The Keynesian models fundamentally divorce consumption spending from 
investment spending. In the work of the early pioneers of neoclassical 
theory (Walras, Jevons, Menger, Clark, Marshall, Pareto), decisions to 
invest meant decisions that ultimately led to the production of consumer 
goods. Some consumer goods could be produced directly but the vast 
majority required the prior production of capital goods. Thus, real 
investment typically meant the production of capital goods at one link 
in a supply chain with the help of capital goods that had already been 
produced at another link in the supply chain that was "farther away" 
from the ultimate consumer good. The idea of "investing" without there 
being supply chains would have been unthinkable to the general 
equilibrium, marginal productivity theorists of the late 19th and early 
20th century.

I am amazed when I reflect on the fact that Keynesian macroeconomics is 
built on a foundation that almost completely disregards this most 
important idea. In this field, the decision to produce capital goods is 
completely disconnected from the decision to produce consumption goods. 
Similarly, there is virtually no connection between the early general 
equilibrium theory of the neoclassicals mentioned above and the later 
Keynesian macro. Arguably the most important lesson in economics since 
Adam Smith's invisible hand was expunged from the "science" and 
generations of "macroeconomics" students were trained as if it had never 
existed.

The reason why I object to AD/AS analysis being taught in the principles 
classroom and to IS/LM being taught in the intermediate and graduate 
texts is that it amounts to a retrogression. To teach this is, to me, 
analogous to burning the libraries at Alexandria as opposed to learning 
from them. To carry the analogy a step farther, I regard it as part of a 
dark age in the history of economics.

Pat Gunning