I. Doubtless the best-remembered lesson of our World War experience concerns inflation. The various discomforts of rapidly advancing prices made a deep impression on people; and largely by Newtonian analogy, the post-war deflation was associated closely with the previous advance. Retrospectively, much of the price advance of the war years has been attributed to poor management, which means that the associated discomfort and dislocation could be considered unnecessary. It has been agreed that next time – i.e., this time – prices must be kept under better control. So whatever the state of feeling about the present war, any sort of poll would show a grim determination to defeat inflation.

Inflation clearly is undesirable; we should try if possible to avoid the inequitable distribution of burden, the demoralizing effects of unstable profit expansion, the arbitrary shifts in income distribution, and (so far as anyone knows) the increased difficulties of post-war readjustment, which are likely to be associated with rapid price advances.

But simply to control these price movements is not important in itself; there is nothing inherently iniquitous about a price spiral. The consequences of price behavior are the important considerations, and in avoiding these consequences we do not want others which are equally bad. This point is important for inflation can be controlled so that the effects or results of the controls are as bad or worse than the effects or results of inflation. It is possible so to select or time the controls that they handicap or restrict the use of resources for both military and civilian production; it is possible to employ controls that will increase the real burden of the armament program and make it arbitrary and regressive; and it is perfectly possible to adopt controls which will be more definite and certain in their destabilizing influence during the post-armament period than inflationary price movements in the period ahead. Professor Hansen touches on the question of inflation control in the closing section of his paper. My own comments are confined largely to this section, and in the main they are supplementary to rather than at issue with his argument.

II. In common with many others, Professor Hansen distinguishes between rising prices associated with specific bottlenecks and rising prices associated with an approach to full employment. For the first, he recommends an aggressive policy of breaking the bottlenecks by providing increased plant capacity and training additional workers, and by use of priorities, rationing, and, presumably, specific measures of price regulation. He proposes to meet general inflation by one measure or another for reducing the general volume of spending in the economy.

I believe it important at the outset to scrutinize the distinction between bottleneck inflation and general inflation. I am not sure that complete rejection of the distinction would not be wise.

The American economy has been operating at a relatively low level of resource use for a decade. Within that decade rates of obsolescence and replacement in plant capacity have been widely different in different industries. The same is true as between different classes of skilled labor. Within the framework of general excess capacity shifts in consumer tastes and requirements have taken place. Obviously, an expansion of demand associated with armament production is channelled to different industries than an expansion associated with peacetime consumption. And present expansion takes place concurrently with drastic, war-enforced changes in export markets and in our foreign sources of supply of both industrial raw materials and consumers’ goods.

In this context the idea of bottlenecks, carried over from our discussion of the process of peacetime expansion, -is an understatement of the problem. What we face is a great and progressively more difficult problem in the reorganization of resources. In my judgment, our ability to reorganize our resources – and not our aggregate resources – now becomes the limiting factor in the expansion process. The notion of the total -labor force as the limit on our total productive power has only theoretical value. The practical question is how rapidly and to what extent we can produce the radically revised requirements of a war economy in a world that is at war.

The real limiting factors, in other words, are the institutional and technical resistances to the reorganization that will permit the employment of the employable labor force. The total number of employables in Pennsylvania is not important compared with the problem of so organizing steel production that these people may be employed in filling the exaggerated demands of the war economy for steel. The size of the labor force left idle or in disguised un-
employment by the loss of cotton and tobacco markets is of little significance compared with the organizational problem of developing, say, a rubber or tin industry to replace former sources of supply of rubber and tin. At full employment the supply function for all industries, taken as a whole, has become inelastic. But as matters now stand – from the present situation to full use of the labor force -we will encounter a steadily increasing number of industries where the supply function will be inelastic while management is being induced to expand, or plant is being built, or skilled workers are being trained. The expansion process, to put it somewhat roughly, will encounter an ever widening area of resistances. The removal of each set of resistances makes it possible to move on to a new level of output and employment and brings an encounter with shortages over a larger area of industry.

III. When the problem is so projected, it becomes clear that the important decisions in price policy and fiscal policy will have to be made long before full employment is reached. A year hence fifty per cent of our industrial production may be “bottle-necked,” but the fact that unemployed or redundant labor exists elsewhere will not prevent an inflation, wholly indistinguishable from Professor Hansen’s general inflation, from manifesting itself in the resistance area. And that the resistances could eventually be removed by new capacity, new sources of materials, or new supplies of skilled labor (in short, by aggressive action in breaking bottle-necks on all fronts) will not alter the situation. We shall have price advances in the interim; and, if my analysis is correct, the new output, when it comes, will merely move us along to the next set of resistances. Full employment will have little or no relation to the appearance of inflation.

This means that the question of timing in the application of inflation controls is considerably more important and quite a bit more difficult than Professor Hansen implies. If we wait for full employment before reducing the volume of spending we shall already have had a good deal of inflation. But the danger of reducing the amount of spending too soon is equally serious. That inflation can be checked by a sufficient over-all reduction of expenditures is not in doubt; but this measure has the crudity in application of any general or blanket measure. Further, if it is applied too soon it will check inflation by checking the whole expansion process. As a reductio ad absurdum, the price of aluminum scrap, an item with inflationary tendencies at the moment, could now theoretically be reduced by taxing enough out of consumer income so that purchases of pots, pans, and other aluminum items would be checked. But the astringent effect would be the same for cotton, tobacco, and food (to mention a few of many commodities which offer no conceivable reason for reducing consumer purchases); and equally important, it would remove a very desirable pressure for expansion of capacity and elimination of resistances. Could we look forward to a reduction of consumer demand for steel during the coming year, the pressure for expansion of this industry would be greatly eased. If consumers cannot buy, the effective criteria of expansion – viz., large backlogs, long delayed deliveries and outright rationing-would not be present. For the inequitable distribution of burden associated with inflation, we should be substituting, simply, an unnecessary sacrifice in real income. And to the extent that our total steel capacity is a factor in our total armament potential, we should be handicapping our military strength as well.

Obviously then, even at some risk we must avoid premature reduction of spending. It is not something to be done “at the first sign of inflation.” It is not clearly desirable even when inflationary tendencies become relatively wide-spread, for at such time, if my analysis is correct, there will still be important unused resources. Can we obtain a reasonably full use of resources without inflation?

IV. Reasonably full use of resources without serious inflation can be-achieved; but we cannot rely entirely, or even in major part, upon measures which reduce the general volume of spending in the economy. We shall need a portfolio of measures appropriate to different stages in the expansion process.

First and at all stages we need the most vigorous efforts to reorganize resources to the demands of war economy. This reorganization-expansion of capacity, provision of necessary skills, development of domestic sources of materials supply, and so forth-should anticipate rather than wait upon the development of resistance points in the expansion. Our showing along these lines to date has been disappointing.

Second in the areas where resistances develop, as they are now developing, we shall need specific price controls or price fixing. These controls will necessarily be accompanied by the use of priorities for such part of the supply as may be required for military purposes and by a greater or less measure of civilian rationing.

Until the expansion process is well advanced these specific price controls must be the major reliance. They cannot be expected to be completely effective. To control any considerable number of prices in all of their dimensions is a Gargantuan task. The results are almost certain to look irregular or even erratic. But this control, assuming proper consideration of supply elasticity, will check inflation without hampering expansion or curbing the consumption of commodities or the use of services which are plentiful. In short, the real burden of the armament effort can be reduced by price controls, and a- careful organization of this phase of the defense effort is most worthwhile.

Before a general reduction of spending is implemented as a control measure, a third step should at least receive careful consideration. This step is the withdrawal of expenditure in specific areas for specific types of consumers’ or capital goods. One possibility of this sort would be a limitation or ban on consumer credit for -goods where supply is temporarily or permanently inelastic. Such a measure would be applicable, for example, to automobiles, refrigerators, or to home construction. At this same stage licensing of certain types of new investment would also be appropriate. The limitation would be confined to investment items which are not essential either directly or indirectly for defense production and which do not act directly to remove shortages in existing plant for supplying scarce consumers’ goods. Office buildings, certain types of public works, or plant and machinery investment which is merely to reduce costs or implement competitive changes in design, are examples. Reduction of these classes of investment will, via the multiplier, reduce the aggregate of consumer spending in the economy. But they have the greater advantage that the primary curtailment can be addressed to specific bottleneck areas.

Finally, in this array of control measures, come those which seek to check the increase or to reduce the volume of spending in the economy as a whole. These measures should be taken only when supply is becoming inelastic over a very substantial portion of the economy and when the prospect of relief is comparatively small. But that the criterion in applying these measures will be the approach to full employment is doubtful. Rather, it will be the inability any longer to maintain control over price movements by specific measures. To the extent that specific measures are successful, the longer the general withdrawal of spending may be delayed. Possibly one of the best clues to the timing of this measure will be the prospect of a sustained (as contrasted with a temporary) upward movement in the prices of those commodities that are not affected by defense procurement.

Considerable discussion of alternative methods of effecting this general control has followed Mr. Keynes’ proposals in England. While I cannot comment on these measures here, I would like to make one suggestion. It is that at this stage, as at other stages, the question of price or inflation control be kept separate, where possible, from the question of revenue. Beginning immediately, much may be said for increasing taxes — the proper taxes. This increase is necessary because it is expected and demanded by the public, because it avoids the danger of further attenuating the unbalance in income distribution in a period of high profits, and because it seems wise at least to minimize the rate of increase in the transfer required on public debt account. But the volume of revenue raised, or the size of the deficit, bears little relation to the problem of inflation. An inflation is theoretically possible with a balanced budget just as stable prices are wholly consistent with a sizeable deficit.

For my own part I would go one step further and suggest that we should be very careful about using the revenue structure for the control of inflation. Deflationary taxes are precisely the taxes which, as a matter of long time practice, we want least in the revenue system. If revenue and inflation control are tied together during the armament period, they may stay tied together afterward. In other words we will emerge from this period with a revenue structure loaded with consumer taxes. These taxes might save us from the inflation only to leave us with a diminished propensity to consume much worse in its effects than the aftermath of inflation itself.

— J.K. Galbraith Princeton University May 1, 1941