Profs. Serrano and Petri
I somehow missed this paper so far. It explain the several additional problems with Say's Law even if one does not rely on the capital debates to show the limitations of the downward-sloping investment curve.
From the abstract:
Neoclassical capital-labour substitution correctly understood is unable to prove a tendency toward the full employment of resources because it leaves investment indeterminate if the full employment of labour is not assumed to start with; then Say's Law loses plausibility because of the inevitable presence of accelerator-type influences on investment, even neglecting the inconsistencies of neoclassical capital theory; and wage decreases cause a decrease of investment, undermining the 'neoclassical synthesis' criticism of Keynes. The way a negatively interest-elastic investment function is obtained by Romer without assuming the full employment of labour, that is through adjustment costs, relies on several grave mistakes. The recent DSGE models which directly assume that investment equals savings are not supported by general equilibrium theory because the latter theory is admitted by the specialists not to be a positive theory, nor can those models rely on the neoclassical synthesis or monetarism because of the critique of this paper (besides the capital critique), so they must be discarded too.Read rest here.