Gil, thanks for your interesting post (3939 on Monday). Two replies below. On Mon, 2 Oct 2000, Gil Skillman wrote: > There is no problem flowing simply from the fact that a given theoretical > entity, in this case m, is unobservable. For example, > the notion that capitalist firms act to maximize profit is central to > neoclassical theory, even though profits are not perfectly measurable > because of imputed costs and other difficulties. However--and this is the > key point--the testable hypotheses of the theory do not depend on measuring > profit. Based on the assumption of profit maximization, you can derive > unambiguous predictions about relationships between strictly observable > variables, such as firm output and market price, without attempting to > measure profit. Whether or not one believes in neoclassical theory, it, > like other real economic theories, generates hypothetical relationships > among potentially observable variables, even if it includes theoretical > elements which are not themselves observable. So in the present case, > Fred's assertion of the existence of m would be theoretically innocuous if > doing so leads to testable hypotheses about economic relationships among > strictly observable variables. I am glad that you (unlike Ajit) agree that it is at least conditionally legitimate to assume unobservable givens, on the condition that the theory derives some "unambiguous predictions", which can be empirically tested. I argue that Marx's theory does indeed derive "unambiguous predictions" about important phenomena in capitalist economies, which can be empirically tested. These "unambiguous predictions" are that capitalist economies will be characterized by: (1) conflicts over the length of the working day; (2) conflicts over the intensity of labor; and inherent technological change. All these important phenomena are derived as logical deductions from the basic assumptions of a given L and a given m. These initial givens are unobservable, but they are used to derive important "unambiguous predictions" about observable phenomena. It is obvious that these "unambiguous predictions" of Marx's theory are strongly supported by the history of all capitalist economies over the last several centuries. Therefore, Marx's theory satisfies Gil's condition of "unambiguous predictions" that can be empirically tested. In which case, Marx's assumption of the unobservable givens of L and m would seem to be justified, right Gil? > > In any case, I assume it's settled that in the expression S = mL- V, S > *cannot* be said to be determined "up to a factor of proportionality," > contrary to Fred's previous claim. No, Gil, this is not "settled". I argued in (3922) that surplus-value is UNIQUELY determined, not just determined up to a factor of proportionality, because m is taken as given as a unique value. Marx's theory assumes that there is an actual m in the economy, with a definite magnitude; i.e. that each hour of abstract labor produces a definite quantity m of money new-value. It is this actual m, with a definite magnitude, that is taken as given in Marx's theory of surplus-value. If surplus-value is uniquely determined, then it is certainly determined up to a factor of proportionality. As I explained in (3922), my earlier argument in (3831) - that s is determined only up to a factor of proportionality - was based on an erroneous assumption that, because m is not theoretically determined, m could be any value. But I now realize more clearly this is not true. Even though m is not theoretically determined, m cannot have any value. At a given point in time, m has a unique value, the actual m in the economy. If m could have any value, then you would be right - S would not be determined up to a factor of proportionality. But m cannot have any value. m is taken as given as a definite magnitude, as the actual m that is assumed to exist in the real capitalist economy. As a result, S is uniquely determined, not just up to a factor of proportionality. Again I ask: why is it not permissible to take as given this unique, actual m? Comradely, Fred
This archive was generated by hypermail 2b29 : Tue Oct 31 2000 - 00:00:08 EST