"Next, Keynes -- noting, like Bittlingmayer, cost and demand conditions under which competition doesn't emerge, Keynes goes further, to explain how economists move from simplifying assumptions to abandonment of the actual facts, and to conclude that their model is what reality is."
As an aside, Wassily Leontief makes a similar point about simplifying assumptions in modelling in one of his Essays on Economics, but I cannot provide the exact reference just now. Leontief's point tends to be more that modelling is abused in the sense that dodgy inferences are drawn from modelling phenomena on the basis of simplistic assumptions, causing pradoxical results, whereas much better inferences would result if a comprehensive study of available empirical data about the subject area was made. That is to say, modelling might cause us to abstract precisely from those data sources, which are required for a balanced objective assessment, and for the conditions for a solution to the given economic problem. This is another way of saying, that in modelling we may assume precisely what we need to explain: modelling becomes a way both for evading serious theoretisation (substituting simplistic analogies and metaphors) and evading a comprehensive confrontation with the facts which already exist (statistically known trends, circumstantial evidence). The inferential wizardry of the model might be impressive, but it might lead to very little result, precisely because of the initial conditions (the simplifying assumptions). The significance of Marx's Capital is I think precisely that he builds a "model" which not only incorporates more and more assumptions, but also is capable of overturning those assumptions, while at the same time integrating ever more empirical data and inferences into the picture. Marx knew very well that the circuits of economic life (the overall process of economic reproduction) could be approached from many different angles, but not every starting point or initial assumption could lead to a coherent theoretical structure being built up, since, in order to understand one set of economic phenomena, another set of economic phenomena needed to be understood first, or could not be understood in separation from another set of economic phenomena. He felt therefore there should be an order (sequence) of discussion and analysis, but in his critique of political economy, this ordering principle is not given within economic science, but external to it, namely by the materialist conception of history, in which the foundational role of production and a genetic-historical approach are primary. J.
