Can We Still Avoid Inflation?
Friedrich A. Hayek
This essay was originally given as a lecture before the Trustees and guests of the Foundation for Economic Education at Tarrytown, New York on May 18, 1970, and was first published in the first edition of this book (1978).
The current edition: Copyright 1996 by the Ludwig von Mises Institute, Auburn, Alabama 36849-5301.
In one sense the question asked in the title of this lecture is purely rhetorical. I hope none of you has suspected me of doubting even for a moment that technically there is no problem in stopping inflation. If the monetary authorities really want to and are prepared to accept the consequences, they can always do so practically overnight. They fully control the base of the pyramid of credit, and a credible announcement that they will not increase the quantity of bank notes in circulation and bank deposits, and, if necessary, even decrease them, will do the trick. About this there is no doubt among economists. What I am concerned about is not the technical but the political possibilities. Here, indeed, we face a task so difficult that more and more people, including highly competent people, have resigned themselves to the inevitability of indefinitely continued inflation. I know in fact of no serious attempt to show how we can overcome these obstacles which lie not in the monetary but in the political field. And cannot myself claim to have a patent medicine which I am sure is applicable and effective in the prevailing conditions. But do not regard it as a task beyond the scope of human ingenuity once the urgency of the problem is generally understood. My main aim tonight is to bring out clearly why we must stop inflation if we are to preserve a viable society of free men. Once this urgent necessity is fully understood, I hope people will also gather the courage to grasp the hot irons which must be tackled, if the political obstacles are to be removed and we are to have a chance of restoring a functioning market economy.
In the elementary textbook accounts, and probably also in the public mind generally, only one harmful effect of inflation is seriously considered, that on the relations between debtors and creditors. Of course, an unforeseen depreciation of the value of money harms creditors and benefits debtors This is important but by no means the most important effect of inflation. And since it is the creditors who are harmed and the debtors who benefit, most people do not particularly mind, at least until they realize that in modern society the most important and numerous class of creditors are the wage and salary earners and the small savers, and the representative groups of debtors who profit in the first instance are the enterprises and credit institutions.
But I do not want to dwell too long on this most familiar effect of inflation which is also the one which most readily corrects itself. Twenty years ago I still had some difficulty to make my students believe that if an annual rate of price increase of five per cent were generally expected, we would have rates of interest of 9-10 per cent or more. There still seem to be a few people who have not yet understood that rates of this sort are bound to last so long as inflation continues. Yet, so long as this is the case, and the creditors understand that only part of their gross return is net return, at least short term lenders have comparatively little ground for complaint -- even though long term creditors, such as the owners of government loans and other debentures, are partly expropriated.
There is, however, another more devious aspect of this process which I must at least briefly mention at this point. It is that it upsets the reliability of all accounting practices and is bound to show spurious profits much in excess to true gains. Of course, a wise manager could allow for this also, at least in a general way, and treat as profits only what remains after he has taken into account the depreciation of money as affecting the replacement costs of his capital. But the tax inspector will not permit him to do so and insist on taxing all the pseudo-profits. Such taxation is simply confiscation of some of the substance of capital, and in the case of a rapid inflation may become a very serious matter.
But all this is familiar ground -- matters of which I merely wanted to remind you before turning to the less conspicuous but, for that very reason, more dangerous effects of inflation. The whole conventional analysis reproduced in most textbooks proceeds as if a rise in average prices meant that all prices rise at the same time by more or less the same percentage, or that this at least was true of all prices determined currently on the market, leaving out only a few prices fixed by decree or long term contracts, such as public utility rates, rents and various conventional fees. But this is not true or even possible, The crucial point is that so long as the flow of money expenditure continues to grow and prices of commodities and services are driven up, the different prices must rise, not at the same time but in succession, and that in consequence, so long as this process continues, the prices which rise first must all the time move ahead of the others. This distortion of the whole price structure will disappear only sometime after the process of inflation has stopped. This is a fundamental point which the master of all of us, Ludwig von Mises, has never tired from emphasizing for the past sixty years. It seems nevertheless necessary to dwell upon it at some length since, as I recently discovered with some shock, it is not appreciated and even explicitly denied by one of the most distinguished living economists [Sir John Hicks is referenced in a footnote here].
That the order in which a continued increase in the money stream raises the different prices is crucial for an understanding of the effects of inflation was clearly seen more than two hundred years ago by David Hume -- and indeed before him by Richard Cantillon. It was in order deliberately to eliminate this effect that Hume assumed as a first approximation that one morning every citizen of a country woke up to find the stock of money in his possession miraculously doubled. Even this would not really lead to an immediate rise of all prices by the same percentage. But it is not what ever really happens. The influx of the additional money into the system always takes place at some particular point. There will always be some people who have more money to spend before the others. Who these people are will depend on the particular manner in which the increase in the money stream is being brought about. It may be spent in the first instance by government on public works or increased salaries, or it may be first spent by investors mobilizing cash balances or borrowing for the purpose; it may be spent in the first instance on securities, on investment goods, on wages or on consumer's goods. It will then in turn be spent on something else by the first recipients of the additional expenditure, and so on. The process will take very different forms according to the initial source or sources of the additional money stream; and all its ramifications will soon be so complex that nobody can trace them. But one thing all these different forms of the process will have in common: that the different prices will rise, not at the same time but in succession, and that so long as the process continues some prices will always be ahead of the others and the whole structure of relative prices therefore very different from what the pure theorist describes as an equilibrium position. There will always exist what might be described as a prices gradient in favor of those commodities and services which each increment of the money stream hits first and to the disadvantage of the successive groups which it reaches only later -- with the effect that what will rise as a whole will not be a level but a sort of inclined plane -- if we take as normal the system of prices which existed before inflation started and which will approximately restore itself sometime after it has stopped.
To such a change in relative prices, if it has persisted for some time and comes to be expected to continue, will of course correspond a similar change in the allocation of resources: relatively more will be produced of the goods and services whose prices are now comparatively higher and relatively less of those whose prices are comparatively lower. This redistribution of the productive resources will evidently persist so long, but only so long, as inflation continues at a given rate.
We shall see that this inducement to activities, or a volume of some activities, which can be continued only if inflation is also continued, is one of the ways in which even a contemporary inflation places us in a quandary because its discontinuance will necessarily destroy some of the jobs it has created.
But before I turn to those consequences of an economy adjusting itself to a continuous process of inflation, I must deal with an argument that, though I do not know that it has anywhere been clearly stated, seems to lie at the root of the view which represents inflation as relatively harmless. It seems to be that, if future prices are correctly foreseen, any set of prices expected in the future is compatible with an equilibrium position, because present prices will adjust themselves to expected future prices. For this it would, however, clearly not be sufficient that the general level of prices at the various future dates be correctly foreseen, and these, as we have seen, will change in different degrees. The assumption that the future prices of particular commodities can be correctly foreseen during a period of inflation is probably an assumption which never can be true: because, whatever future prices are foreseen, present prices do not by themselves adapt themselves to the expected higher prices of the future, but only through a present increase in the quantity of money with all the changes in the relative height of the different prices which such changes in the quantity of money necessarily involve.
More important, however, is the fact that if future prices were correctly foreseen, inflation would have none of the stimulating effects for which it is welcomed by so many people.
Now the chief effect of inflation which makes it at first generally welcome to business is precisely that prices of products turn out to be higher in general than foreseen. It is this which produces the general state of euphoria, a false sense of wellbeing, in which everybody seems to prosper. Those who without inflation would have made high profits make still higher ones. Those who would have made normal profits make unusually high ones. And not only businesses which were near failure but even some which ought to fail are kept above water by the unexpected boom. There is a general excess of demand over supply -- all is saleable and everybody can continue what he had been doing. It is this seemingly blessed state in which there are more jobs than applicants which Lord Beveridge defined as the state of full employment -- never understanding that the shrinking value of his pension of which he so bitterly complained in old age was the inevitable consequence of his own recommendations having been followed.
But, and this brings me to my next point, "full employment" in his sense requires not only continued inflation but inflation at a growing rate. Because, as we have seen, it will have its immediate beneficial effect only so long as it, or at least its magnitude, is not foreseen. But once it has continued for some time, its further continuance comes to be expected. If prices have for some time been rising at five percent per annum, it comes to be expected that they will do the same in the future. Present prices of factors are driven up by the expectation of the higher prices for the product -- sometimes, where some of the cost elements are fixed, the flexible costs may be driven up even more than the expected rise of the price of the product -- up to the point where there will be only a normal profit.
But if prices then do not rise more than expected, no extra profits will be made. Although prices continue to rise at the former rate, this will no longer have the miraculous effect on sales and employment it had before. The artificial gains will disappear, there will again be losses, and some firms will find that prices will not even cover costs. To maintain the effect inflation had earlier when its full extent was not anticipated, it will have to be stronger than before. If at first an annual rate of price increase of five percent had been sufficient, once five percent comes to be expected something like seven percent or more will be necessary to have the same stimulating effect which a five percent rise had before. And since, if inflation has already lasted for some time, a great many activities will have become dependent on its continuance at a progressive rate, we will have a situation in which, in spite of rising prices, many firms will be making losses, and there may be substantial unemployment. Depression with rising prices is a typical consequence of a mere braking of the increase in the rate of inflation once the economy has become geared to a certain rate of inflation.
All this means that, unless we are prepared to accept constantly increasing rates of inflation which in the end would have to exceed any assignable limit, inflation can always give only a temporary fillip to the economy, but must not only cease to have stimulating effect but will always leave us with a legacy of postponed adjustments and new maladjustments which make our problem more difficult. Please note that I am not saying that once we embark on inflation we are bound to be drawn into a galloping hyper-inflation. I do not believe that this is true. All I am contending is that if we wanted to perpetuate the peculiar prosperity-and-job-creating effects of inflation we would have progressively to step it up and must never stop increasing its rate. That this is so has been empirically confirmed by the Great German inflation of the early 1920s. So long as that increased at a geometrical rate there was indeed (except towards the end) practically no unemployment. But till then every time merely the increase of the rate of inflation slowed down, unemployment rapidly assumed major proportions. I do not believe we shall follow that path -- at least not so long as tolerably responsible people are at the helm -- though I am not quite so sure that a continuance of the monetary policies of the last decade may not sooner or later create a position in which less responsible people will be put into command. But this is not yet our problem. What we are experiencing is still only what in Britain is known as the "stop-go" policy in which from time to time the authorities get alarmed and try to brake, but only with the result that even before the rise of prices has been brought to a stop, unemployment begins to assume threatening proportions and the authorities feel forced to resume expansion. This sort of thing may go on for quite some time, but I am not sure that the effectiveness of relatively minor doses of inflation in rekindling the boom is not rapidly decreasing. The one thing which, I will admit, has surprised me about the boom of the last twenty years is how long the effectiveness of resumed expansion in restarting the boom has lasted. My expectation was that this power of getting investment under way by a little more credit expansion would much sooner exhaust itself -- and it may well be that we have now reached that point. But I am not sure. We may well have another ten years of stop-go policy ahead of us, probably with decreasing effectiveness of the ordinary measures of monetary policy and longer intervals of recessions. Within the political framework and the prevailing state of opinion the present chairman of the Federal Reserve Board will probably do as well as can be expected by anybody. But the limitations imposed upon him by circumstances beyond his control and to which I shall have to turn in a moment may well greatly restrict his ability of doing what we would like to do.
On an earlier occasion on which several of you were present, I have compared the position of those responsible for monetary policy after a full employment policy has been pursued for some time to "holding a tiger by the tail. " It seems to me that these two positions have more in common than is comfortable to contemplate. Not only would the tiger tend to run faster and faster and the movement bumpier and bumpier as one is dragged along, but also the prospective effects of letting go become more and more frightening as the tiger becomes more enraged. That one is soon placed in such a position is the central objection against allowing inflation to run on for some time. Another metaphor that has often been justly used in this connection is the effects of drug-taking. The early pleasant effects and the later necessity of a bitter choice constitute indeed a similar dilemma. Once placed in this position it is tempting to rely on palliatives and be content with overcoming short-term difficulties without ever facing the basic trouble about which those solely responsible for monetary policy indeed can do little.
Before I proceed with this main point, however, I must still say a few words about the alleged indispensability of inflation as a condition of rapid growth. We shall see that modern developments of labor union policies in the highly industrialized countries may there indeed have created a position in which both growth and a reasonably high and stable level of employment may, so long as those policies continue, make inflation the only effective means of overcoming the obstacles created by them. But this does not mean that inflation is, in normal conditions, and especially in less developed countries, required or even favorable for growth. None of the great industrial powers of the modern world has reached its position in periods of depreciating money. British prices in 1914 were, so far as meaningful comparisons can be made over such long periods, just about where they had been two hundred years before, and American prices in 1939 were also at about the same level as at the earliest point of time for which we have data, 1749. Though it is largely true that world history is a history of inflation, the few success stories we find are on the whole the stories of countries and periods which have preserved a stable currency; and in the past a deterioration of the value of money has usually gone hand in hand with economic decay.
There is of course, no doubt that temporarily the production of capital goods can be increased by what is called "forced saving" -- that is, credit expansion can be used to direct a greater part of the current services of resources to the production of capital goods. At the end of such a period the physical quantity of capital goods existing will be greater than it would otherwise have been. Some of this may be a lasting gain -- people may get houses in return for what they were not allowed to consume. But I am not so sure that such a forced growth of the stock of industrial equipment always makes a country richer, that is, that the value of its capital stock will afterwards be greater -- or by its assistance all-round productivity be increased more than would otherwise have been the case. If investment was guided by the expectation of a higher rate of continued investment (or a lower rate of interest, or a higher rate of real wages, which all come to the same thing) in the future than in fact will exist, this higher rate of investment may have done less to enhance overall productivity than a lower rate of investment would have done if it had taken more appropriate forms. This I regard as a particularly serious danger for underdeveloped countries that rely on inflation to step up the rate of investment. The regular effect of this seems to me to be that a small fraction of the workers of such countries is equipped with an amount of capital per head much larger than it can hope within the foreseeable future to provide for all its workers, and that the investment of the larger total in consequence does less to raise the general standard of living than a smaller total more widely and evenly spread would have done. Those who counsel underdeveloped countries to speed up the rate of growth by inflation seem to me wholly irresponsible to an almost criminal degree. The one condition which, on Keynesian assumptions, makes inflation necessary to secure a full utilization of resources, namely the rigidity of wage rates determined by labor unions, is not present there. And nothing I have seen of the effects of such policies, be it in South America, Africa, or Asia, can change my conviction that in such countries inflation is entirely and exclusively damaging -- producing a waste of resources and delaying the development of that spirit of rational calculation which is the indispensable condition of the growth of an efficient market economy.
The whole Keynesian argument for an expansionist credit policy rests entirely and completely on the existence of that union determined level of money wages which is characteristic of the industrially advanced countries of the West but is absent in underdeveloped countries -- and for different reasons less marked in countries like Japan and Germany. It is only for those countries where, as it is said, money wages are "rigid downward" and are constantly pushed up by union pressure that a plausible case can be made that a high level of employment can be maintained only by continuous inflation -- and I have no doubt that we will get this so long as those conditions persist. What has happened here at the end of the last war has been that principles of policy have been adopted, and often embodied in the law, which in effect release unions of all responsibility for the unemployment their wage policies may cause and place all responsibility for the preservation of full employment on the monetary and fiscal authorities. The latter are in effect required to provide enough money so that the supply of labor at the wages fixed by the unions can be taken off the market. And since it cannot be denied that at least for a period of years the monetary authorities have the power by sufficient inflation to secure a high level of employment, they will be forced by public opinion to use that instrument. This is the sole cause of the inflationary developments of the last twenty-five years, and it will continue to operate as long as we allow on the one hand the unions to drive up money wages to whatever level they can get employers to consent to -- and these employers consent to money wages with a present buying power which they can accept only because they know the monetary authorities will partly undo the harm by lowering the purchasing power of money and thereby also the real equivalent of the agreed money wages.
This is the political fact which for the present makes continued inflation inevitable and which can be altered not by any changes in monetary but only by changes in wage policy. Nobody should have any illusion about the fact that so long as the present position on the labor market lasts we are bound to have continued inflation. Yet we cannot afford this, not only because inflation becomes less and less effective even in preventing unemployment, but because after it has lasted for some time and comes to operate at a high rate, it begins progressively to disorganize the economy and to create strong pressure for the imposition of all kinds of controls. Open inflation is bad enough, but inflation repressed by controls is even worse: it is the real end of the market economy.
The hot iron which we must grasp if we are to preserve the enterprise system and the free market is, therefore, the power of the unions over wages. Unless wages, and particularly the relative wages in the different industries, are again subjected to the forces of the market and become truly flexible, in particular groups downwards as well as upwards, there is no possibility for a non-inflationary policy. A very simple consideration shows that, if no wage is allowed to fall, all the changes in relative wages which become necessary must be brought about by all the wages except those who tend to fall relatively most being adjusted upwards. This means that practically all money wages must rise if any change in the wage structure is to be brought about. Yet a labor union conceding a reduction of the wages of its members appears today to be an impossibility. Nobody, of course, gains from this situation, since the rise in money wages must be offset by a depreciation of the value of money if no unemployment is to be caused. It seems, however, a built-in necessity of that determination of wages by collective bargaining by industrial or craft unions plus a full employment policy.
I believe that so long as this fundamental issue is not resolved, there is little to be hoped from any improvement of the machinery of monetary control. But this does not mean that the existing arrangements are satisfactory They have been designed precisely to make it easier to give in to the necessities determined by the wage problem, i.e., to make it easier for each country to inflate. The gold standard has been destroyed chiefly because it was an obstacle to inflation. When in 1931 a few days after the suspension of the gold standard in Great Britain Lord Keynes wrote in a London newspaper that "there are few Englishmen who do not rejoice at the breaking of our gold fetters," and fifteen years later could assure us that Bretton Woods arrangements were "the opposite of the gold standard," all this was directed against the very feature of the gold standard by which it made impossible any prolonged inflationary policy of any one country. And though I am not sure that the gold standard is the best conceivable arrangement for that purpose, it has been the only one that has been fairly successful in doing so. It probably has many defects, but the reason for which it has been destroyed was not one of them; and what has been put into its place is no improvement. If, as I have recently heard it explained by one of the members of the original Bretton Woods group, their aim was to place the burden of adjustment of international balances exclusively on the surplus countries, it seems to me the result of this must be continued international inflation. But I only mention this in conclusion to show that if we are to avoid continued world-wide inflation, we need also a different international monetary system. Yet the time when we can profitably think about this will be only after the leading countries have solved their internal problems. Till then we probably have to be satisfied with makeshifts, and it seems to me that at the present time, and so long as the fundamental difficulties I have considered continue to be present, there is no chance of meeting the problem of international inflation by restoring an international gold standard, even if this were practical policy. The central problem which must be solved before we can hope for a satisfactory monetary order is the problem of wage determination.