Stability in neo-classical and neo-keynesian growth models

The aim of this paper is to investigate the properties of neo-classical and neo-keynesian models of economic growth with respect to the stability of dynamic equilibria. the first question that immediately arises is: what do we mean by stability? to answer this question we must formulate our specific problem in such a way that we can first of all find out what concepts of stability we could theoretically apply and, secondly which one make sense in this context. we are considering dynamic systems, the structure of which we know via a set of equations. what we are interested in are the time-paths of the endogenous variables and whether there exists an equilibrium time path of these variables. when these problems are solved, we have to ask ourselves how this equilibrium can be attained and what happens to it when disturbances occur. stability can now mean that: (1) whatever the initial conditions of the system are it will tend to move towards a steady-state growthpath, which (a) is the same as if the relation of the relevant variables had been the ("right" being the steady state solution ratios) from the outset on; (b) is different from this steady-state path (2) whatever changes in the size or the sign of the structural parameters of the system may occur there will be a path back to equilibrium. the present paper is mainly concerned with initial conditions as most of the other questions have been dealt with in great detail in other publications. another problem of great interest to us was to find out what values these parameters may take to yield a stable equilibrium and to what limit they may alter until the system does not return to its steady state. the basis of our investigation were the growth models of meade, solow, etc. on the neo-classical side and the kaldor-model on the other, both of them in non-vintage form.;