Inflation and Deflation

views updated

Inflation and Deflation

Theoretical aspects

Empirical aspects

Redistributive effects of inflation


Inflation is a fall and deflation is a rise in the purchasing power of money, as measured ordinarily by an index number of prices. When the price index rises, economists speak of the purchasing power of money falling—of inflationary conditions—and the converse when this index falls. Associated expressions relating to the external purchasing power of money (that is, its purchasing power in terms of foreign exchange) are devaluation and upward revaluation (or sometimes up-valuation). When a currency loses purchasing power in terms of foreign currency, by reason of inflation at home or deflation abroad, it is effectively devalued; and again, .the converse. The order of events may also be reversed, with devaluation leading to inflation through rises in the import components of the domestic price level; upward revaluation may have deflationary consequences through declines in the import components.

On the side of inflation, particularly, it has become customary to speak of creeping, trotting, galloping, and runaway inflation (also called hyperinflation). Clear distinctions are lacking. However, a general price increase of less than 1 per cent per quarter (approximately 4 per cent per year) is often considered a creeping inflation, despite its eventual consequences if unchecked. At the other extreme, Phillip Cagan’s definition of hyperinflation as a price rise of 50 per cent a month has received wide acceptance (1956).

Despite substantial agreement in principle as to the definitions of inflation and deflation, there remains disagreement in detail, a disagreement symptomatic of some confusion in the general theory of the subject. We list some issues often slurred over by the elementary definitions:

(1)Which of an infinite number of possible price levels is used to measure the purchasing power of money? Much recent literature (since World War II) has tended to use a national income or gross national product (GNP) “deflator,” whereas earlier literature often relied on wholesale price indexes. In many cases price indexes will show widely differing rates of inflation or deflation, and the differences may become political issues. When wages, for example, are rising more rapidly than wholesale prices, the income deflators and consumer price indexes will also rise more rapidly.

(2)How are index numbers to be computed? In particular, what allowance should be made for the tendency of buyers to shift their purchases in the direction of those products that have fallen most or risen least in price from one period to the next; and what adjustment should be made for possible upward biases in price indexes, reflecting increases in living standards? These problems are often solved in practice by using so-called Laspeyres price indexes, in which price quotations are “weighted” by consumption in the “base” period. An upward bias results, which is reduced by revising base periods relatively frequently. [Seeindex numbers.]

(3)What allowance should be made for new products, quality improvements, product availabilities. and so on? In the United States an AFLCIO task force emphasized these factors during World War II, alleging that the official indexes understated the magnitude of wartime inflation. In the late 1950s a National Bureau of Economic Research (1961) task force headed by George Stigler argued that the indexes overstated the rate of American inflation after the Korean War by confusing product improvement with inflation.

(4)When price controls are in effect, should the indexes be based on black market as well as on official prices? The terms repressed inflation and suppressed inflation are used for cases in which only official prices are used, as distinguished from open inflation, when controls are not in effect or when black market prices are used to compute price index numbers. An unsettled issue is whether repressed inflation and suppressed inflation are most usefully considered as types of inflation or as alternatives thereto.

(5)When commodity taxes, sales taxes, and subsidies are used widely, should prices be taken gross or net of such taxes and subsidies? Suppose also that, as in France before 1959, such taxes and subsidies are influenced by price index considerations, with commodities in the official indexes treated differently from others. How reliable are price indexes under this sort of sampling bias?

(6)Following destruction and disruption due to war, flood, earthquake, or other calamity, there are often sharp rises in prices. Is such a result to be called inflationary when no change in monetary or fiscal policy has occurred? Does it make any difference, in answering the last question, whether the same destruction and disruption simultaneously create unemployment?

(7)For any of a number of reasons, including income redistribution and “money illusion,” a rise in the price level may lead to at least a short-run increase in real output, in employment, and in economic growth [seeincome and employment theory]. Is such a price increase really inflationary, or should the term be limited to cases in which no significant increases in output or employment occur?

(8)During the 1920s technical progress in western Europe and North America brought about widespread reduction in production costs with no corresponding reduction in prices (in most countries). The resulting increases in money wages and profits were later called inflationary by Friedrich Hayek and other adherents (primarily Austrian) of the “neutral money” school. Is this a defensible position?

Our answers to these and related questions are often affected by extraneous issues, such as whether we consider inflation or commodity taxes or price controls to be a “good thing” or not. If one considers inflation, for example, a “bad thing,” it is difficult to avoid a question-begging definition that will render “good results” somehow noninflationary even when accompanied by price level increases.

For those who cannot accept the evidence of price indexes alone, a number of alternative definitions of inflation and deflation are available. We cite four, but the list is not exhaustive.

(1)Inflation is a condition of generalized excess demand for stocks of goods and flows of real income, in which “too much money chases too few goods.” Conversely, deflation is a condition of generalized excess demand for stocks and flows of money, in which “too many goods chase too little money.”

(2)Inflation is a rise of the money stock or money income in a society, either total or per capita. Conversely, deflation is a fall in one or more of these magnitudes. These definitions are favored by the apostles of monetary neutrality.

(3)Inflation (deflation) is a rise (fall) in price levels with a number of additional characteristics or conditions: it does not increase real output and employment; it leads (through cost changes) to further price movements; it is faster than some “safe” rate; it arises “from the side of money”; it is measured by prices net of commodity taxes and subsidies; it has been imperfectly anticipated.

(4)Ralph Turvey (1949) has formulated a comprehensive definition of inflation, which has no obvious counterpart on the deflationary side. To Turvey inflation is the process resulting from competition in attempting to maintain total real income, total real expenditure, and/or total output at a level which has become physically impossible or attempting to increase any of them to a level which is physically impossible.

Historical record. During the century preceding World War I, periods of inflation and deflation largely canceled each other in the Western world. It became customary to consider price levels as stable in the long run, to assume that “what goes up must come down,” and to employ accounting conventions based on the principle that “a dollar is a dollar.” These views, and plans based upon them, embody what has come to be known as the “money illusion.”

There have in fact been relatively few periods of equally protracted balance between inflationary and deflationary forces. Currency debasement, gold and silver discoveries, fractional reserve banking, and, more recently, the appeal of “full employment at whatever cost” have more than outweighed periods of strict adherence to inflexible metallic standards. The secular trend of prices has been upward, perhaps as a social rebuttal to the mathematical force of compound interest. This has led Murray Rothbard, presenting “The Case for a 100 Per Cent Gold Dollar,” to state: “The natural tendency of the state is inflation” (1962, p. 109). This may be an exaggeration, but consider the pound sterling (originally a pound of silver, at 90.5 cents per fine ounce), currently pegged at $2.80; the French franc (formerly “livre,” meaning “pound”), currently worth slightly more than 20 cents after a nominal hundredfold upward revaluation; and the Italian lira (with the same meaning), currently 625 to the dollar. Or consider the evaporation into the infinitesimal of such values as the British farthing, the Italian centesimo, the Japanese rin and sen, etc. Developments from 1939 to 1951 have been summarized on a close to world-wide basis by Arthur J. Brown in The Great Inflation (1955), while Lester Chandler has treated American developments in Inflation in the United States, 1940-1948 (1951). Studies of many past hyperinflations are also available, a notable one being Constantino Bresciani-Turroni’s Economics of Inflation (1931). The principal interruptions of the inflationary trend since 1945 have occurred in centrally planned economies, such as the Soviet Union and the People’s Republic of China. These countries have held official prices down, and often lowered them, by rationing and by heavy commodity taxes (not included in the price quotations). It is with particular respect to these countries that the question whether a suppressed inflation is a form of inflation or an alternative thereto is most significant.

In discussing the theory of price level changes, we shall therefore follow contemporary practice and concentrate on inflationary cases.

Theoretical aspects

We may classify economic theories of price level movement into three principal types:

(1) Demand theories.

(a)Demand for nonmonetary assets: the quantity theory of money.

(b)Demand for real income.

(2)Cost theories.

(a)Wage-price spiral.

(b)Administrative inflation.

(c)Sectoral inflation.

(3)Structural theories.

(a)Resource disproportionality.

(b)Economic growth.

(c)Income distribution.

Demand theories

The quantity theory of money may be classified as a demand theory—the demand for nonmonetary assets.

Quantity theory of money. Prior to 1929 the dominant explanation of price level movements was the quantity theory of money [seemoney]. If, in the so-called equation (identity) of exchange devised by Irving Fisher, we let M represent the quantity of money, including bank deposits and (sometimes) time deposits; V the number of times per period a unit of money, including bank deposits, changes hands; P the price level; Y the level of real income; and T the volume of real transactions, we have

MVv = PvY and MVt = PtT,

where the subscripts y and t relate to income and transactions respectively. (For example, transactions involving the purchase of existing securities and houses are taken into account in Vt and Pt but not in Vy and Py.) To return to the quantity theory: if we suppose V, P, and T approximately constant, we then have, for any periods 1 and 2, (P2/P1) = (M2/M1). Over short periods, this involves some difficulty because velocity rises in periods of prosperity and falls in periods of depression, and over long periods there is difficulty because of secular growth in Y and T, not to mention a possible secular fall in velocity. Nevertheless, the quantity theory usually yields accurate forecasts of price level changes, both in this elementary form and in the more advanced one (developed in Friedman 1956), in which other variables, principally wealth, interest rates, and the proportion of “human” to non-human wealth, are introduced explicitly. The implication of the quantity theory is that price level movements can be controlled by varying M, but it does not in itself explain why M varies as it does. One common criticism, in fact, interprets the level of prices as an active factor and the volume of money as a passive one, adjusting to meet the so-called needs of trade as prices change.

The quantity theory of money is classified as a demand theory because of the psychology underlying its operation: As the nominal amount of money expands (in the inflationary case), people who hold the increased stock of money desire to exchange some part of the increase for other assets,

meaning consumer goods, investment goods, and securities. Their attempts to “balance their portfolios” increase the demand for these alternatives; the prices of consumer and investment goods rise, and the yields (interest rates) on securities fall.

Demand for real income. The exposition in this article of the quantity theory approach to inflation and deflation has been based on the work of Fisher and Friedman. In the same way, the income theory approach, embodying the notions of inflationary and deflationary gaps, is due to John Maynard Keynes (1940) and his disciples in the decade 1936 to 1946.

In Figure 1 money income is measured along the horizontal axis and money expenditures along the vertical axis. The resulting expenditure function is called (C + I + G), since it represents the vertical sum of three principal components: personal consumption (C), private domestic investment (I), and government expenditures for goods and services (G). The equilibrium money income level is Y, since only at Y does aggregate income equal aggregate expenditures in money terms. (The 45-degree line through the origin 0 indicates this equality.) Suppose, however, that the capacity of the economy at full employment were only X0, measured in the prices of the preceding period. At the full employment income level, Y0, there is an inflationary gap, A0B0, with aggregate expenditures exceeding aggregate capacity. This gap causes a rise in prices. If the initial price increase raises nominal capacity (measured in new prices) to X1 full employment money income becomes Y1 and the inflationary gap falls to A1B1. The process continues until nominal capacity rises to X2, full employment income rises to Y2 (= Y), and the inflationary gap is eliminated.

A difficulty with this analysis is that the expenditure function (C + I + G) does not seem to be a stable function of money income. It tends, in most cases, to rise as money income rises; a representative consumer, with a fixed $5,000 income, who spent $4,000 at a price level of 100, for example, may spend $4,500 in an effort to maintain his accustomed living standard when the level rises to 120; $5,000 when it rises to 150; and $5,500 (dipping into capital) when it rises to 200. When we make allowance for the instability of the expenditure function in money terms, it is not at all clear that the inflationary gap is in fact reduced by price increases or that any equilibrium price level exists.

A more sophisticated variant of the income theory is represented by Figure 2. Here the income measured along the horizontal axis is real income, deflated for price changes; and the vertical axis represents “the” price level. If the quantity of money is held constant, the demand for income (D0, D1) slopes downward. This means that more is demanded at lower prices than at higher prices. One rationalization of this result is that aggregate demand depends on wealth as well as income and that wealth moves with “real cash balances,” that is, with the constant nominal amount of money deflated by the price level. At the same time, with constant money wage rates, the aggregate supply curve, S, slopes upward with the price level until a capacity ceiling is reached (a vertical line on

Figure 2). The argument runs here that an increase in prices increases profits and thence the demand for productive resources, particularly labor. The upward slope of S is generally slight until the point A is reached, christened by A. P. Lerner “low full employment.” Subsequently it increases more rapidly, becoming almost vertical at B, which corresponds to Lerner’s “high full employment” (1951).

The policy implication of income theory is that major emphasis should be placed on fiscal policy, particularly public expenditures and taxation, in controlling price level movements by shifting the aggregate demand curve D to the right or left, as the case may be.

Cost theories

While the two positions we have just examined consider price level movements as resulting from shifts in demand for assets and incomes, respectively, the view has gathered strength since the end of the Korean War (particularly from the persistence of price level increases in periods of mild recession) that the Western world is faced with a “new inflation,” whose motive force comes primarily from the cost or supply side.

Wage-price spiral. The most general form of cost-inflation argument is known by a number of titles, “cost-push,” “wage-price spiral,” and so on. The argument is that organized economic pressure groups raise prices in excess of any rise in productivity, particularly when full employment is approached, and leave the government with the responsibility of preventing any fall in real output and in employment that might normally result. In many forms of this argument one particular pressure group, the trade unions, is assigned major responsibility for inflation, since its objective is considered a dynamic one (maximum wage increase per period) rather than a static maximization of income or profits. Sidney Weintraub (1959) represents this position most straightforwardly, on the basis of statistics showing net sales (gross sales minus purchases) to be a near-constant multiple of payrolls.

Figure 3 illustrates this position graphically. It is similar to Figure 2, except that the supply function shifts upward as shown. Starting at P0, the price level rises to P1 and P2 if real income is maintained at Y0, by increases in aggregate demand from D0 to D1 and D2. If aggregate demand is held at D0, real income falls from Y0 to Y1 and Y1, and employment likewise falls; the price rise, while reduced, is not eliminated. A compromise path, denoted by abode, shows alternations of recession and recovery as a country’s attention shifts back and forth between inflation and unemployment situations.

Cost-push theories suggest that inflation in a pressure-group economy at high employment is difficult to control without direct regulation of prices and/or wages. The London Economist has spoken of “The Uneasy Triangle” (1952) of full employment, strong labor organizations, and stable prices, of which no more than two are obtainable simultaneously. Comparing the rates of change of money wage rates with the rates of unemployment in the prior year, A. W. Phillips (1958) has led in the development of “Phillips curves,” of which Figure 4 is an idealized version. If the vertical line A is a maximum politically acceptable rate of unemployment (possibly 4 per cent) and the horizontal line B is the maximum economically possible rate of noninflationary wage increase (in the absence of drastic redistribution of personal incomes), the Phillips curves for most countries (including the

United States but apparently not Great Britain) pass outside their intersection at P. This, like the Economist’s uneasy triangle, illustrates the dilemma of the anti-inflation policy maker in a parliamentary democracy and in a basically free economy.

Administrative and sectoral inflation. The anti-union animus of many, if not most, of the views summarized in the last section has not gone unanswered. Gardiner C. Means (1962) believes he has evidence of “administered inflation,” as industry inches a bit at a time toward a maximum profit position by raising prices in prosperity and maintaining them in recession. Labor, in this view, has merely protected its real wage and traditional income share.

A neutral variant, “blaming” neither labor nor capital, is the sectoral-inflation theory of Charles Schultze (1959). Imagine an economy producing two goods, a and b, whose prices are both rigid downward. Suppose that demand shifts from sector b to sector a. The price of a rises in response to the demand shift. The price of b remains constant despite the demand shift. An index number combining the prices of a and b must rise. If higher profits in a induce wage demands there that are won in collective bargaining, similar increases may also be won in b, forcing prices to rise there as well.

Structural theories

Structural theories of inflation have resulted from doubts about the efficacy of policies based on either demand-inflation or cost-inflation views. Of the three examples presented, one stresses factor disproportionality, one income distribution, and one the economic growth rate. There are many other forms, since “structuralism”

is seldom defined and covers a wide variety of economic phenomena, particularly in Latin America but also in other underdeveloped areas.

Resource disproportionality. Within the structuralist camp (in Spanish, estructuralismo or cepa-lismo) one variant is the disproportionality theory. In our simplified picture of this theory (Figure 5), there are only two productive resources, homogeneous capital goods and homogeneous labor. A number of industries use capital and labor, each in fixed proportions. In the bundle of rays through the origin in Figure 5, each ray depicts a different industry. There is a labor force of L workers and a capital stock of C machines. Of the L workers, at least NL are in open or disguised unemployment, since their marginal productivity is. zero. A capital stock of K is required to employ all L workers. In this situation, inflation can impose forced abstinence upon the community by concentrating credit in capital goods industries. These can drive the capital stock toward K from C, raise warranted employment above N, and increase the national income above Y.

Economic growth. A principal argument of structuralists, exemplified by Raúl Prebisch of the Economic Commission for Latin America, is that monetary or fiscal controls on inflation in developing economies check precisely those structural changes that bring about sustainable growth. Many structuralists dislike static diagrammatic and mathematical treatment of their dynamic problems; it is difficult to formalize their views.

Consider, however, Figure 6, and turn to the right-hand panel. Structuralists believe that increases in the nominal amount of money increase both aggregate supply and aggregate demand functions when these relate to domestic output currently produced. A rise in money and prices is a partial expropriation of rentiers, who prefer capital export and land speculation to domestic production as investment outlets. It subsidizes the active elements who both produce and purchase current domestic output, since these are the initial recipients of easy money and credit. On the diagram the (vertical) shift in the aggregate supply function from S” through S1, to S2 reflects higher money wages and other costs. The (horizontal) shift in the aggregate demand function from D0 through Da to D2 reflects increased money incomes. A curve labeled Q passes through the intersections (Q”, Qi, Q”). Its positive slope in the range from Q” to R should not be confused with the slopes of true supply functions. This slope illustrates the structuralist opinion that aggregate output and employment vary in the same direction as the price level (at least up to

a point). If this is the case, premature application of indirect controls slows down the growth rates of output and employment.

According to the left-hand panel, an increase in the nominal money supply raises prices from P0 to P1, but the real output rises from Y0 to Y1. A further increase in the money supply raises real output further, to Y2, but it raises prices by a larger proportionate amount, to P2 The policy implication is that the first increase is desirable; the second increase, if it is in the range beyond R, may not be.

Income distribution. The issues of income and wealth distribution are brought into the foreground of the inflation discussion by another structuralist group, the “sociological school” of economists, centered in France. These writers view inflation in terms of conflicting money-income demands by rival social classes. The fixed-income receivers (salariat, rentiers, etc.) have a claim to F units of money income (see Figure 7, based upon Reder 1948). The variable-income receivers (wage earners and entrepreneurs) claim varying amounts of money income (V) as the price level rises. The total of income claims (demands), related to the price level, is plotted as F + V in Figure 7. Let the (given) total of real income be a ray through the origin, 0 —the higher the real income, the steeper the slope of the ray. As prices rise, the total of income claims rises more slowly, up to a point; the fixed-income element provides a stabilizer. In fact (P1,Y1) in Figure 7 is a stable-equilibrium price and income level. If prices rise further, however, the total of income claims eventually rises faster than the price level. Some fixed claims are “escalated,” and others include in their demands protection against the inflation they foresee in the short run. On the diagram, the (F + V) function becomes concave upward and the second equilibrium level (P2,Y2) becomes unstable. A price level 10 per cent above P2, for example, generates income claims 15 per cent above Y2; if granted, these spiral, in turn, to further price increases.

The implication of this picture (Figure 7) is disturbing. Once P1 is well behind, stable equilibrium at a stable price level is hard to restore, even though real income increases, without refusing the

income claims of one or more variable-income classes—workers or entrepreneurs. This, in turn, is difficult to do without some form of major social disturbance.

Empirical aspects

Statistical testing

While the foregoing theories of inflation are not mutually exclusive, there is room for substantial disagreement as to their relative importance. The theoretical dispute has called forth a substantial volume of statistical research designed to illuminate this issue of relative importance, but the results have not been accepted unequivocally.

Much of the statistical research since World War II has been carried on by advocates of the sophisticated “modern” version of the quantity theory of money, in attempted refutation of one or another of the rival views presented above. The nature of a few selected studies may be indicated briefly:

(1)Studies of historical changes in the quantity of money, the velocity of circulation, the price level, and the level of real income have tended to show that monetary changes come first—the others come later (with relatively long and variable lags)—and that income can be forecast as well from monetary changes as from changes in the income theory’s “autonomous” variables (investment, government spending, and the constant term in the consumption function). In addition to Friedman himself (see Friedman & Meiselman 1963), outstanding work along this line has been done by Clark War-burton (1951).

(2)Studies of hyperinflation tend to show that, in addition to the quantity of money, an index of expected future price changes (extrapolated from the recent past) is all that is needed for satisfactory explanation and prediction. Here the work of Cagan (1956) has blazed the trail, at least on the statistical side.

(3)Correlation studies at the microeconomic level, between individual price and quantity changes, have cast some doubt on cost theories of inflation, since the observed correlations have been large and positive. Edmund Phelps (1961) and Richard Selden (1959) have worked along these lines.

(4)Studies of inflation and economic growth attempt to locate and estimate within broad limits an optimum rate of inflation (from the viewpoint of growth). Results here are much in doubt because of the impinging of so many additional variables, not to mention the suppression of inflation.

(5)The number and variety of Phillips-curve studies, involving a number of countries, periods, and additional variables has already been mentioned. It remains open to question, however, whether the Phillips curve is independent of monetary and fiscal policy. That is to say, a higher degree of unemployment may well be required, to avoid inflationary wage increases, when the monetary authorities have a long record of validating such increases by expansionary policies than when they have not.

(6)An interesting recent variant of ordinary functional relations between M (the money supply) and P (the price level) is provided by A. C. Harberger’s study of Chilean monetary history (1963). Given a high statistical correlation between Pt(dependent variable) and the lagged money supply (Mt-1, Mt-2), Harberger inquiries: By how much could this correlation be increased by including such additional dependent variables as an index of wage rates, representing cost-push effects, and the force of inflationary acceleration (Pt-1 –Pt-2), representing an inflationary or deflationary economic structure? Harberger’s initial results for Chile suggest that the additional correlation is very small.

(7)For the United States, the National Bureau of Economic Research (1961) task force headed by George Stigler has supported the thesis that published price indexes exaggerate the degree of inflation and understate the degree of deflation, by reason of a number of defects, the most significant being the omission of new products and quality changes and the use of quoted, rather than actual, prices. The last-named point also bears upon the validity of Means’s thesis (1962) of inflationary price administration.

Speed, duration, and extent of inflation

Let us agree that an inflationary process occurs when active claims for shares in real national income or expenditure exceed the amount that can be produced in time to satisfy them. Inflation then continues until the conflicting claims are somehow reconciled or modified. The speed of the inflation that follows an initial inflationary shock depends upon the reactions of various income and spending groups. If there are few reactions and these are delayed and/or mild, the pace of inflation will be slow and will not go far beyond the initial shock. However, if the successive reactions are numerous, rapid, and vigorous, the resulting inflation may be explosive (hyperinflation). The properties of the inflation, given an inflationary shock of specific size, will then depend on a number of factors:

Effects on expenditure demands. Rising prices do not, in the aggregate, necessarily imply increases in the quantity of goods and services supplied or reductions in the quantity demanded, as in the case of individual commodities. At least, they do not do so if the nominal quantity of money and the money wage rate are both free to vary. This is because rising prices and wages are incomes to some groups, as well as costs to others. While it is therefore possible for an inflation to continue forever, this seldom happens, because the inflationary process reduces the level of real expenditures in one or more of the following ways: (a) “money illusion” on the part of spenders who plan their purchases in money rather than real terms; (b) a change in the distribution of income away from rentiers, pensioners, aged persons, and others with high marginal-spending propensities; (c) a rise in the marginal tax rate of progressive income and profits taxes; (d) real balance effects, as represented in Figure 3, if prices rise more rapidly than the quantity of money; (e) rising real interest rates, sometimes reinforced by credit rationing; (f) balance of payments difficulties if the country is inflating more rapidly than its trading partners.

Effects on income claims. The process of inflation will affect the ability of different economic groups to push for higher incomes. In a demand inflation, for example, there may be an initial fall in the level of unemployment and an initial rise in the rate of profit. Inflation then accelerates as labor achieves large wage increases that lead, in turn, to higher prices. In a cost inflation unemployment tends to increase. This reduces subsequent cost pushes until the inflation comes to an eventual halt. This process is thwarted, of course, to whatever extent the government’s full-employment policy “validates” wage and price increases that would otherwise raise the level of unemployment.

Lags. Lags influence strongly the character of inflationary movements. Without them, an inflation would run its course almost instantaneously. Their existence is largely responsible for the manageable rates of inflation in most Western nations since 1950. Perhaps the most important lag is the wage lag, which results from the fact that collective bargaining is seldom initiated more frequently than once a year and that negotiations may drag on for weeks or months. The process is also slowed by the fact that some months usually elapse before price changes are reflected in published index numbers. The wage lag is usually even greater for the so-called passive, or fixed income, groups. The greater the size of these elements, with their frequently anti-inflationary political views, the slower and smaller the rise in prices.

In industries that administer prices on the basis of cost changes, the lag between cost and price changes is probably less than the lag of wage rates behind changes in the cost of living. (On the other hand, such prices may be insensitive to demand changes in a demand inflation.) In the public sector, there are lags in both receipts and expenditures. Increases in tax receipts, induced by inflation, are not received immediately in the Treasury. Government expenditures often lag behind appropriations. In both legislative and administrative processes of inflation control, furthermore, there are “inside” lags delaying the application of whatever controls are decided upon and “outside” lags delaying their effectiveness. These public-sector lags are probably the principal exceptions to our generalization that long lags mean slow inflations and vice versa.

Escalation. Escalated increases in income are no different in principle from wage increases or farm prices negotiated directly. When escalated wage increases, for example, are no greater than warranted by the excess demand for labor, they can be considered on a par with wage increases on the open market. When wages are forced up by escalation, despite deficient demand for labor, the results are analogous to a cost push. In general, when escalated increases substitute for increases that would otherwise have been attained through other means, the institutional arrangements of escalation cannot be considered to have contributed to the inflationary process.

The impact of escalation will depend, in addition, upon the rate and frequency of adjustment as well as the percentage of population covered. If the total income-earning population were covered, with incomes escalated both immediately and proportionately to changes in a cost-of-living index, a crucial stabilizing lag would be eliminated from the system. Typically, these conditions are not met in full and the lag is preserved in part.

Statistical studies summarized by Franklyn D. Holzman (1964) suggest, surprisingly enough, that formal wage escalation has had no discernible influence on the speed of inflation in the United States, Great Britain, or Denmark. The results suggest that escalation in these countries has served primarily as a substitute for, rather than an addition to, other income increases and that its extent has been determined largely by market forces.

Consumer expectations. It is well known that expectations play a major role in accelerating in flation, once price increases are already rapid. Hyperinflations, in particular, are characterized by a so-called flight from cash, which Cagan explains statistically as a consequence of an index of expected price changes.

In less rapid inflations, however, the public seems willing to pay a high price for the convenience of holding its customary stock of cash balances. In the first place, people differ in their perceptions of past price trends, so that in creeping inflations not everyone realizes what has gone on. In the second place, recognition that prices have risen need not lead to anticipations of future increases. In the third place, even expectations of future increases do not necessarily lead to an acceleration of purchases. In the United States, high and rising prices in both 1951 and 1957 led to a drop, rather than a rise, in expenditures for a wide range of consumer durable items. Interview data suggest that this occurred because consumers resented the high level of prices and went on a buyers’strike. This attitude was more widespread than any incentive to “buy early and beat the price rise.” And finally, increases in expenditures based on expected price increases are more likely to cause “once-and-for-all” increases in expenditures than are the continuous increases that accelerate inflation. As long as people are willing to hold cash balances at a sacrifice of real income, continuous inflationary shifts in demand from money to goods require an increasing certainty about the upward price trend or an increase in the supply of cash or credit, or (more usually) both, as in the hyperinflationary cases.

Redistributive effects of inflation

Inflation alters the distributions of both income and wealth, partly in a random and partly in a systematic manner. Prices of different goods and services, different productive resources, and different monetary assets, real assets, and liabilities are not equally flexible and respond differently to inflationary pressure. Furthermore, when inflation and deflation change the real value of cash balances, supply and demand functions for other goods and services change in different degrees, depending on their complementarities with cash balance holdings. Purely monetary assets and liabilities (cash, insurance policies, bonds) are completely inflexible unless escalated. Their nominal market values are independent of price level changes. Most monetary rent and interest incomes in America and western Europe are fixed contractually for long periods and so react slowly to price increases. In underdeveloped countries, on the other hand, rent and interest obligations in kind are favorite inflation hedges. The values of physical assets held for use, rather than sale, adjust more or less proportionately to the price level. Most other prices move with the price level but at different rates. The prices of durables tend to rise faster than the price of food; the wages of government workers and professionals usually lag behind those of industrial workers; and common stock and real estate prices usually overadjust to changes in commodity price levels. (For this reason, these assets are often used as inflation hedges, along with precious metals, jewelry, art objects, and stable foreign currencies, whose prices have also shown overadjusting tendencies.)

Assessment of the redistributive impact of inflation requires consideration of the relative impact on individuals and sectors of all these elements of inflation sensitivity. It is also difficult to separate the effects of general price movements on relative prices from the effects of other changes, including general economic growth.

In general, the differential impact of inflation on individuals and groups is a function of two factors: their ability to forecast the course of inflation and their ability to adjust their economic behavior to take account of what they foresee. If everyone had equal abilities along these lines, inflation and deflation would have little or no redistributive effects. There are, in fact, great differences on both accounts. These differences are responsible for the “inequities” of inflation. It should also be remembered, on the other hand, that the larger the number of people who anticipate price changes correctly and adjust to them promptly, the more rapid the changes are likely to be.

Two related hypotheses on the redistributive effects of inflation on income have been accepted for many years but have only recently been analyzed with care. The first of these hypotheses is that in inflation money wages lag behind prices, so that there is a shift away from wages and in favor of profits. The second hypothesis states that business firms gain through inflation at the expense of households and public bodies, and vice versa for deflation.

The wage-lag hypothesis has been argued in a number of historical studies. Earl J. Hamilton (1952) used it as an explanation of the financing of western European industrialization following the discovery of Mexican and Peruvian gold and silver deposits. Similar studies cover the periods of the American Civil War and of the inflations that followed World War I. Their conclusions have, however, been challenged, particularly by Armen Al-chian and Reuben Kessel (1959). These and other critics ascribe the orthodox results to an arbitrary choice of beginning and ending years and to other factors, such as population growth and migration to the cities, that would have presumably caused wage lags regardless of the course of general prices. Alchian and Kessel conclude their study by considering the United States after World War II, when labor organization had achieved substantial strength. They argue that, if the wage-lag hypothesis is valid, firms with large annual wage bills (relative to equity) would show proportionately higher profit increases. The opposite seems to have been in fact the case.

Several American studies have examined the changes in the functional distribution of income since 1945, in attempts to test the wage-lag hypothesis more aggregatively. With regard to wages and profits, the picture is mixed. Some periods, such as 1946 to 1948, apparently conform to the wage-lag hypothesis whereas others, such as 1949 to 1953 and 1955 to 1958, do not. Phelps (1961) has suggested that a demand inflation conforms to the hypothesis more closely than does a cost inflation. The studies of labor share, moreover, include both the effects of wage-rate changes and the effects of changes in factor proportions—as capital is (or is not) substituted for labor, as technical improvements occur, and as wage rates rise. It is, in any case, agreed that “active elements” (wages and profits) gain relatively to “passive elements” (interest and rent) whenever prices rise, and vice versa, so that the wage-lag controversy reduces to the question of which active elements gain most from the partial expropriation of the passive ones. Harold M. Levinson (1960) approached this by dividing the private economy of the United States into two sectors, one in which trade unions are strong (manufacturing, mining, transportation, and utilities) and one in which they are weak (commerce, finance, trade, services, and agriculture). His results indicate that the wage-lag hypothesis is confirmed in the “nonunion” group but not confirmed in the “union” one.

The other conclusion—that business, or the corporate sector, gains through inflation and loses through deflation—involves two assumptions. If it is true that wages lag behind prices, it would follow that corporate profits tend to rise during inflations. A more common assumption, associated with both Fisher and Keynes, asserts that business is a net debtor and gains during inflation by the opportunity to pay debts in cheaper money. A corollary is theproposition that banks, being particularly large debtors (with an unusually large ratio of debt to equity), should gain disproportionately in inflation. (This does not seem to be the case, perhaps because bank investments are concentrated in fixed money claims.)

The several studies by Alchian and Kessel confirm the Keynes-Fisher assertion of the importance of the debtor-creditor status in inflation. They take issue, however, with the assertion that business firms are net debtors. This assumption was true before World War i in the United States, but the Alchian-Kessel samples (1959) for the period after World War II found business firms distributed evenly between net debtors and net creditors. The stock prices of debtor companies rose by larger percentages than those of creditor companies, but even this aspect of conventional wisdom was questioned in another study, by G. L. Bach and Albert Ando (1957), covering the period 1939 to 1954. Bach and Ando found the debtor-creditor position often overbalanced by the effects of a number of other factors, most importantly the volume of sales.

Turning from income to wealth, Bach and Ando studied the redistributive effect of inflation in terms largely of the net monetary (fixed price) asset or liability positions of different population groups. The aggregative figures show that in the United States over the decade 1939 to 1949 the household sector became a large net creditor and the government sector a large net debtor. Unincorporated business, nonfinancial corporations, and financial corporations were all nearly neutral. More revealing was the percentage of net assets held in monetary form by income groups classified by income size. Except for the lowest income group, this percentage was approximately constant at from 13 per cent to 15 per cent of total assets. Calculations suggest that, as compared with personal income taxes particularly, losses on net monetary account from inflation are regressive.

The real locus of the burden of holding monetary assets in inflation is revealed by making similar calculations for different occupational and age groups. By far the highest concentration of monetary assets is found among the retired. High proportions are also indicated for professional and semiprofessional workers and for unskilled workers. Farm operators and managers, on the other hand, suffer only minimally. The high figure for retired people is emphasized by the age-group breakdown, which records a higher proportion of monetary assets in the oldest age group (55 years old and over) than in any other. Finally, renters tend to lose relatively to homeowners when prices are rising. Since most members of the lower income groups cannot afford their own homes, this aspect of the cost of inflation weighs most heavily upon them.

M. bronfenbrenner

[See alsobusiness cycles].


Ackley, Gardner 1961 Macroeconomic Theory. New York: Macmillan. → A useful textbook presentation of the problems of inflation and deflation. See especially Chapter 16, “The Theory of Inflation.”

Alchian, Armen A.; and Kessel, Reuben A. 1959 Redistribution of Wealth Through Inflation. Science 130:535-539. → A summary of the authors’several studies.

Bach, G. L.; and Ando, Albert 1957 The Redistributional Effects of Inflation. Review of Economics and Statistics 39:1-13.

Ball, R. J. 1964 Inflation and the Theory of Money. Chicago: Aldine.

Biacabe, Pierre 1962 Analyses contemporaines de Vin-flation. Paris: Sirey. → A sympathetic presentation of structural and allied theories.

Rresciani-Turroni, Constantino (1931) 1937 The Economics of Inflation: A Study of Currency Depreciation in Post-war Germany. London: Allen & Unwin. → First published as Le vicende del marco tedesco.

Bronfenbrenner, Martin; and Holzman, Franklyn D. 1963 Survey of Inflation Theory. American Economic Review 53:593-661. → A parallel discussion of many of the points in this article.

Brown, Arthur J. 1955 The Great Inflation: 1939-1951. Oxford Univ. Press. → Contains theoretical as well as historical material.

Cagan, Phillip 1956 The Monetary Dynamics of Hyperinflation. Pages 25-117 in Milton Friedman (editor), Studies in the Quantity Theory of Money. Univ. of Chicago Press.

Candler, Lester V. 1951 Inflation in the United States: 1940-1948. New York: Harper. → Contains theoretical as well as historical material.

Conference on Inflation, Helsing0r, Denmark,19591962 Inflation: Proceedings of a Conference Held by the International Economic Association. Edited by Douglas Hague. New York: St. Martins. → Valuable as a confrontation of rival theories.

De Podwin, Horace J.; and Selden, Richard T. 1963 Business Pricing Policies and Inflation. Journal of Political Economy 71:116-127.

Dernburg, Thomas F.; and McDougall, Duncan M. (1960) 1963 Macro-economics: The Measurement, Analysis, and Control of Aggregate Economic Activity. 2d ed. New York: McGraw-Hill. → A good textbook presentation of the topic. See especially pages 203-218, “Fundamentals of Growth Economics.”

Friedman, Milton (editor) 1956 Studies in the Quantity Theory of Money. Univ. of Chicago Press. → See especially pages 5-21 on “The Quantity Theory of Money: A Restatement,” by Milton Friedman.

Friedman, Milton; and Meiselman, David 1964 The Relative Stability of Monetary Velocity and the Investment Multiplier in the United States, 1897-1958. Pages 165-268 in Stabilization Policies: A Series of Research Studies Prepared for the Commission on Money and Credit. Englewood Cliffs, N.J.: Prentice-Hall.

Hamilton, Earl J. 1952 Prices as a Factor in Business Growth: Prices and Progress. Journal of Economic History 12, no. 4:325-349.

Hansen, Bent 1951 A Study in the Theory of Inflation. London: Allen & Unwin.

Harberger, Arnold C. 1963 The Dynamics of Inflation in Chile. Pages 219-250 in Measurement in Economics: Studies in Mathematical Economics and Econometrics in Memory of Yehuda Grunfeld, by Carl F. Christ et al. Stanford Univ. Press.

Holzman, Franklyn D. 1950 Income Determination in Open Inflation. Review of Economics and Statistics 32:150-158.

Holzman, Franklyn D. 1964 Escalation and Its Use to Mitigate the Inequities of Inflation. Pages 177-229 in Inflation, Growth and Employment: A Series of Research Studies Prepared for the Commission on Money and Credit. Englewood Cliffs, N.J.: Prentice-Hall.

Keynes, John M. 1940 How to Pay for the War: A Radical Plan for the Chancellor of the Exchequer. New York: Harcourt. → The standard presentation of income theory in relation to price level changes.

Koopmans, Tjalling 1942 The Dynamics of Inflation. Review of Economics and Statistics 24:53-65. → A comment by A. Smithies and a reply by Koopmans appear on pages 189-190.

Lerner, Abba P. 1951 Economics of Employment. New York: McGraw-Hill. → See especially pages 191-241 on “Full Employment and Inflation.”

Levinson, Harold M. 1954 Collective Bargaining and Income Distribution. American Economic Review 44, no. 2:308-316.

Levinson, Harold M. 1960 Postwar Movement of Prices and Wages in Manufacturing Industries. Supplementary Technical Material to the Staff Report, by George W. Bleile and Thomas A. Wilson. U.S. Congress, Joint Economic Committee, Study Paper No. 21. Washington : Government Printing Office. → Material prepared in connection with the study of employment, growth, and price levels.

Means, Gardiner C. 1962 Pricing Power and the Public Interest. New York: Harper.

National Bureau of Economic Research, Price Statistics Review Committee 1961 The Price Statistics of the Federal Government: Review, Appraisal, and Recommendations. Edited by George J. Stigler. New York: The Bureau.

Phelps, Edmund S. 1961 A Test for the Presence of Cost Inflation in the United States: 1955-1957. Yale Economic Essays 1:28-69.

Phillips, A. W. 1958 The Relation Between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom: 1861-1957. Economica New Series 25:283-299. → Statistical test of compatibility between full employment and price level stability.

Reder, Melvin W. 1948 The Theoretical Problems of a National Wage-Price Policy. Canadian Journal of Economics and Political Science 14:46-61.

Rothbard, Murray N. 1962 The Case for a 100 Per Cent Gold Dollar. Pages 94-136 in Leland B. Yeager (editor), In Search of a Monetary Constitution. Cambridge, Mass.: Harvard Univ. Press.

Schultze, Charles L. 1959 Recent Inflation in the United States. U.S. Congress, Joint Economic Committee, Study Paper No. 1. Washington: Government Printing Office. → Discusses cost theories.

Seers, Dudley 1962 A Theory of Inflation and Growth in Under-developed Economies Based on the Experience of Latin America. Oxford Economic Papers New Series 14, no. 2:173-195. → A sympathetic presentation of structural and allied theories.

Selden, Richard T. 1959 Cost-push Versus Demand-pull Inflation: 1955-1957. Journal of Political Economy 67:1-20. → A good empirical study.

Smithies, Arthur 1943 The Behavior of Money National Income Under Inflationary Conditions. Quarterly Journal of Economics 57:113-128.

Thorp, Willard L.; and Quandt, Richard E. 1959 The New Inflation. New York: McGraw-Hill. → Discusses cost theories.

Turvey, Ralph 1949 Period Analysis and Inflation. Economica New Series 16:218-227.

The Uneasy Triangle. 1952 Economist 164:322-323, 376-378, 434-435.

Warrurton, Clark 1951 The Misplaced Emphasis in Contemporary Business-fluctuation theory. Pages 284-318 in American Economic Association, Readings in Monetary Theory. Homewood, III.: Irwin; Philadelphia: Blakiston.

Weintraub, Sidney 1959 Forecasting the Price Level, Income Distribution, and Economic Growth. Philadelphia: Chilton. → A college edition was published simultaneously as A General Theory of Price Level, Output, Income Distribution, and Economic Growth.