Paul Krugman continues to invoke Keynes’s famous statement. I wish Krugman and others would give some serious thought about what it is supposed to mean and the errors it involves.
In the first place, Keynes was complaining about the “classical” economics, that is, the ideas of the economists before him who believed that the market, if unhampered after a recession, could reduce or eliminate the unemployment associated with the business cycle.
Of course, this puts many economists – with different ideas – in the same category and treats the issue of cyclical unemployment in a grossly simplified way. But this, in general, is how Keynes treated those who disagreed with him. Keynes, the polemicist, was without inhibition.
Some basic methodology is in order. When economists talk about “the long run” they do not mean calendar time. Yes, that’s right. They do not mean long in the sense of many years or perhaps even many decades. (On this see Fritz Machlup, ”Equilibrium and Disequilibrium: Misplaced Concreteness and Disguised Politics,” Essays in Economic Semantics (1963))
Then what do they mean? It is actually a technical term that has meaning only in the context of a specific model. I believe that Keynes knew this but used the everyday meaning of the term to sow confusion about the ideas of his predecessors and to use that to his advantage.
The long run happens when all of the variable elements in a model are fully adjusted. It is an intellectual experiment. Suppose, simply, that a model says there are firms which face prices, make quantities of outputs and have a certain size. Now let the “permanent” demand for their product increase. Firms will face higher prices and their outputs will increase. But until firm-capacity is adjusted, there is no long run. Not everything in the model has adjusted.
Will this complete adjustment take a long (calendar) time? Maybe or maybe not. It will depend on many specific circumstances in the industry. It could be extremely rapid.
The impression that Krugman wants to convey is that only the stimulus-crowd wants to make things better now. The opponents, on the other hand, want to let present suffering continue as they fear some possible problems in the distant future of calendar time.
But let’s look at the arguments made by the opponents of fiscal stimulus.
Some have argued that, as deficits increase, people now offset the putative stimulus by increasing their savings in anticipation of future tax increases. So there is no stimulus now.
Others have argued that, for example, extending unemployment insurance (again) to those unemployed for more than six months will increase the length of unemployment now (by subsidizing it) while failing to stimulate.
The stimulus failure is due to the relatively small increase in spending induced by non-permanent increases in income (as unemployment insurance is certainly not permanent source of income). Even more, producers know that the spending is non-permanent so it is unlikely to result in increased employment of labor. Thus, there is no stimulus now; in fact if unemployment continues there is a kind of anti-stimulus now.
Austrians have argued that failing to allow the housing market to adjust by both fiscal and monetary propping-up measures, worsens the situation now by prolonging the inevitable adjustment to a bubble sector. As the adjustment is dragged out and the rest of the economy suffers the dampening effects now. This must include the uncertainty as to when (in calendar time) the market will be allowed to adjust.
In empirical work, John Taylor finds that to the extent there was some effect of the fiscal stimulus it was very small and lasted only a matter of two or three months for each major injection. So I guess the long run is four or five months by this reckoning:
Compared with the 2008 stimulus, the 2009 stimulus was larger, but the amount paid in checks was smaller and more drawn out. Nevertheless, there is still no noticeable effect on consumption. I also show the timing of the “Cash for Clunkers” program in Figure 7; it did encourage some consumption, but did not last and cannot be considered an effective method to stimulate the economy. In addition, my analysis of the government spending part of the stimulus is that it too had little positive impact.
Even frameworks that stress future consequences of current stimulus need not be long-run theories in the calendar sense. For example, if the anticipated taxes required to pay off or service current deficits consist of rises in marginal income tax rates, output will be considerably lower and the real interest rates higher in a matter of a couple of years than without stimulus.
The upshot of all of this is that the anti-stimulus economists are not claiming we must trade off benefits now for some long-term pie-in-the-sky benefits. Most are saying: The stimulus route leads to (almost) no benefits now as well as costs later.
If we are all dead in the long run, then Keynes-Krugman must argue that we are all dead in the short run as well, too.
All of this confusion could have been avoided if Krugman did not follow Keynes in his disrespect and disregard for those who dare to disagree.