Today (Friday the 13 of September, 2013), Paul Krugman's blog (http://krugman.blogs.nytimes.com/2013/09/13/wynne-godley-and-the-hydraulics/) compares the "hydraulic" macroeconomics of the late Wynne Godley to the currently dominant school and finds Godley's view to be lacking.
One key fact that Krugman misses is that "hydraulic Keynesianism" had a completely different focus than the "modern" neoclassical version of Keynesianism that largely replaced it. While hydraulic Keynesian focused on flows of money (with the correction for the effects of inflation being only an afterthought), the neoclassical Keynesians try to see the world totally in terms of "real" variables. Maybe money is crucial in the short run, the neoclassical reasoning goes, but in the long run (which is what's really important) money doesn't matter. Money is only a veil that must be removed to understand what's really happening. To "hydraulic" Keynesians such as Godley, the short run can be extremely crucial, sometimes causing economic collapse. The neoclassical Keynesian focus on the long run (as a result predetermined by the "supply side") totally misses that possibility.
There's also a big difference between hydraulic Keynesianism and the old-school Keynesianism that once dominated academic macroeconomics. In fact, the failures to predict macroeconomic behavior that Krugman points to are not really part of the "hydraulic" aspect of Keynesianism. The prediction that saving would rise (and consumption would fall) relative to total income after World War II was based on an incomplete understanding of the determinants of aggregate consumer spending. Because it was entirely new topic when Keynes wrote The General Theory, it should not be surprising that he got some of the analysis wrong.
This problem was solved by James Duesenberry, who was very much part of old-school Keynesianism. He was not a "hydraulic" thinker. Just as the more sophisticated economists are doing today, he imported elements from psychology into economics in order to understand what was going on. (Duesenberry used math, which should have endeared him to the orthodox economists, but he's largely been forgotten.)
Milton Friedman's innovation on this issue arose from the fact that (as a neoclassical economist) he did not like Duesenberry's rejection of a hard-core individualistic vision of humanity. Instead, he embraced that hard-core theory. This lead him to assume (surrealistically) that people were narrow-minded and isolated maximizers. Somehow, in his view, these "people" were able to plan ahead far into the future despite the blatant uncertainty that will always be associated with future events. (Keynes understood uncertainty's role, but not Friedman.) By the way, both Duesenberry's and Friedman's theories do just as well in terms of predicting changes of aggregate consumer spending over time.
As for the failure in predicting inflation (which became obvious in the 1970s), again the error really wasn't due to a hydraulic focus (an emphasis on flows of money). Instead, it was based on faith in the real-world data held by respected economists such as Paul Samuelson and Robert Solow. In this view, if the data showed that there was a stable trade-off between unemployment and inflation, there must be one. But it wasn't Friedman or Phelps who first pointed out the problems with this perspective. Rather, it was the old-school Keynesian Abba Lerner (who, like Duesenberry, lacked a "hydraulic" bent) who very early on noted that if unemployment stays "too low" and inflation rates stay "too high," inflation will get worse and worse. By the way, Lerner's theory is much more sophisticated than the Friedman/Phelps one, because he never accepted the idea that the "full employment" unemployment rate could be represented by a unique number.
(By the way, economist Nathan Tankus has pointed out that Godley was an early critic of the Phillips Curve.)
If the failure of either old-school or hydraulic Keynesianisms to predict real-world events encourages people to reject these views, we should also reject the neoclassical Keynesianism that expresses itself so clearly in Dynamic Stochastic General Equilibrium theory. That perspective bombed totally when it came to understanding the boom of the early 2000s and the crash of 2008-09.
-- Jim Devine