Published on

1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. THE EVOLUTION AND POLICY IMPLICATIONS OF PHILLIPS CURVE ANALYSIS Thomas M. Humphrey At the core of modern macroeconomics is some steps that led to this change. Accordingly, the para- version or another of the famous Phillips curve rela- graphs below sketch the evolution of Phillips curve tionship between inflation and unemployment. The analysis, emphasizing in particular the theoretical Phillips curve, both in itspriginal and more recently innovations incorporated into that analysis at each reformulated expectations-augmented versions, has stage and the policy implications of each innovation. two main uses. In theoretical models of inflation, it provides the so-called “missing equation” that ex- plains how changes in nominal income divide them- I. selves into price and quantity components. On the EARLY VERSIONS OF THE PHILLIPS CURVE policy front, it specifies conditions contributing to the effectiveness (or lack thereof) of expansionary The idea of an inflation-unemployment trade-off is and disinflationary policies. For example, in its hardly new. It was a key component of the monetary expectations-augmented form, it predicts that the doctrines of David Hume (1752) and Henry Thorn- power of expansionary measures to stimulate real ton (1802). It was identified statistically by Irving activity depends critically upon how price anticipa- Fisher in 1926, although he viewed causation as tions are formed. Similarly, it predicts that disinfla- running from inflation to unemployment rather than tionary policy will either work slowly (and painfully) vice versa. It was stated in the form of an econo- or swiftly (and painlessly) depending upon the speed metric equation by Jan Tinbergen in 1936 and again of adjustment of price expectations. In fact, few by Lawrence Klein and Arthur Goldberger in 1955. macro policy questions are discussed without at least Finally, it was graphed on a scatterplot chart by A. J. some reference to an analytical framework that might Brown in 1955 and presented in the form of a dia- be described in terms of some version of the Phillips grammatic curve by Paul Sultan in 1957. Despite curve. these early efforts, however, it was not until 1958 As might be expected from such a widely used tool, that modern Phillips curve analysis can be said to,Phillips curve analysis has hardly stood still since its have begun. That year saw the publication of Pro- beginnings in 1958. Rather it has evolved under the fessor A. W. Phillips’ famous article in which he pressure of events and the progress of economic fitted a statistical equation w=f (U) to annual data theorizing, incorporating at each stage such new on percentage rates of change of money wages (w) elements as the natural rate hypothesis, the adaptive- and the unemployment rate (U) in the United King- expectations mechanism, and most recently, the ra- dom for the period 1861-1913. The result, shown tional expectations hypothesis. Each new element in a chart like Figure 1 with wage inflation measured expanded its expIanatory power. Each radically verticaIIy and unempIoyment horizontally, was a altered its policy implications. As a result, whereas smooth, downward-sloping convex curve that cut the the Phillips curve was once seen as offering a stable horizontal axis at a positive level of unemployment. enduring trade-off for the policymakers to exploit, The curve itself was given a straightforward inter- it is now widely viewed as offering no trade-off at all. pretation : it showed the response of wages to the In short, the original Phillips curve notion of the excess demand for labor as proxied by the inverse of potency of activist fine tuning has given way to the the unemployment rate. Low unemployment spelled revised Phillips curve notion of policy ineffectiveness. high excess demand and thus upward pressure on The purpose of this articIe is to trace the sequence of wages. The greater this excess Iabor demand the FEDERAL RESERVE BANK OF RICHMOND 3
  2. 2. exist even when the market was in equilibrium, that is, when excess labor demand was zero and wages were stable. Accordingly, this frictional unemploy- Figure 1 ment was indicated by the point at which the Phillips EARLY PHILLIPS CURVE curve crosses the horizontal axis. According to Phillips, this is also the point to which the economy w Wage Inflation Rate 1%) returns if the authorities ceased to maintain dis- equilibrium in the labor market by pegging the excess Phillips Curve Trade-off demand for labor. Finally, since increases in excess demand would likely run into diminishing marginal Inflation and Unemployment returns in reducing unemployment, it followed that the curve must be convex-this convexity showing that successive uniform decrements in unemployment would require progressively larger increments in . Frictional (and Structural) excess demand (and thus wage inflation rates) to. Unemployment Rate at achieve them. Which Overall Excess Demand for Labor is Zero and Wages are Popularity of the Phillips Paradigm therefore Stable Once equipped with the foregoing theoretical foun- dations, the Phillips curve gained swift acceptance Unemployment among economists and policymakers alike. It is important to understand why this was so. At least three factors probably contributed to the attractive- ness of the Phillips curve. One was the remarkable temporal stability of the relationship, a stability re- At unemployment rate Uf the labor market is in equilibrium and wages are stable. At vealed by Phillips’ own finding that the same curve lower unemployment rates excess demand estimated for the pre-World War I period 1861-1913 exists to bid up wages. At higher unemploy- fitted the United Kingdom data for the post-World ment rates excess supply exists to bid down War II period 1948-1957 equally well or even better. wages. The curve’s convex shape shows that increasing excess demand for labor runs into Such apparent stability in a two-variable relationship diminishing marginal returns in reducing un- over such a long period of time is uncommon in employment. Thus successive uniform de empirical economics and served to excite interest in creases in unemployment (horizontal gray the curve. arrows) require progressively larger increases in excess demand and hence wage inflation A second factor contributing to the success of the rates (vertical black arrows) as we go from Phillips curve was its ability to accommodate a wide point a to b to c to d along the curve. variety of inflation theories. The Phillips curve itself explained inflation as resulting from excess demand that bids up wages and prices. It was en- tirely neutral, however, about the causes of that phenomenon. Now excess demand can of course be generated either by shifts in demand or shifts infaster the rise in wages. Similarly, high unemploy- supply regardless of the causes of those shifts.ment spelled negative excess demand (i.e., excess Thus a demand-pull theorist could argue that excess-labor supply) that put deflationary pressure on demand-induced inflation stems from excessivelywages. Since the rate of change of wages varied expansionary aggregate demand policies while a cost-directly with excess demand, which in turn varied push theorist could claim that it emanates from trade-inversely with unemployment, wage inflation would union monopoly power and real shocks operating onrise with decreasing unemployment and fall with labor supply. The Phillips curve could accommodateincreasing unemployment as indicated by the negative both views. Economists of rival schools could acceptslope of the curve. Moreover, owing to unavoidable the Phillips curve as offering insights into the naturefrictions in the operation of the labor market, it of the inflationary process even while disagreeing onfollowed that some frictional unemployment would the causes of and appropriate remedies for inflation.4 ECONOMIC REVIEW, MARCH/APRIL 1985
  3. 3. Finally, the Phillips curve appealed to policy- (1) P = =GJ>makers because it provided a convincing rationale for where p is the rate of price inflation, x(U) is overalltheir apparent failure to achieve full employment excess demand in labor and hence product markets-with price stability-twin goals that were thought to this excess demand being an inverse function of thebe mutually compatible before Phillips’ analysis. unemployment rate-and a is a price-reaction coeffi-When criticized for failing to achieve both goals cient expressing the response of inflation to excesssimultaneously, the authorities could point to the demand. From this equation the authorities couldPhillips curve as showing that such an outcome was determine how much unemployment would be asso-impossible and that the best one could hope for was ciated with !ny given target rate of inflation. Theyeither arbitrarily low unemployment or price stability could also use it to measure the effect of policies-but not both. Note also that the curve, by offering a undertaken to obtain a more favorable Phillips curve,menu of alternative inflation-unemployment combi- i.e., policies aimed at lowering the price-responsenations from which the authorities could choose, coefficient and the amount of unemployment associ-provided a ready-made justification for discretionary ated with any given level of excess demand.intervention and activist fine tuning. Policymakershad but to select the best (or least undesirable) Trade-Offs and Attainable Combinationscombination on the menu and then use their policyinstruments to achieve it. For this reason too the The foregoing equation specifies the position (orcurve must have appealed to some policy authorities, distance from origin) and slope of the Phillips curvenot to mention the economic advisors who supplied -two features stressed in policy discussions of thethe cost-benefit analysis underlying their choices. early 1960s. As seen by the policymakers of that era, the curve’s position fixes the inner boundary, orFrom Wage-Change Relation to frontier, of feasible (attainable) combinations ofPrice-Change Relation inflation and unemployment rates (see Figure 2). As noted above, the initial Phillips curve depicted a Determined by the structure of labor and productrelation between unemployment and wage inflation. markets, the position of the curve defines the set ofPolicymakers, however, usually specify inflation tar- all coordinates of inflation and unemployment ratesgets in terms of rates of change of prices rather than the authorities could achieve via implementation ofwages. Accordingly, to make the Phillips curve more monetary and fiscal policies. Using these macroeco-useful to policymakers, it was therefore necessary to nomic demand-management policies the authoritiestransform it from a wage-change relationship to a could put the economy anywhere on the curve. Theyprice-change relationship. This transformation was could not, however, operate to the left of it. Theachieved by assuming that prices are set by apply- Phillips curve was viewed as a constraint preventinging a constant mark-up to unit labor cost and so move them from achieving still lower levels of both inflationin step with wages--or, more precisely, move at a and unemployment. Given the structure of labor andrate equal to the differential between the percentage product markets, it would be impossible for mone-rates of growth of wages and productivity (the latter tary and fiscal policy alone to reach inflation-assumed zero here).’ The result of this transforma- unemployment combinations in the region to the lefttion was the price-change Phillips relation of the curve. The slope of the curve was interpreted as showing1 Let prices P be the product of a fixed markup K (in- the relevant policy trade-offs (rates of exchangecluding normal profit margin and provision for depreci- between policy goals) available to the authorities. Asation) applied to unit labor costs C, (1) P = K C . explained in early Phillips curve analysis, theseUnit labor costs by definition are the ratio of hourly trade-offs arise because of the existence of irrecon-wages W to labor productivity or output per labor hour cilable conflicts among policy objectives. When theQ? goals of full employment and price stability are not (2) C = W/Q.Substituting (2) into (I), taking logarithms of both sides simultaneously achievable, then attempts to move theof the resultinn exoression. and then differentiating with economy closer to one will necessarily move it furtherrespect to time yields away from the other. The rate at which one objective (3) P = w - qwhere the lower case letters denote the percentage rates must be given up to obtain a little bit more of theof change of the price, wage, and productivity variables. other is measured by the slope of the Phillips curve.Assuming productivity growth q is zero and the rate of For example, when the Phillips curve is steeplywage change w is an inverse function of the unemploy-ment rate yields equation (1) of the text. sloped, it means that a small reduction in unemploy- FEDERAL RESERVE BANK OF RICHMOND 5
  4. 4. rates of unemployment in exchange for permanently higher rates of inflation or vice versa. Put differ- ently, the curve was interpreted as offering a menu Figure 2 of alternative inflation-unemployment combinations TRADE-OFFS AND from which the authorities could choose. Given the menu, the authorities’ task was to select the particular ATTAINABLE COMBINATIONS inflation-unemployment mix resulting in the smallest social cost (see Figure 3). To do this, they would p Price I&ffation Rate (96) have to assign relative weights to the twin evils of Phillips Curve / Frontier p=axKl): Shows Attainable Inflation-Unemployment. Combinations Figure 3 THE BEST SELECTION ON Slope A--- Indicates THE MENU OF CHOICES Trade-off p Price Inflation Rate Phillips Curve Constraint The position or location of the Phillips curve defines the frontier or set of Optimum Feasible attainable inflation-unemployment combi- Inflation-Unemployment nations. Using monetary and fiscal policies, Combination the authorities can attain all combinations lying upon the frontier itself but none in the shaded region below it. In this way the curve acts as a constraint on demand- management policy choices. The slope of the curve shows the tradeoffs or rates of exchange between the two evils of inflation and unemployment. U The bowed-out curves are social disutility contours. Each contour shows all the com- binations of inflation and unemployment ment would be purchased at the cost of a large in- resulting in a given level of social cost or crease in the rate of inflation. Conversely, in rela- harm. The closer to the origin, the lower tively flat portions of the curve, considerably lower the social cost. The slopes of these contours unemployment could be obtained fairly cheaply, that reflect the relative weights that society (or the policy authority) assigns to the evils of is at the cost of only slight increases in inflation. inflation and unemployment. The best Knowledge of these trade-offs would enable the combination of inflation and unemploy- authorities to determine the price-stability sacrifice ment that the policymakers can reach, given necessary to buy any given reduction in the unem- the Phillips curve constraint, is the mix appearing on the lowest attainable social ployment rate. disutility contour. Here the additional social benefit from a unit reduction in The Best Selection on the Phillips Frontier unemployment will just be worth the extra inflation cost of doing so. The preceding has described the early view of the Phillips curve as a stable, enduring trade-off per- mitting the authorities to obtain permanently lower 6 ECONOMIC REVIEW, MARCH/APRIL 1985
  5. 5. inflation and unemployment in accordance with theirviews of the comparative harm caused by each. Then,using monetary and fiscal policy, they would movealong the Phillips curve, trading off unemployment Figure 4for inflation (or vice versa) until they reached the DIFFERENT PREFERENCES,point at which the additional benefit from a further DIFFERENT POLICY CHOICESreduction in unemployment was just worth the extrainflation cost of doing so. Here would be the opti- p Price Inflation Ratemum, or least undesirable, mix of inflation and unem- b I/ployment. At this point the economy would be on itslowest attainable social disutility contour (the bowed- Phillips Curve Constraintout curves radiating outward from the origin ofFigure 3) allowed by the Phillips curve constraint. Inflation-Unemployment Choice of anHere the unemployment-inflation combination chosen ’ Unemployment-Aversewould be the one that minimized social harm. It was Administrationof course understood that if this outcome involved apositive rate of inflation, continuous excess money Social Disutility Contours:growth would be required to maintain it. For without Unemployment Weightedsuch monetary stimulus, excess demand would dis-appear and the economy would return to the pointat which the Phillips curve crosses the horizontalaxis.Different Preferences, Different Outcomes It was also recognized that policymakers mightdiffer in their assessment of the comparative social 0cost of inflation vs. unemployment and thus assigndifferent policy weights to each. Policymakers who Different political administrations maybelieved that unemployment was more undesirable differ in their evaluations of the socialthan rising prices would assign a much higher relative harmfulness of inflation relative to that ofweight to the former than would policymakers who unemployment. Thus in their policy delib erations they will attach different relativejudged inflation to be the worse evil. Hence, those weights to the two evils of inflation and un-with a marked aversion to unemployment would employment. These weights will be reprefer a point higher up on the Phillips curve than fleeted in the slopes of the social disutilitywould those more anxious to avoid inflation, as shown contours (as those contours are interpreted by the policymakers). The relatively flatin Figure 4. Whereas one political administration contours reflect the views of those attachingmight opt for a high pressure economy on the higher relative weight to the evils of infla-grounds that the social benefits of low unemployment tion; the steep contours to those assigningexceeded the harm done by the inflation necessary to higher weight to unemployment. An unem-achieve it, another administration might deliberately ployment-averse administration will choose a point on the Phillips curve involving moreaim for a low pressuse economy because it believed inflation and less unemployment than thethat some economic slack was a relatively painless combination selected by an inflation-aversemeans of eradicating harmful inflation. Both groups administration.would of course prefer combinations to the southwestof the Phillips constraint, down closer to the figure’sorigin (the ideal point of zero inflation and zero un-employment). As pointed out before, however, thiswould be impossible given the structure of the econ- Pessimistic Phillips Curve and theomy, which determines the position or location of the “Cruel Dilemma”Phillips frontier. In short, the policymakers wouldbe constrained to combinations lying on this bound- In the early 196Os, there was much discussion ofary, unless they were prepared to alter the economy’s the so-called “cruel-dilemma” problem imposed by anstructure. unfavorable Phillips curve. The cruel dilemma refers FEDERAL RESERVE BANK OF RICHMOND 7
  6. 6. to certain pessimistic situations where none of theavailable combinations on the menu of policy choicesis acceptable to the majority of a country’s voters (see Figure 5). For example, suppose there is some Figure 5maximum rate of inflation, A, that voters are just PESSIMISTIC PHILLIPS CURVEwilling to tolerate without removing the party in AND THE “CRUEL DILEMMA”power. Likewise, suppose there is some maximumtolerable rate of unemployment, B. As shown in p Price Inflation RateFigure 5, these limits define the zone of acceptable or Ipolitically feasible combinations of inflation and Pessimistic or Unfavorableunemployment. A Phillips curve that occupies a Phillips Curve; Liesposition anywhere within this zone will satisfy soci- / Outside the Zone ofety’s demands for reasonable price stability and high Tolerable Outcomes employment. But if both limits are exceeded and thecurve lies outside the region of satisfactory outcomes,the system’s performance will fall short of what wasexpected of it, and the resulting discontent may I Phillips Curveseverely aggravate political and social tensions. If, as some analysts alleged, the Phillips curvetended to be located so far to the right in the chartthat no portion of it fell within the zone of acceptablecombinations, then the policymakers would indeed beconfronted with a painful dilemma. At best theycould hold only one of the variables, inflation orunemployment, down to acceptable levels. But they A = Maximum Tolerable Rate of Inflationcould not hold both simultaneously within the limits B = Maximum Tolerable Rate of Unemploymentof toleration. Faced with such a pessimistic Phillipscurve, policymakers armed only with traditional Given the unfavorable Phillips curve, policy-demand-management policies would find it impossible makers are confronted with a cruel achieve combinations of inflation and unemploy- They can achieve acceptable rates of inffa-ment acceptable to society. tion (point a) or unemployment (point b) but not both. The rationale for i n c o m e s (wage-price) and strucrural (labor marketlPolicies to Shift the Phillips Curve policies was to shift the Phillips curve down into the zone of tolerable outcomes. It was this concern and frustration over the seem-ing inability of monetary and fiscal policy to resolvethe unemployment-inflation dilemma that inducedsome economists in the early 1960s to urge the adop-tion of incomes (wage-price) and structural (labor-market) policies. Monetary and fiscal policies alone coefficient to be zero (as with wage-price freezes),were thought to be insufficient to resolve the cruel or by replacing it with an officially mandated rate ofdilemma since the most these policies could do was to price increase, or simply by persuading sellers toenable the economy to occupy alternative positions on moderate their wage and price demands, such policiesthe pessimistic Phillips curve. That is, monetary would lower the rate of inflation associated with anyand fiscal policies could move the economy along the given level of unemployment and thus twist down thegiven curve, but they could not move the curve itself Phillips curve. The idea was that wage-price controlsinto the zone of tolerable outcomes. What was would hold inflation down while excess demand wasneeded, it was argued, were new policies that would being used to boost employment.twist or shift the Phillips frontier toward the origin Should incomes policies prove unworkable or pro-of the diagram. hibitively expensive in terms of their resource- Of these measures, incomes policies would be misallocation and restriction-of-freedom costs, thendirected at the price-response coefficient linking infla- the authorities could rely solely on microeconomiction to excess demand. Either by decreeing this structural policies to improve the trade-off. By en-8 ECONOMIC REVIEW, MARCH/APRIL 19R5
  7. 7. inflationary expectations had become too prominent to ignore and many analysts were perceiving them as the dominant cause of observed shifts in the Phillips curve. Figure 6 Coinciding with this perception was the belated EFFECTS OF UNEMPLOYMENT recognition that the original Phillips curve involved a DISPERSION misspecification that could only be corrected by the incorporation of a price expectations variable in the w Wage Inflation Rate trade-off. The original Phillips curve has expressed in terms of nominal wage changes, w=f (U) . Since neoclassical economic theory teaches that real rather I al Phillips Curve: H LoCThe Same for than nominal wages adjust to clear labor markets, Micromarkets A and B however, it follow,s that the Phillips curve should R WA have been stated in terms of real wage changes. I Better still (since wage bargains are made with an Line Showing Weighted / Average eye to the future), it should have been stated in terms i of Local Wage 1 Inflation Rates WA and WB of expected real wage changes, i.e., the differential between the rates of change of nominal wages and i Macro Wage expected future prices, w-pe=f(U). In short, the Inflation Rate: original Phillips curve required a price expectations Dispersion Case term to render it correct. Recognition of this fact No-Dispersion Case led to the development of the expectations-augmented Phillips curve described below. III. U THE EXPECTATIONS-AUGMENTED PHILLIPS CURVE AND THE ADAPTIVE-EXPECTATIONS MECHANISM The original Phillips curve equation gave way to If aggregate unemployment at rate U” were the expectations-augmented version in the early evenly distributed across individual labor markets such that the same rate prevailed 1970s. Three innovations ushered in this change. everywhere, then wages would inflate at the The first was the respecification of the excess de- rate w* both locally and nationally. But if mand variable. Originally defined as an inverse aggregate unemployment U” is unequally function of the unemployment rate, x(U), excess distributed such that rate UA exists in market A and UB in market B, then wages demand was redefined as the discrepancy or gap will inflate at rate WA in the former market between the natural and actual rates of unemploy- and wB in the latter. The average of these ment, UN-U. The natural (or full employment) local inflation rates at aggregate unemploy- rate of unemployment itself was defined as the rate ment rate U” is wo which is higher than inflation rate w* of the no-dispersion case. that prevails in steady-state equilibrium when expec- tations are fully realized and incorporated into all Conclusion: The greater the dispersion of wages and prices and inflation is neither accelerating unemployment, the higher the aggregate nor decelerating. It is natural in the sense (1) that inflation rate associated with any given level of aggregate unemployment. Unem- it represents normal full-employment equilibrium in ployment dispersion shifts the aggregate the labor and hence commodity markets, (2) that it Phillips curve rightward. is independent of the steady-state inflation rate, and (3) that it is determined by real structural forces (market frictions and imperfections, job information and labor mobility costs, tax laws, unemployment subsidies, and the like) and as such is not susceptible to manipulation by aggregate demand policies.10 ECONOMIC REVIEW, MARCH/APRIL 1955
  8. 8. The second innovation was the introduction of an amount equal to half the error, or 3 percentageprice anticipations into Phillips curve analysis re- points. Such revision will continue until the expec-sulting in the expectations-augmented equation tational error is eliminated. Analysts also demonstrated that equation 4 is (3) p = a(UN-U)+p” equivalent to the proposition that expected inflationwhere excess demand is now written as the gap is a geometrically declining weighted average of allbetween the natural and actual unemployment rates past rates of inflation with the weights summing toand pe is the price expectations variable representing one. This unit sum of weights ensures that any con-the anticipated rate ofg inflation. This expectations stant rate of inflation eventually will be fully antici-variable entered the equation with a coefficient of pated, as can be seen by writing the error-learningunity, reflecting the assumption that price expecta- mechanism astions are completely incorporated in actual pricechanges. The unit expectations coefficient implies (5) p” = &P--ithe absence of money illusion, i.e., it implies that where S indicates the operation of summing the pastpeople are concerned with the expected real purchas- rates of inflation, the subscript i denotes past timeing power of the prices they pay and receive (or, periods, and vi denotes the weights attached to pastalternatively, that they wish to maintain their prices rates of inflation. With a stable inflation rate prelative to the prices they expect others to be charg- unchanging over time and a unit sum of weights, theing) and so take anticipated inflation into account. equation’s right-hand side becomes simply p, indi-As will be shown later, the unit expectations coeffi- cating that when expectations are formulated adap-cient also implies the complete absence of a trade-off tively via the error-learning scheme, any constantbetween inflation and unemployment in long-run rate of inflation will indeed eventually be fully antici-equilibrium when expectations are fully realized. pated. Both versions of the adaptive-expectationsNote also that the expectations variable is the sole mechanism (i.e., equations 4 and 5) were combinedshift variable in the equation. All other shift vari- with the expectations-augmented Phillips equation toables have been omitted, reflecting the view, prevalent explain the mutual interaction of actual inflation,in the early 197Os, that changing price expectations expected inflation, and excess demand.were the predominant cause of observed shifts inthe Phillips curve. The Natural Rate HypothesisExpectations-Generating Mechanism These three innovations-the redefined excess de- mand variable, the expectations-augmented Phillips The third innovation was the incorporation of an curve, and the error-learning mechanism-formed theexpectations-generating mechanism into Phillips basis of the celebrated natural rate and accelerationistcurve analysis to explain how the price expectations hypotheses that radically altered economists’ andvariable itself was determined. Generally a simple policymakers’ views of the Phillips curve in the lateadaptive-expectations or error-learning mechanism 1960s and early 1970s. According to the naturalwas used. According to this mechanism, expecta- rate hypothesis, there exists no permanent trade-offtions are adjusted (adapted) by some fraction of the between unemployment and inflation since real eco-forecast error that occurs when inflation turns out nomic variables tend to be independent of nominalto be different than expected. In symbols, ones in steady-state equilibrium. To be sure, trade- offs may exist in the short run. For example, sur- ( 4 ) + = b(p-pe) prise inflation, if unperceived by wage earners, may,where the dot over the price expectations variable by raising product prices relative to nominal wagesindicates the rate of change (time derivative) of that and thus lowering real wages, stimulate employmentvariable, p-p” is the expectations or forecast error temporarily. But such trade-offs are inherently(i.e., the difference between actual and expected price transitory phenomena that stem from unexpectedinflation), and b is the adjustment fraction. Assum- inflation and that vanish once expectations (and theing, for example, an adjustment fraction of s, equa- wages and prices embodying them) fully adjust totion 4 says that if the actual and expected rates of inflationary experience. In the long run, when infla-inflation are 10 percent and 4 percent, respectively- tionary surprises disappear and expectations arei.e., the expectational error is 6 percent-then the realized such that wages reestablish their preexist-expected rate of inflation will be revised upward by ing levels relative to product prices, unemployment FEDERAL RESERVE BANK OF RICHMOND 11
  9. 9. returns to its natural (equilibrium) rate. This rateis compatible with all fully anticipated steady-staterates of inflation, implying that the long-run Phillips figure 7curve is a vertical line at the natural rate of unem- THE NATURAL RATEployment. HYPOTHESIS AND ADJUSTMENT Equation 3 embodies these conclusions. That equa- TO STEADY-STATE EQUILIBRIUMtion, when rearranged to read p-p”=a(Ux-U),states that the trade-off is between unexpected infla- p Price Inflation Rate btion (the difference between actual and expected Short-run Linflation, p-p”) and unemployment. That is, only Phillips Curvessurprise price increases could induce deviations of A / Long-run Verticalunemployment from its natural rate. The equationalso says that the trade-off disappears when inflation “I” I Phillips Curveis fully anticipated (i.e., when p-pe equals zero), a 1 Sl ’ Iresult guaranteed for any steady rate of inflation bythe error-learning mechanism’s unit sum of weights.Moreover, according to the equation, the right-handside must also be zero at this point, which impliesthat unemployment is at its natural rate. The naturalrate of unemployment is therefore compatible withany constant rate of inflation provided it is fullyanticipated (which it eventually must be by virtue ofthe error-learning weights adding to one). In short, -4equation 3 asserts that inflation-unemployment trade-offs cannot exist when inflation is fully anticipated.And equation 5 ensures that this latter conditionmust obtain for all steady inflation rates such that the of Unemployment ’’long-run Phillips curve is a vertical line at the natural The vertical line L through the natural raterate of unemployment.2 of unemployment UN is the long-run steady state Phillips curve along which all rates of The message of the natural rate hypothesis was inflation are fully anticipated. The down-clear. A higher stable rate of inflation could not ward-sloping lines are short-run Phillipsbuy a permanent drop in joblessness. Movements to curves each corresponding to a different given expected rate of inflation. Attemptsthe left along a short-run Phillips curve only provoke to lower unemployment from the naturalexpectational wage/price adjustments that shift the rate UN to U1 by raising inflation to 3 per-curve to the right and restore unemployment to its cent along the short-run tradeoff curve So will only induce shifts in the short-run curvenatural rate (see Figure 7). In sum, Phillips curve to S,, SR, SS as expectations adjust to thetrade-offs are inherently transitory phenomena. At- higher rate of inflation. The economytempts to exploit them will only succeed in raising travels the path ASIDE to the new steady state equilibrium, point E, where unemploy-the permanent rate of inflation without accomplish- ment is at its preexisting natural rate buting a lasting reduction in the unemployment rate. inflation is higher than it was originally.2 Actually, the long-run Phillips curve may become posi-tively sloped in its upper ranges as higher inflation leadsto greater inflation variability (volatility,unpredictability) that raises the natural Prate of unemployment. H i g h e r a n dhence more variable and erratic infla-tion can raise the equilibrium level of The Accelerationist Hypothesisunemployment by generating increased 0 u Uuncertainty that inhibits business ac- The expectations-augmented Phillips curve, whentivity and by introducing noise into market price signals, combined with the error-learning process, alsothus reducing the efficiency of the price system as acoordinating and allocating mechanism. yielded the celebrated accelerationist hypothesis that12 ECONOMIC REVIEW, MARCH/APRIL 1985
  10. 10. dominated many policy discussions in the inflationary1970s. This hypothesis, a corollary of the naturalrate concept, states that since there exists no long-run Figure 8trade-off between unemployment and inflation, at-tempts to peg the former variable below its natural THE ACCELERATIONIST(equilibrium) level must produce ever-increasing HYPOTHESISinflation. Fueled by progressively faster monetary % p Price Inflation Rateexpansion, such price acceleration would keep actualinflation always running ahead of expected infktion,thereby perpetuating the inflationary surprises thatprevent unemployment from returning to its equilib-rium level (see Figure 8). I -,, Long-run Vertical Q Accelerationists reached these conclusions via the P3’ I---following route. They note8 that equation 3 positsthat unemployment can differ from its natural levelonly so long as actual inflation deviates from ex- C P2’ I---pected inflation. But that same equation togetherwith equation 4 implies that, by the very nature of S2 B Ithe error-learning mechanism, such deviations cannot Pl ‘I---persist unless inflation is continually accelerated sothat it always stays ahead of expected inflation.3 If I -1inflation is not accelerated, but instead stays con- I A 0 ’ Ustant, then the gap between actual and expectedinflation will eventually be closed. Therefore acceler- “’ r”” ‘s,ation is required to keep the gap open if unemploy- Natural Ratement is to be maintained below its natural equilibrium of Unemployment 1level. In other words, the long-run trade-off impliedby the accelerationist hypothesis is between unem- Since the adjustment of expected to actual inflation works to restore unemployment toployment and the rate of acceleration of the inflation its natural equilibrium level UN a t a n y rate, in contrast to the conventional trade-off between steady rate of inflation, the authorities mustunemployment and the inflation rate itself as implied continually raise (accelerate) the inflationby the original Phillips curve.4 rate if they wish to peg unemployment at some arbitrarily low level such as U,. Such acceleration, by generating a continuousPolicy Implications of the Natural Rate succession of inflation surprises, perpetuallyand Accelerationist Hypotheses frustrates the full adjustment of expecta- tions that would return unemployment to At least two policy implications stemmed from the its natural rate. Thus attempts to pegnatural rate and accelerationist propositions. First, unemployment at U, will provoke explo- sive, ever-accelerating inflation. The economy will travel the path ABCD with3 Taking the time derivative of equation 3, then assuming the rate of inflation rising from zero to p,that the deviation of U from U, is pegged at a constant to p2 to P3 etc.level by the authorities such that its rate of change iszero, and then substituting equation 4 into the resultingexpression yields ; = b(p-pe)which says that the inflation rate must accelerate to stayahead of expected inflation.4The proof is simple. Merely substitute equation 3 into the authorities could either peg unemployment orthe expression presented in the preceding footnote to stabilize the rate of inflation but not both. If theyobtain i = ba(U,-U) pegged unemployment, they would lose control of the rate of inflation because the latter accelerateswhich says that the trade-off is between the rate of when unemployment is held below its natural level.acceleration of inflation i and unemployment U relativeto its natural rate. Alternatively, if they stabilized the inflation rate, FEDERAL RESERVE BANK OF RICHMOND 13
  11. 11. they would lose control of unemployment since the Iv. latter returns to its natural level at any steady STATISTICAL TESTS OF THE rate of inflation. Thus, contrary to the original NATURAL RATE HYPOTHESIS Phillips hypothesis, they could not peg unemployment at a given constant rate of inflation. They could, The preceding has examined the third stage of however, choose the steady-state inflation rate at Phillips curve analysis in which the natural rate hy- which unemployment returns to its natural level. pothesis was formed. The fourth stage involved A second policy implication stemming from the statistical testing of that hypothesis. These tests, natural rate hypoth&is was that the authorities could conducted in the early- to mid-1970s, led to criticisms choose from among alternative transitional adjust- of the adaptive-expectations or error-learning model ment paths to the desired steady-state rate of infla- of inflationary expectations and thus helped prepare tion. Suppose the authorities wished to move from a the way for the introduction of the alternative Qigh inherited inflation rate to a zero or other low rational expectations idea into Phillips curve analysis.. target inflation rate. To do so, they must lower The tests themselves were mainly concerned with inflationary expectations, a major determinant of the estimating the numerical value of the coefficient on inflation rate. But equations 3 and 4 state that the the price-expectations variable in the expectations- only way to lower expectations is to create slack augmented Phillips curve equation. If the coefficient capacity or excess supply in the economy. Such is one, as in equation 3, then the natural rate hypothe- slack raises unemployment above its natural level and sis is valid and no long-run inflation-unemployment thereby causes the actual rate of inflation to fall trade-off exists for the policymakers to exploit. But below the expected rate so as to induce a downward if the coefficient is less than one, the natural rate revision of the latter.5 The equations also indicate hypothesis is refuted and a long-run trade-off that how fast inflation comes down depends on the exists. Analysts emphasized this fact by writing the amount of slack created.e Much slack means fast expectations-augmented equation as adjustment and a relatively rapid attainment of the (6) P = a(UN-U)++pe inflation target. Conversely, little slack means slug- gish adjustment and a relatively slow attainment of where + is the coefficient (with a value of between the inflation target. Thus the policy choice is between zero and one) attached to the price expectations vari- adjustment paths offering high excess unemployment able. In long-run equilibrium, of course, expected for a short time or lower excess unemployment for a inflation equals actual inflation, i.e., p*=p. Setting long time (see Figure 9) .’ expected inflation equal to actual inflation as required for long-run equilibrium and solving for the actual rate of inflation yields s The proof is straightforward. Simply substitute equa- tion 3 into equation 4 to obtain (7) P = & (UN-U). $ = ba(U,-U). This expression says that expectations will be adjusted Besides showing that the long-run Phillips curve is downward (;e will be negative) only if unemployment steeper than its short-run counterpart (since the slope exceeds its natural rate. parameter of the former, a/(1-+), exceeds that of s Note that the equation developed in footnote 4 states the latter, a), equation 7 shows that a long-run trade- that disinflation will occur at a faster pace the larger the off exists only if the expectations coefficient + is less unemployment gap. than one. If the coefficient is one, however, the slope 7 Controls advocates proposed a third policy choice: use term is infinite, which means that there is no relation wage-price controls to hold actual below expected infla- tion so as to force a swift reduction of the latter. Over- between inflation and unemployment so that the looked was the fact that controls would have little impact trade-off vanishes (see Figure 10). on expectations unless the public was convinced that the trend of prices when controls were in force was a reliable Many of the empirical tests estimated the coeffi- indicator of the future orice trend after controls were cient to be less than unity and concluded that the lifted. Convincing the piblic would be difficult if controls had failed to stop inflation in the past. Aside from this, natural rate hypothesis was invalid. But this con- it is hard to see why controls should have a stronger clusion was sharply challenged by economists who impact on expectations than a preannounced, demon- strated policy of disinflationary money growth. contended that the tests contained statistical bias that 14 ECONOMIC REVIEW, MARCH/APRIL 1985
  12. 12. Figure 9 ALTERNATIVE DISINFLATION PATHS p Price Inflation Rate p Price Inflation Rate I SA L Long-run Vertical Phillips Curve Initial Short-run Phillips Curve for Expected Inflation pa ‘U High Excess / ’ ACB = Fast disinflation path involving high ADEB = Gradualist disinflation path involv- excess unemployment for a short ing low excess unemployment for time. a long time. To move from high-inflation point A to zero-inflation point B the authorities must first travel along short-run Phillips curve SA, lowering actual relative to expected inflation and thereby inducing the downward revision of expectations that shifts the short-run curve leftward until point B is reached. Since the speed of adjustment of expectations depends upon the size of the unemployment gap, it follows that point B will be reached faster via the high excess unem- ployment path ACB than via the low excess unemployment path ADEB. The choice is between high excess unemployment for a short time or low excess unemployment for a long time.tended to work against the natural rate hypothesis. Shortcomings of the Adaptive-ExpectationsThese critics pointed out that the tests typically used Assumptionadaptive-expectations schemes as empirical proxies In connection with the foregoing, the critics arguedfor the unobservable price expectations variable. that the adaptive-expectations scheme is a grosslyThey further showed that if these proxies were in-appropriate measures of inflationary expectations inaccurate representation of how people formulatethen estimates of the expectations coefficient could price expectations. They pointed out that it postu-well be biased downward. If so, then estimated coeffi- lates naive expectational behavior, holding as it doescients of less than one constituted no disproof of the that people form anticipations solely from a weightednatural rate hypothesis. Rather they constituted evi- average of past price experience with weights thatdence of inadequate measures of expectations. are fixed and independent of economic conditions and FEDERAL RESERVE BANK OF RICHMOND 15
  13. 13. tional errors. That people would fail to exploit infor- mation that would improve expectational accuracy seems implausible, however. In short, the critics Figure 10 contended that adaptive expectations are not wholly THE EXPECTATIONS rational if other information besides past price COEFFICIENT AND THE changes can improve inflation predictions. LONG-RUN STEADY-STATE Many economists have since pointed out that it is PHILLIPS CURVE hard to accept the notion that individuals would con- ‘tinually form price anticipations from any scheme p Price Inflation Rate that is inconsistent with the way inflation is actually I Long-run generated in the economy. Being different from the SteadyState true inflation-generating mechanism, such schemes Phillips Curve: will produce expectations that are systematically wrong. If so, rational forecasters will cease to use them. For example, suppose inflation were actually accelerating or decelerating. According to equation 5, the adaptive-expectations model would systematically underestimate the inflation rate in the former case and overestimate it in the latter. Using a unit weighted average of past inflation rates to forecast a steadily rising or falling rate would yield a succes- sion of one-way errors. The discrepancy between actual and expected inflation would persist in a per- fectly predictable way such that forecasters would be provided free the information needed to correct their mistakes. Perceiving these persistent expecta- tional mistakes, rational individuals would quickly abandon the error-learning model for more accurate expectations-generating schemes. Once again, the Statistical tests of the natural rate hypo- thesis sought to determine the magnitude adaptive-expectations mechanism is implausible be- of the expectations coefficient @J in the cause of its incompatibility with rational behavior. lonprun steady-state Phillips curve equation P= * (U,-U). V. A coefficient of one means that no perma- nent tradeoff exists and the steady-state FROM ADAPTIVE EXPECTATIONS TO Phillips curve is a vertical line through the RATIONAL EXPECTATIONS natural rate of unemployment. Conversely, a coefficient of less than one signifies the The shortcomings of the adaptive-expectations existence of a long-run Phillips curve trade approach to the modeling of expectations led to the off with negative slope for the policymakers incorporation of the alternative rational expectations to exploit. Note that the long-run curves are steeper than the short-run ones, indi- approach into Phillips curve analysis. According to cating that permanent trade-offs are less the rational expectations hypothesis, individuals will favorable than temporary ones. tend to exploit all available pertinent information about the inflationary process when making their price forecasts. If true, this means that forecasting errors ultimately could arise only from random (unforeseen) shocks occurring to the economy. At first, of course, price forecasting errors might alsopolicy actions. It implies that people look only at arise because individuals initially possess limited orpast price changes and ignore all other pertinent incomplete information about, say, an unprecedentedinformation-e.g., money growth rate changes, ex- new policy regime, economic structure, or inflation-change rate movements, announced policy intentions generating mechanism. But it is unlikely that thisand the like-that could be used to reduce expecta- condition would persist. For if the public were16 ECONOMIC REVIEW, MARCH/APRIL 1985
  14. 14. truly rational, it would quickly learn from these infla- Policy actions, to the extent they are systematic,tionary surprises or prediction errors (data on which are predictable. Systematic policies are simply feed-it acquires costlessly as a side condition of buying back rules or response functions relating policy vari-goods) and incorporate the free new information ables to past values of other economic variables.into its forecasting procedures, i.e., the source of These policy response functions can be estimated andforecasting mistakes would be swiftly perceived and incorporated into forecasters’ price predictions. Insystematically eradicated. As knowledge of policy other words, rational individuals can use past obser-and the inflationary process improved, forecasting vations on the behavior of the authorities to discovermodels would be continually revised to produce moreaccurate predictions. Soon all systematic (predict- the policy rule. Once they know the rule, they can)able) elements influencing the rate of inflation would use current observations on the variables to whichbecome known and fully understood, and individuals’ the policymakers respond to predict future policyprice expectations would constitute the most accu- moves. Then, on the basis of these predictions, theyrate (unbiased) forecast consistent with that knowl- can correct for the effect of anticipated policies be-edge.* When this happened the economy would con- forehand by making appropriate adjustments to nomi-verge to its rational expectations equilibrium and nal wages and prices. Conseq’uently, when stabiliza-people’s price expectations would be the same as tion actions do occur, they will have no impact onthose implied by the actual inflation-generating mech- real variables like unemployment since they will haveanism. As incorporated in natural rate Phillips curve been discounted and neutralized in advance. In short,models, the rational expectations hypothesis implies rules-based policies, being in the information set usedthat thereafter, except for unavoidable surprises due by rational forecasters, will be perfectly anticipatedto purely random shocks, price expectations wouldalways be correct and the economy would always be and for that reason will have no impact on unemploy-at its long-run steady-state equilibrium. ment. The only conceivable way that policy can have even a short-run influence on real variables is for itPolicy Implications of Rational Expectations to be unexpected, i.e., the policymakers must either act in an unpredictable random fashion or secretly The strict (flexible price, instantaneous market change the policy rule. Apart from such tactics,clearing) rational expectations approach has radical which are incompatible with most notions of thepolicy implications. When incorporated into natural proper conduct of public policy, there is no way therate Phillips curve equations, it implies that system-atic policies-i.e., those based on feedback control authorities can influence real variables, i.e., causerules defining the authorities’ response to changes in them to deviate from their natural equilibrium levels.the economy-cannot influence real variables such as The authorities can, however, influence a nominaloutput and unemployment even in the short run, variable, namely the inflation rate, and should con-since people would have already anticipated what the centrate their efforts on doing so if some particularpolicies are going to be and acted upon those antici- rate (e.g., zero) is desired.pations. To have an impact on output and employ- As for disinflation strategy, the rational expecta-ment, the authorities must be able to create a diver- tions approach generally calls for a preannouncedgence between actual and expected inflation. This sharp swift reduction in money growth-provided offollows from the proposition that inflation influences course that the government’s commitment to endingreal variables only when it is unanticipated. To lowerunemployment in the Phillips curve equation p-p”= inflation is sufficiently credible to be believed. Hav-a (UN-U), the authorities must be able to alter the ing chosen a zero target rate of inflation and havingactual rate of inflation without simultaneously causing convinced the public of their determination to achievean identical change in the expected future rate. This it, the policy authorities should be able to do somay be impossible if the public can predict policy without creating a costly transitional rise in unem-actions. ployment. For, given that rational expectations adjust infinitely faster than adaptive expectations to a credible preannounced disinflationary policy (and8 Put differently, rationality implies that current expecta-tional errors are uncorrelated with past errors and with also that wages and prices adjust to clear marketsall other known information, such correlations already continuously) the transition to price stability shouldhaving been perceived and exploited in the process ofimproving price forecasts. be relatively quick and painless (see Figure 11). FEDERAL RESERVE BANK OF RICHMOND 17
  15. 15. natural rate Phillips curve models. Under adaptive- expectations, short-run trade-offs exist because such expectations, being backward looking and slow to Figure 11 respond, do not adjust instantaneously to elimi- COSTLESS DISINFLATION nate forecast errors arising from policy-engineered UNDER RATIONAL changes in the inflation rate. With expectations EXPECTATIONS AND adapting to actual inflation with a lag, monetary policy can generate unexpected inflation and conse- POLICY CREDIBILITY quently influence reel variables in the short run. This cannot happen under rational expectations where both actual and expected inflation adjust identically p Price Inflation Rate Steady-State and instantaneously to anticipated policy changes. I In short, under rational expectations, systematic policy cannot induce the expectational errors that generate short-run Phillips curves.e Phillips curves may exist, to be sure. But they are purely adventi- tious phenomena that are entirely the result of unpre- dictable random shocks and cannot be exploited by policies based upon rules. In sum, no role remains for systematic counter- Disinflation Path cyclical stabilization policy in Phillips curve models embodying rational expectations and the natural rate hypothesis. The only thing such policy can influ- Inflation Rate ence in these models is the rate of inflation which adjusts immediately to expected changes in money growth. Since the models teach that the full effect 0 U of rules-based policies is on the inflation rate, it follows that the authorities-provided they believe that the models are at all an accurate representation of the way the world works-should concentrate Assuming expectational rationality, wage/ their efforts on controlling that nominal inflation price flexibility, and full policy credibility, a variable since they cannot systematically influence preannounced permanent reduction in money growth to a level consistent with real variables. These propositions are demonstrated Stable prices theoretically lowers expected with the aid of the expository model presented in the and thus actual inflation to zero with no Appendix on page 21. accompanying transitory rise in unemploy- ment. The economy moves immediately from point A to point B on thevertical steady-state Phillips curve. Here is the basic prediction of VI. the rational expectations-natural rate model: that fully anticipated policy changes EVALUATION OF RATIONAL EXPECTATIONS (including credible preannounced ones) affect only inflation but not output and The preceding has shown how the rational expec- employment. tations assumption combines with the natural rate hypothesis to yield the policy-ineffectiveness conclu- sion that no Phillips curves exist for policy to exploitNo Exploitable Trade-Offs s Note that the rationat expectations hypothesis also rules out the accelerationist notion of a stable trade-off between unemployment and the rate of acceleration of the inflation To summarize, the rationality hypothesis, in con- rate. If exnectations are formed consistentlv with thejunction with the natural rate hypothesis, denies the way inflatioi is actually generated, the authbrities will not be able to fool people by accelerating inflation or byexistence of exploitable Phillips curve trade-offs in accelerating the rate bf acceleration, etc. Indeed, nbthe short run as well as the long. In so doing, it systematic policy will work if expectations are formed consistently with the way inflation is actually generateddiffers from the adaptive-expectations version of in the economy.18 ECONOMIC REVIEW, MARCH/APRIL 1985
  16. 16. even in the short run. Given the importance of the expectations are basically nonrational, i.e., that most rational expectations component in modern Phillips people are too naive or uninformed to formulate un- curve analysis, an evaluation of that component is biased price expectations. Overlooked is the coun- now in order. terargument that relatively uninformed people often One advantage of the rational expectations hy- delegate the responsibility for formulating rationalpothesis is that it treats expectations formation as a forecasts to informed specialists and that professional part of optimizing behavior. By so doing, it brings forecasters, either through their ability to sell supe- the theory of price anticipations into accord with the rior forecasts or to act in behalf of those without rest of economic analysis. The latter assumes that same, will ensure that the economy will behave as ifpeople behave as rational optimizers in the production all people were rational. One can also note that theand purchase of goods, in the choice of jobs, and in rational expectations hypothesis is merely an impli- the making of investment decisions. For consistency, cation of the uncontroversial assumption of profitit should assume the same regarding expectational (and utility) maximization and that, in any case,behavior. economic analysis can hardly proceed without the In this sense, the rational expectations theory is rationality assumption. Other critics insist, however,superior to rival explanations, all of which imply that that expectational rationality cannot hold during theexpectations may be consistently wrong. It is the transition to new policy regimes or other structuralonly theory that denies that people make systematic changes in the economy since it requires a long timeexpectation errors. Note that it does not claim that to understand such changes and learn to adjust topeople possess perfect foresight or that their expec- them. Against this is the counterargument that suchtations are always accurate. What it does claim is changes and their effects are often foreseeable fromthat they perceive and eliminate regularities in their the economic and political events that precede themforecasting mistakes. In this way they discover the and that people can quickly learn to predict regimeactual inflation generating process and use it in form- changes just as they learn to predict the workings of aing price expectations. And with the public’s rational given regime. This is especially so when regimeexpectations of inflation being the same as the mean changes have occurred in the past. Having experi-value of the inflation generating process, those expec- enced such changes, forecasters will be sensitive totations cannot be wrong on average. Any errors will their likely future random, not systematic. The same cannot be said Most of the criticism, however, is directed not atfor other expectations schemes, however. Not being the rationality assumption per se but rather atidentical to the expected value of the true inflation another key assumption underlying its policy-generating process, those schemes will produce biased ineffectiveness result, namely the assumption of noexpectations that are systematically wrong. policymaker information or maneuverability advan- tage over the private sector. This assumption states Biased expectations schemes are difficult to justifytheoretically. Systematic mistakes are harder to that private forecasters possess exactly the sameexplain than is rational behavior. True, nobody information and the ability to act upon it as do thereally knows how expectations are actually formed. authorities. Critics hold that this assumption is im-But a theory that says that forecasters do not con- plausible and that if it is violated then the policytinually make the same mistakes seems intuitively ineffectiveness result ceases to hold. In this case, anmore plausible than theories that imply the opposite. exploitable short-run Phillips curve reemerges, allow-Considering the profits to be made from improved ing some limited scope for systematic monetary poli-forecasts, it seems inconceivable that systematic ex- cies to reduce unemployment.pectational errors would persist. Somebody would For example, suppose the authorities possess moresurely notice the errors, correct them, and profit by and better information than the public. Having thisthe corrections. Together, the profit motive and information advantage, they can predict and hencecompetition would reduce forecasting errors to ran- respond to events seen as purely random by thedomness. public. These policy responses will, since they are unforeseen by the public, affect actual but not ex-Criticisms of the Rational Expectations pected inflation and thereby change unemploymentApproach relative to its natural rate in the (inverted) Phillips curve equation UN-U= ( l/a) (p-p”). Despite its logic, the rational expectations hypothe- Alternatively, suppose that both the authoritiessis still has many critics. Some still maintain that and the public possess identical information but that FEDERAL RESERVE BANK OF RICHMOND 19
  17. 17. the latter group is constrained by long-term con- it and make it freely available. Finally, if contractualtractual obligations from exploiting that information. wages and prices are sticky and costly to adjust, thenFor example, suppose workers and employers make the authorities should minimize these price adjust-labor contracts that fix nominal wages for a longer ment costs by following policies that stabilize theperiod of time than the authorities require to change general price level.the money stock. With nominal wages fixed and In short, advocates of the rational expectationsprices responding to money, the authorities are in a approach argue that feasibility alone constitutes in-position to lower real wages and thereby stimulateemployment with an inflationary monetary policy. sufficient justification for activist policies. Policies should also be gocially beneficial. Activist policies In these ways, contractual and informational con- hardly satisfy this latter criterion since their effective-straints are alleged to create output- and employment- ness is based on deceiving people into making expec-stimulating opportunities for systematic stabilization tational errors. The proper role for policy is not topolicies. Indeed, critics have tried to demonstrate as influence real activity via deception but rather tomuch by incorporating such constraints into rational . reduce information deficiencies, to eliminate erraticexpectations Phillips curve models similar to the one . variations of the variables under the policymakers’outlined in the Appendix of this article. control, and perhaps also to minimize the costs of Proponents of the rational expectations approach, adjusting prices.however, doubt that such constraints can restore thepotency of activist policies and generate exploitablePhillips curves. They contend that policymaker VII.information advantages cannot long exist when gov-ernment statistics are published immediately upon CONCLUDING COMMENTScollection, when people have wide access to datathrough the news media and private data services, The preceding paragraphs have traced the evolu-and when even secret policy changes can be pre- tion of Phillips curve analysis. The chief conclusionsdicted from preceding observable (and obvious) can be stated succinctly. The Phillips curve concepteconomic and political pressures. Likewise, they has changed radically over the past 25 years as thenote that fixed contracts permit monetary policy to notion of a stable enduring trade-off has given wayhave real effects only if those effects are so inconse- to the policy-ineffectiveness view that no such trade-quential as to provide no incentive to renegotiate off exists for the policymakers to exploit. Instru-existing contracts or to change the optimal type of mental to this change were the natural rate and contract that is negotiated. And even then, they note, rational expectations hypotheses, respectively. Thesuch monetary changes become ineffective when the former says that trade-offs arise solely from expec-contracts expire. More precisely, they question the tational errors while the latter holds that systematic whole idea of fixed contracts that underlies the sticky macroeconomic stabilization policies, by virtue ofwage case for policy activism. They point out that their very predictability, cannot possibly generatecontract duration is not invariant to the type of policy such errors. Taken together, the two hypotheses being pursued but rather varies with it and thus imply that systematic demand management policies provides a weak basis for activist fine-tuning. are incapable of influencing real activity, contrary to Finally, they insist that such policies, even if effec- the predictions of the original Phillips curve analysis.tive, are inappropriate. In their view, the proper role On the positive side, the two hypotheses do implyfor policy is not to exploit informational and con- that the government can contribute to economic sta-tractual constraints to systematically influence real bility by following policies to minimize the expecta-activity but rather to neutralize the constraints or to tional errors that cause output and employment tominimize the costs of adhering to them. Thus if deviate from their normal full-capacity levels. Forpeople form biased price forecasts, then the policy- example, the authorities could stabilize the price levelmakers should publish unbiased forecasts. And if the so as to eliminate the surprise inflation that generatespolicy authorities have informational advantages over confusion between absolute and relative prices andprivate individuals, they should make that informa- that leads to perception errors. Similarly, they couldtion public rather than attempting to exploit the ad- direct their efforts at minimizing random and erraticvantage. That is, if information is costly to collect variations in the monetary variables under their con-and process, then the central authority should gather trol. In so doing, not only would they lessen the20 ECONOMIC REVIEW, MARCH/APRIL 1985