Summary

The past two decades have seen a sharp renewal of interest in the Kaleckian models of growth and distribution. These models have been put to so much use that it made us wonder if this trend was based on their ability to generate macroeconomic equilibrium variables located within a reasonable range of some conceivable economic reality. While the Kaleckians spend much time debating how much their models are wage-led vs. profit-led, our work lies upstream of these questions, for we are investigating the Kaleckian models ’ existence properties. We conducted a series of simulations based on a basic, three-equation model: an investment function, a savings function and an equation for the firms ’ objectives. By estimating the parameters composing these different functions without recourse to econometrics but by means of an economic diagnostic, and by specifying an assumed range of value for the equilibrium variable, we tested these models ’ ability to produce plausible results. We also tested the models ’ stability property. Although stability is often put forth as a pre-requisite in the study of such models, in the Kaleckian literature the question of how plausible is the stability is rarely raised. Our results suggest that there is a need to reconsider these models, their investment functions and, more particularly, certain of the parameters appearing in these functions. By comparing the relative performance of the models in terms of stability and plausibility, we are able to establish an interesting property: the most plausible models are unstable