The Re-Examining of Keynesian Economics in the Light of Employment Experience

Abstract

If the plenitude of observations since 1940 were equaled by the ease of discovering their total meaning, we should perhaps be able by a few twists of the time series to lay to rest the controversy between Keynesism and old-fashioned orthodoxy. But choice between theories requires the and one of the most difficult of tasks is the designing of such-especially those of the sort available in the laboratory of history. So it is not strange that students of society-most of us-fall into one of two categories of imperfect scientists. One group, much concerned that the bases of design be sound, rightly devote great effort to exploring the internal logic of theoretical systems but, wrongly from an ideal standpoint, often stop short of discovering new experiments or even of repeating old ones. The other group, in proper eagerness to look at the world in order to know the world, often settle cheap with regard to making sure the experiment is a significant one, and so allocate effort unduly in favor of observation or statistical manipulation. Both groups are proper objects of concern, but at present the latter will serve as my whipping boys. I shall not dwell on oversimple conclusions from experience, although, however absent they may be from the formal positions of members of such groups as this, they have, one suspects, been rather influential on the climate of thought. (I have in mind, for example, the conclusion that our high since the early forties proves Keynes wrong, or that the same thing in its relation to government spending proves Keynes right.) However, with deliberate naivete, I have for the most part followed the simple lead and taken literally employment experience as the thing from which light is presumably to be shed upon Keynesian economics, and for two reasons. First, this approach has enabled me to run through an exercise which may serve to discourage oversimple tests while restating a central part of an orientation making for sound union between reasoning and observation. Second, it has furnished me an excuse for bringing up issues which most desperately need to be fought through to consensus. We cannot, of course, assure ourselves that any theory tells the ultimate truth; and scientific study becomes a matter of trying to discover which of the many theories we can devise is the least bad one-to a large degree a matter of finding which best fits the observations. A critical experiment, then, will consist of an