In the United States, for example, each unit of new capital formation seems to bring more increase in output now than was the case in earlier periods; an increased rate of technological-organizational progress (more plentiful innovating activity) appears to be the most plausible explanation of this. There exists no good reason to assume that historically we would be running into diminishing returns on any major scale, or with much consistency, even if the saving-output or investment-output ratio were the same now as it was two or three generations ago. The saving-output or investment-output ratio has become lower, but it would be far-fetched to argue that with somewhat higher saving Western economies would be running into diminishing returns to the extent of generating chronic deflationary pressure.
In the Western world, investment opportunities have so far stayed plentiful indeed, except in periods of cyclical depression (some of which, however, were of considerable duration). Technological-organizational advance may very well prove to be a self-accelerating process. At any rate, the thesis of secular stagnation must by now explain away five decades or more as atypical decades when it represents a single decade as reflecting allegedly typical conditions in advanced industrial systems. Innovations have been sufficiently plentiful and their land and laborsaving character has stayed sufficiently pronounced to prevent any appreciable secular lowering of rates of return to investors.
Even if economic stagnation should in the future become the typical state of affairs in Western societies, the question of whether the reasons are Keynesian or Schumpeterian would, of course, stay controversial, since net rates of return to investors are influenced by redistributive political measures as well as by the relative factor supplies in our economies. But the main point here is that stagnationist predictions cannot at present properly be derived by projecting into the future the observable long-run relations of the past.
Notwithstanding favorable secular trends, the influence of stagnationist Keynesianism may wane only gradually. In fact, its influence has been on the decline for some time, but this has been a slow decline. The dangers of a stagnationist bias in professional opinion are not negligible.
Experts with a stagnationist bias are apt to suggest strong antideflationary measures when we move into a cyclical phase which later turns out to be one of minor recession; in other words, a policy of cyclical Keynesianism may become distinctly lopsided (inflationary on balance) if the policy is administered by experts who are under the influence of stagnationist Keynesianism. Furthermore, in times of severe cyclical depression such
experts may show a tendency to favor not merely those antideflationary public policies which at the same time provide incentives for a subsequent full recovery of private investment, but also some public investment policies which discount, so to speak, the inability of private investment ever to recover to its former significance.
On the whole, my own view is therefore that while stagnationist Keynesianism has raised questions of very great analytical and practical interest (and in this sense has been very fruitful), much harm could develop from a stagnationist attitude to problems of economic policy. I feel that it is exceedingly important to divorce a reasonable version of cyclical Keynesianism from stagnationist Keynesianism or, as we might prefer to put it, to hasten the gradual process by which the two are becoming divorced from one another.
The third and last meaning of Keynesianism to be considered in the present paper is that which we have called "fundamental-theoretical Keynesianism." I here mean the theoretical position that denies the Pigou effect or Patinkin's "real balance effect." The positions which in this paper were called those of cyclical Keynesianism and of stagnationist Keynesianism can afford to by-pass the fundamental question of whether, in the event of threatening demand insufficiency at a given price level, a fall in money wage rates and in prices would not always raise effective demand to the size consistent with full utilization. If we disregard problems of transition (expectational spirals and the like), perfect flexibility of money wage rates and of prices could have this beneficial consequence by raising the real value of the already existing money hoards and by thus encouraging the use of money for the purchase of goods. This was pointed out already by Haberler in the early stages of the controversy about the General Theory. Cyclical and stagnationist Keynesianism were defined as implying either that the reasoning which establishes the real balance effect of price reductions is invalid or that the reasoning, while theoretically valid, lacks practical significance. This is somewhat evasive.