Few real-world circumstances are as challenging to economists as recessions. First of all, economists are notoriously unable to predict recessions, a fact that became glaringly obvious when the Great Recession began eight years ago. The reason is their over-confident reliance on one kind of forecasting tool and their weak, even non-existent reliance on economic theory for their analytical work. (Some of my peers would protest here by challenging the meaning of "economic theory"; for the record, I do not consider a recursively soluble equation system "economic theory".)
Beyond that, you will also find a surprisingly large number of practitioners of macroeconomics who are simply uninterested in the business cycle. They tend to specialize in a small area of macroeconomics, such as the labor market, monetary theory and policy, or public economics. Nothing wrong with that, but for every specialist there should be a macroeconomic generalist out there who can analyze and explain the entire macroeconomic system. We are too few and too far in between to be heard when it really counts...
Another problem is the shifting tides of intellectual fashion. Up until circa 1980 one version or another of Keynesian economics dominated macroeconomics. That famous line "We are all Keynesians now" often attributed to President Richard Nixon was replaced rapidly in the early 1980s with "We are all new classicals now". The pivotal moment was the publication of the paper After Keynesian Economics by Robert Lucas and Thomas Sargent in 1979. Since then, macoreconomics as taught at major universities shifted away from business-cycle analysis and system-wide thinking to a kind of aggregate microeconomics that, quite frankly, reduced the entire branch of macroeconomics to eclectic flea killing.
For reasons that I partly touched on in my article Keynesianism: The Real Story, macroeconomics is not a simple aggregation of microeconomics as new classical economists tend to suggest. It is more complex than that. New classicals, spearheaded by - and to some degree misinterpreting - Lucas and Sargent, believe that the macroeconomic system automatically produces full employment and price stability, if only left alone by government. Recessions are explained by so called singularities, namely one single factor that causes the economic system to temporarily divert from full employment. Were it not for that single factor, the economy would be producing at full use of all resources and everyone would be employed.
The consequence of this singularity thinking is that economists always configure their forecasting models so that, over the long term, the economy will always revert back to full employment. They will tell you that there is always a recovery coming after every recession. But just as they cannot tell you when the recovery is coming - other than "in the long term" - they are unable to explain when the recessions break out. To make matters worse, the economy in real life, the one we all live in, does not behave the way new classicals suggest. A recession is not caused by a singularity such as "sticky prices" or "imperfect information" or "search costs" on the labor market. Recessions are caused by a systemic shift in economic activity, which is either minor, thus leading to a recession, or major, thus throwing us into a depression.
In the world of new classical economics, depressions do not exist and recessions solve themselves if left alone by government. I am not going to disagree wholeheartedly with their suggestion as to how to deal with a recession, but if a recession is mishandled or even misinterpreted by legislators it is entirely possible that a passive government (just like an over-active government) can escalate a recession into a depression.
New classicals lack both the theoretical foundation and methodological mechanics to explain any part of what I just said.
Keynesians, on the other hand, can explain the difference between depressions and recessions, as well as the fork in the road that can take us from the latter into the former. There will be time later for an in-depth discussion of this moment in the business cycle; the point here is that pure Keynesian macroeconomics (as opposed to hybrids like "New Keynesian Economics") does not by default predict that the economy will either remain on, or return to, a path of full employment and price stability.
Simply put, it is an open question where the economy will go once there is a systemic shift in economic activity.
That does not mean you cannot forecast the future based on Keynesian macroeconomics. You can, as I demonstrated in three articles in The Street in October last year (see here, here and here). But you have to understand what drives shifts in economic activity:
1. Uncertainty caused by the "net diminution" of economic activity I pointed to in Keynesianism: The Real Story; and
2. Flock behavior, which is driven not by rationality or perfect information, but by what Keynes summarized as the "animal spirit".
The last point is eminently analyzed and explained by Armen Alchian in his brilliant, and thoroughly challenging essay Uncertainty, Evolution and Economic Theory. The point is that economic agents, when resorting to flock behavior, can drive the economy in any direction, downward as well as upward, and it can happen at any point in a business cycle.
The role of government is of course critical for the turn of events along the business cycle. New classical economists treat government as an aberration from full-employment, price-stable general equilibrium; as a Keynesian, on the other hand, you have to elevate government to the role it actually plays (that not being an endorsement of any particular size or function of government) and then treat it as the inefficient but major player in the economy that it actually is.
Granted, some of the points here are generalizations, and New classical economists will rightly claim that they have covered all thinkable angles of the problem areas discussed here. However, when they do, they do it with explicit or implicit reference to the core principle of New classical economics: that in the aggregate the economy is nothing but a microeconomic system with flexible prices, flexible resources and at worst one singularity explaining aberrations from long-term full employment and steady growth.
In short: when properly applied, pure, old Keynesian macroeconomics is superior to New classical economics. However, there is yet another theory that, in recent years, has made a strong comeback among economists: Austrian theory. How does it stack up against Keynesianism?