Let me tell my story using the following table, which shows the primary causes of economic fluctuations according to different economists.:
Animal Spirits, Liquidity Preference
Keynes, Akerlof-Shiller, Krugman
Lucas, Dornbusch, Fischer, Blanchard
Clower, Leijonhufvud, Hayek, Kling
Macro practitioners (the people who report that the multiplier is 1.54) use econometric models that do not fit into any of the columns in the table. What is wrong with those models is a whole other story.
Basically, over the past thirty years, the economics profession gravitated toward the center or the table, with a focus on monetary shocks and productivity shocks as the causes of macroeconomic fluctuations. There was an emphasis on a particular "haiku," as Blanchard calls it, of Dynamic Stochastic General Equilibrium (DSGE) modeling.
Krugman, Blanchard, and I were students at MIT in the late 1970's, when Dornbusch and Fischer were young professors who proceeded to dominate macroeconomics in the department, displacing older Keynesians. They put MIT on the path of focusing on monetary shocks, and they created an environment in which the saltwater economists of MIT could join with the freshwater economists of Chicago and Minnesota in what I consider to be the fruitless mathematical exercises of DSGE.
Blanchard embraced Dornbusch and Fischer, and he went on to teach at MIT (although at the moment he is on leave as chief economist at the IMF) and to become a co-author of their macro textbook. Krugman retreated from macroeconomics altogether, and had to settle for Nobel Prize in international trade. I retreated even farther, and I have never held a tenure-track position in economics. As my initial reaction to Krugman's New York Times piece showed, Krugman and I have some common enemies here, even though I would put us in different columns in the table.
A monetary shock is a change in the growth rate of money that surprises the public. It has always seemed to me such a silly way to characterize macroeconomics that it is hard for me to try to explain it with a straight face. Basically, all of us are happy and fully employed, and then one day the central bank secretly withdraws money from the system. This causes great confusion, and real wage rates are in disequilibrium for a while, causing unemployment. Again, I cannot tell this story with a straight face, but if you ask any mainstream macroeconomist who did graduate work in the last thirty years they should be able to make it sound convincing.
A productivity shock is something that affects aggregate productivity. Not by accident did this story emerge in the wake of the surges in oil prices in the 1970's, which raised costs of production and consumption in many industries. However, it never was a convincing story for the Great Depression--although some people, like Cole and Ohanian, certainly have tried it there.
Animal spirits for Keynes represented the impetus to investment in a world of unknowns. For Akerlof and Shiller, animal spirits is a much broader term. I think they have mis-appropriated the term to cover all sorts of deviations from the classical rational model. I like the original Keynesian notion of animal spirits, although I might prefer to supplement changes in the optimism of entrepreneurs with changes in the risk perceptions in financial markets. The latter might loosely be termed a Minsky model. I don't think there is much in Minsky worth adopting, but the idea that risk perceptions are cyclical has great merit.
Liquidity preference says that something special happens when people save in the form of money. According to this view, if instead we saved in the form of goods, then an increase in saving would not decrease the demand for output.
As much as I like animal spirits and cyclical risk perceptions, I am inclined to reject the liquidity preference story. I do not think that attaching significance to money as a store of value is the right way to pursue macro. Hence, I reject Krugman's babysitter co-op parable as a useful story. In that parable, it is clear that liquidity preference destroys the economy, because there is no way to store babysitting over time. Relative to the real world, in which there are many ways to trade future consumption for current consumption, I think that story is highly misleading.
Instead of liquidity preference, I prefer to focus on problems of recalculation. Suppose that the economy needs to get from one equilibrium to another. In the first equilibrium, we build lots of houses and start few high-tech businesses. In the second equilibrium, we build few houses and start lots of high-tech businesses. Getting from one equilibrium to the other requires countless recalculations, wage adjustments, and movements of resources. These do not happen instantly or simultaneously. Instead, along the way there are frictions and miscalculations. These produce unemployment, and indeed we never get to the same equilibrium that we would have arrived at had the recalculation occurred instantly.
In my opinion, the state of macro is very bad (if anything, worse than Krugman makes it out to be), because nobody followed up on the thinking of Clower and Leijonhufvud. The main villains in my story are Dornbusch, Fischer, and Blanchard. I have nothing against them personally, but I think that their influence on the profession has been excessive and adverse.