,p.115,p.122-123) who also quotes 
Schumpeter in support of this view.Goodwin argues that the 
separation of the two theories is based on the superposition 
principle which is justified as long as we work with linear „ 
equations. 
The question. _is whether, even without departing f rt>m simple linear 
equationswe can establish a theory^, in which trend and cycle are 
interdependent,each of them influencing the other. My feeling is 
that the "pure trade cycle” theory of Kalecki might be enlarged by 
a term( or terms) which represents "long term memory". This would 
be an integral over a number of years past of certain variables 
such as for example utilisation of capacity, multiplied with a 
certain reaction coefficient. This term, since it changes 
slowly,would provide for a smooth long term development. Since it 
might cover something 1 ike the— av-erageLength of a ft*41 cycle it 
would involve an influence of the last cycle on the present trend 
value: If this cycle consisted of a long and strong boom and a