advances in business cycle theory - …bcs.worthpublishers.com/webpub/economics/mankiw7/6th...

Advances in Business Cycle Theory

C H A P T E R

19Every great advance in science has issued from a new audacity of imagination.

— John Dewey

Your theory is crazy, but it’s not crazy enough to be true.

— Niels Bohr

What is the best way to explain short-run fluctuations in output andemployment? How should monetary and fiscal policy respond to thesefluctuations? Most economists believe that these questions are best

answered using the model of aggregate demand and aggregate supply, which thisbook has developed and applied thoroughly.Yet as we approach the end of thebook, let’s take a step closer to the frontier of modern economic research andexamine the continuing debate over the theory of short-run economic fluctua-tions.This chapter discusses two recent strands of research—real business cycletheory and new Keynesian economics.

We begin by examining the theory of real business cycles—a viewpoint heldby a small but significant minority of economists.According to this theory, short-run economic fluctuations should be explained while maintaining the assump-tions of the classical model, which we have used to study the long run. Mostimportant, real business cycle theory assumes that prices are fully flexible, evenin the short run. Almost all microeconomic analysis is based on the premise thatprices adjust to clear markets. Advocates of real business cycle theory argue thatmacroeconomic analysis should be based on the same assumption.

Because real business cycle theory assumes complete price flexibility, it is con-sistent with the classical dichotomy: in this theory, nominal variables, such as themoney supply and the price level, do not influence real variables, such as outputand employment.To explain fluctuations in real variables, real business cycle the-ory emphasizes real changes in the economy, such as changes in production tech-nologies. The “real” in real business cycle theory refers to the theory’s exclusionof nominal variables in explaining short-run economic fluctuations.

By contrast, new Keynesian economics is based on the premise that market-clearing models such as real business cycle theory cannot explain short-run eco-nomic fluctuations. In The General Theory, Keynes urged economists to abandon

528 |

528-546_Mankiw_CH19.qxp 3/28/06 2:32 PM Page 528

© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

the classical presumption that wages and prices adjust quickly to equilibrate mar-kets. He emphasized that aggregate demand is a primary determinant of nation-al income in the short run. New Keynesian economists accept these basicconclusions, and so they advocate models with sticky wages and prices.

In their research, new Keynesian economists try to develop more fully theKeynesian approach to economic fluctuations. Many new Keynesians accept theIS-LM model as the theory of aggregate demand and, in their research, try torefine the theory of aggregate supply. This work tries to explain how wages andprices behave in the short run by identifying more precisely the market imper-fections that make wages and prices sticky and that cause output to deviate fromits natural level. We discuss this research in the second half of the chapter.

In presenting the work of these two schools of thought, this chapter takes anapproach that is more descriptive than analytic. Studying recent theoreticaldevelopments in detail would require more mathematics than is appropriate forthis book.Yet, even without the formal models, we can discuss the direction ofthis research and get a sense of how different economists are applying microeco-nomic thinking to better understand macroeconomic fluctuations.1

19-1 The Theory of Real Business Cycles

When we studied economic growth in Chapters 7 and 8, we described a rela-tively smooth process. Output grew as population, capital, and the available tech-nology evolved over time. In the Solow growth model, the economy approachesa steady state in which most variables grow together at a rate determined by theconstant rate of technological progress.

But is the process of economic growth necessarily as steady as the Solowmodel assumes? Perhaps technological progress and economic growth occurunevenly. Perhaps there are shocks to the economy that induce short-run fluc-tuations in the natural levels of output and employment. To see how this mightbe so, we consider a famous allegory that economists have borrowed from authorDaniel Defoe.

The Economics of Robinson CrusoeRobinson Crusoe is a sailor stranded on a desert island. Because Crusoe livesalone, his life is simple. Yet he has to make many economic decisions. Con-sidering Crusoe’s decisions—and how they change in response to changingcircumstances—sheds light on the decisions that people face in larger, morecomplex economies.

C H A P T E R 1 9 Advances in Business Cycle Theory | 529

1 For a more formal treatment of the issues discussed here, see the superb graduate-level textbookby David Romer, Advanced Macroeconomics, 3rd ed. (New York: McGraw-Hill, 2006).


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

To keep things simple, imagine that Crusoe engages in only a few activities.Crusoe spends some of his time enjoying leisure, perhaps swimming at his island’sbeaches. He spends the rest of his time working, either catching fish or collect-ing vines to make into fishing nets. Both forms of work produce a valuable good:fish are Crusoe’s consumption, and nets are Crusoe’s investment. If we were tocompute GDP for Crusoe’s island, we would add together the number of fishcaught and the number of nets made (weighted by some “price” to reflect Cru-soe’s relative valuation of these two goods).

Crusoe allocates his time among swimming, fishing, and making nets based onhis preferences and the opportunities available to him. It is reasonable to assumethat Crusoe optimizes.That is, he chooses the quantities of leisure, consumption,and investment that are best for him given the constraints that nature imposes.

Over time, Crusoe’s decisions change as shocks impinge on his life. For exam-ple, suppose that one day a big school of fish passes by the island. GDP rises inthe Crusoe economy for two reasons. First, Crusoe’s productivity rises: with alarge school in the water, Crusoe catches more fish per hour of fishing. Second,Crusoe’s employment rises. That is, he decides to reduce temporarily his enjoy-ment of leisure to work harder and take advantage of this unusual opportunityto catch fish. The Crusoe economy is booming.

Similarly, suppose that a storm arrives one day. Because the storm makes out-door activity difficult, productivity falls: each hour spent fishing or making netsyields a smaller output. In response, Crusoe decides to spend less time workingand to wait out the storm in his hut. Consumption of fish and investment in netsboth fall, so GDP falls as well. The Crusoe economy is in recession.

Suppose that one day Crusoe is attacked by natives. While he is defendinghimself, Crusoe has less time to enjoy leisure. Thus, the increased demand fordefense spurs employment in the Crusoe economy, especially in the “defenseindustry.”To some extent, Crusoe spends less time fishing for consumption.To alarger extent, he spends less time making nets, because this task is easy to put offfor a while. Thus, defense spending crowds out investment. Because Crusoespends more time at work, GDP (which now includes the value of nationaldefense) rises. The Crusoe economy is experiencing a wartime boom.

What is notable about this story of booms and recessions is its simplicity. Inthis story, fluctuations in output, employment, consumption, investment, and productivityare all the natural and desirable response of an individual to the inevitable changes in hisenvironment. In the Crusoe economy, fluctuations have nothing to do with mon-etary policy, sticky prices, or any type of market failure.

According to the theory of real business cycles, fluctuations in our economyare much the same as fluctuations in Robinson Crusoe’s. Shocks to our ability toproduce goods and services (like the changing weather on Crusoe’s island) alterthe natural levels of employment and output. These shocks are not necessarilydesirable, but they are inevitable. Once the shocks occur, it is desirable for GDP,employment, and other real macroeconomic variables to fluctuate in response.

The parable of Robinson Crusoe, like any model in economics, is not intend-ed to be a literal description of how the economy works. Instead, it tries to get atthe essence of the complex phenomenon that we call the business cycle. Does the

530 | P A R T V I More on the Microeconomics Behind Macroeconomics


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate


parable achieve this goal? Are the booms and recessions in modern industrialeconomies really like the fluctuations on Robinson Crusoe’s island? Economistsdisagree about the answer to this question and, therefore, disagree about the valid-ity of real business cycle theory. At the heart of the debate are four basic issues:

➤ The interpretation of the labor market: Do fluctuations in employmentreflect voluntary changes in the quantity of labor supplied?

➤ The importance of technology shocks: Does the economy’s productionfunction experience large, exogenous shifts in the short run?

➤ The neutrality of money: Do changes in the money supply have onlynominal effects?

➤ The flexibility of wages and prices: Do wages and prices adjust quicklyand completely to balance supply and demand?

Regardless of whether you view the parable of Robinson Crusoe as a plausibleallegory for the business cycle, considering these four issues is instructive, becauseeach of them raises fundamental questions about how the economy works.

The Interpretation of the Labor MarketReal business cycle theory emphasizes the idea that the quantity of labor sup-plied at any given time depends on the incentives that workers face. Just asRobinson Crusoe changes his work effort voluntarily in response to changingcircumstances, workers are willing to work more hours when they are wellrewarded and are willing to work fewer hours when they are poorly rewarded.Sometimes, if the reward for working is sufficiently small, workers choose toforgo working altogether—at least temporarily. This willingness to reallocatehours of work over time is called the intertemporal substitution of labor.

To see how intertemporal substitution affects labor supply, consider the fol-lowing example. A college student finishing her sophomore year has two sum-mer vacations left before graduation. She wishes to work for one of thesesummers (so she can buy a car after she graduates) and to relax at the beach dur-ing the other summer. How should she choose which summer to work?

Let W1 be her real wage in the first summer and W2 the real wage she expectsin the second summer.To choose which summer to work, the student comparesthese two wages. Yet, because she can earn interest on money earned earlier, adollar earned in the first summer is more valuable than a dollar earned in the sec-ond summer. Let r be the real interest rate. If the student works in the first sum-mer and saves her earnings, she will have (1 + r)W1 a year later. If she works inthe second summer, she will have W2. The intertemporal relative wage—that is,the earnings from working the first summer relative to the earnings from work-ing the second summer—is

Intertemporal Relative Wage = .(1 + r)W1⎯

W2


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

Working the first summer is more attractive if the interest rate is high or if thewage is high relative to the wage expected to prevail in the future.

According to real business cycle theory, all workers perform this cost-benefitanalysis when deciding whether to work or to enjoy leisure. If the wage is tem-porarily high or if the interest rate is high, it is a good time to work. If the wageis temporarily low or if the interest rate is low, it is a good time to enjoy leisure.

Real business cycle theory uses the intertemporal substitution of labor toexplain why employment and output fluctuate. Shocks to the economy thatcause the interest rate to rise or the wage to be temporarily high cause people towant to work more; the increase in work effort raises employment and produc-tion. Shocks that cause the interest rate to fall or the wage to be temporarily lowdecrease employment and production.

Critics of real business cycle theory believe that fluctuations in employmentdo not reflect changes in the amount people want to work. They believe thatdesired employment is not very sensitive to the real wage and the real interest rate.They point out that the unemployment rate fluctuates substantially over thebusiness cycle.The high unemployment in recessions suggests that the labor mar-ket does not clear: if people were voluntarily choosing not to work in recessions,they would not call themselves unemployed. These critics conclude that wagesdo not adjust to equilibrate labor supply and labor demand, as real business cyclemodels assume.

In reply, advocates of real business cycle theory argue that unemployment sta-tistics are difficult to interpret.The mere fact that the unemployment rate is highdoes not mean that intertemporal substitution of labor is unimportant. Individ-uals who voluntarily choose not to work may call themselves unemployed sothey can collect unemployment-insurance benefits. Or they may call themselvesunemployed because they would be willing to work if they were offered thewage they receive in most years.


Looking for Intertemporal Substitution

Because intertemporal substitution of labor is central to real business cycle the-ory, much research has been aimed at examining whether it is an importantdeterminant of labor supply. This research looks at data on wages and hours tosee whether people alter the amount they work in response to small changes inthe real wage. If leisure were highly intertemporally substitutable, then individ-uals expecting increases in the real wage should work little today and much inthe future. Those expecting decreases in their real wage should work hard todayand enjoy leisure in the future.

Most studies of labor supply find that expected changes in the real wage leadto only small changes in hours worked. Individuals appear not to respond toexpected real-wage changes by substantially reallocating leisure over time. Thisevidence suggests that intertemporal substitution is not as important a determi-nant of labor supply as real business cycle theorists claim.

C A S E S T U D Y

528-546_Mankiw_CH19.qxp 3/28/06 2:32 PM 32

© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

This evidence does not convince everyone, however. One reason is that thedata are often far from perfect. For example, to study labor supply, we need dataon wages; yet when a person is not working, we do not observe the wage thatperson could have earned by taking a job. Thus, although most studies of laborsupply find little evidence for intertemporal substitution, they do not end thedebate over real business cycle theory.2

The Importance of Technology ShocksThe Crusoe economy fluctuates because of changes in the weather, whichinduce Crusoe to alter his work effort. In real business cycle theory, the anal-ogous variable is technology, which determines an economy’s ability to turninputs (capital and labor) into output (goods and services).The theory assumesthat our economy experiences fluctuations in technology and that these fluc-tuations in technology cause fluctuations in output and employment. Whenthe available production technology improves, the economy produces moreoutput, and real wages rise. Because of intertemporal substitution of labor, theimproved technology also leads to greater employment. Real business cycletheorists often explain recessions as periods of “technological regress.”Accord-ing to these models, output and employment fall during recessions because theavailable production technology deteriorates, lowering output and reducingthe incentive to work.

Critics of real business cycle theory are skeptical that the economy experi-ences large shocks to technology. It is a more common presumption that tech-nological progress occurs gradually. Critics argue that technological regress isespecially implausible: the accumulation of technological knowledge may slowdown, but it is hard to imagine that it would go in reverse.

Advocates respond by taking a broad view of shocks to technology. Theyargue that there are many events that, although not literally technological,affect the economy much as technology shocks do. For example, bad weather,the passage of strict environmental regulations, or increases in world oil priceshave effects similar to adverse changes in technology: they all reduce our abil-ity to turn capital and labor into goods and services. Whether such events aresufficiently common to explain the frequency and magnitude of business cyclesis an open question.


2 The classic article emphasizing the role of intertemporal substitution in the labor market isRobert E. Lucas, Jr. and Leonard A. Rapping,“Real Wages, Employment, and Inflation,” Journal ofPolitical Economy 77 (September/October 1969): 721–754. For some of the empirical work thatcasts doubt on this hypothesis, see Joseph G. Altonji,“Intertemporal Substitution in Labor Supply:Evidence From Micro Data,’’ Journal of Political Economy 94 ( June 1986, Part 2): S176–S215; andLaurence Ball,“Intertemporal Substitution and Constraints on Labor Supply: Evidence From PanelData,’’ Economic Inquiry 28 (October 1990): 706–724. For a recent study reporting evidence in favorof the hypothesis, see Casey B. Mulligan, “Substitution Over Time: Another Look at Life CycleLabor Supply,” NBER Macroeconomics Annual 13 (1998).


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate


The Solow Residual and the Business Cycle

To demonstrate the role of technology shocks in generating business cycles,economist Edward Prescott looked at data on the economy’s inputs (capitaland labor) and its output (GDP). For every year, he computed the Solowresidual—the percentage change in output minus the percentage change ininputs, where the different inputs are weighted by their factor shares. TheSolow residual measures the portion of output growth that cannot beexplained by growth in capital or labor. Prescott interprets it as a measure ofthe rate of technological progress.3

Figure 19-1 shows the Solow residual and the growth in output for the peri-od 1960 to 2002. Notice that the Solow residual fluctuates substantially. It tellsus, for example, that technology worsened in 1982 and improved in 1984. Inaddition, the Solow residual moves closely with output: in years when outputfalls, technology worsens. According to Prescott, these large fluctuations in theSolow residual show that technology shocks are an important source of eco-nomic fluctuations.

Prescott’s interpretation of this figure is controversial, however. Many econo-mists believe that the Solow residual does not accurately represent changes intechnology over short periods of time. The standard explanation of the cyclicalbehavior of the Solow residual is that it results from two measurement problems.

First, during recessions, firms may continue to employ workers they do notneed, so that they will have these workers on hand when the economy recovers.This phenomenon, called labor hoarding, means that labor input is overesti-mated in recessions, because the hoarded workers are probably not working ashard as usual. As a result, the Solow residual is more cyclical than the availableproduction technology. In a recession, productivity as measured by the Solowresidual falls even if technology has not changed simply because hoarded work-ers are sitting around waiting for the recession to end.

Second, when demand is low, firms may produce things that are not easilymeasured. In recessions, workers may clean the factory, organize the inventory,get some training, and do other useful tasks that standard measures of output failto include. If so, then output is underestimated in recessions, which would alsomake the measured Solow residual cyclical for reasons other than technology.

Thus, economists can interpret the cyclical behavior of the Solow residual indifferent ways. Real business cycle theorists point to the low productivity in reces-sions as evidence for adverse technology shocks. Other economists believe thatmeasured productivity is low in recessions because workers are not working ashard as usual and because more of their output is not measured. Unfortunately,

C A S E S T U D Y

3 The appendix to Chapter 8 shows that the Solow residual is

= − a − (1 − a) ,

where A is total factor productivity, Y output, K capital, L labor, and a capital’s share of income.

DL⎯L

DK⎯K

DY⎯Y

DA⎯A


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

there is no clear evidence on the importance of labor hoarding and the cyclicalmismeasurement of output. Therefore, different interpretations of Figure 19-1persist.This disagreement is one part of the debate between advocates and criticsof real business cycle theory.4

The Neutrality of MoneyJust as money has no role in the Crusoe economy, real business cycle theoryassumes that money in our economy is neutral, even in the short run. That is,monetary policy is assumed not to affect real variables such as output andemployment. Not only does the neutrality of money give real business cycle the-ory its name, but neutrality is also the theory’s most radical assumption.


Percent per year

Year

Solow residual

1965 1970 1975 1980 1985 1990 1995 20001960

8

6

4

2

0

�2

�4

Output growth

Growth in Output and the Solow Residual The Solow residual, which some economistsinterpret as a measure of technology shocks, fluctuates with the economy’s output ofgoods and services.

Source: U.S. Department of Commerce, U.S. Department of Labor, and author’s calculations.

figure 19-1

4 For the two sides of this debate, see Edward C. Prescott,“Theory Ahead of Business Cycle Mea-surement,’’ and Lawrence H. Summers, “Some Skeptical Observations on Real Business CycleTheory.’’ Both are in Quarterly Review, Federal Reserve Bank of Minneapolis (Fall 1986).


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

Critics argue that the evidence does not support short-run monetary neutral-ity. They point out that reductions in money growth and inflation are almostalways associated with periods of high unemployment. Monetary policy appearsto have a strong influence on the real economy.

Advocates of real business cycle theory argue that their critics confuse thedirection of causation between money and output. These advocates claim thatthe money supply is endogenous: fluctuations in output might cause fluctuationsin the money supply. For example, when output rises because of a beneficialtechnology shock, the quantity of money demanded rises. The Federal Reservemay respond by raising the money supply to accommodate the greater demand.This endogenous response of money to economic activity may give the illusionof monetary nonneutrality.5


Testing for Monetary Neutrality

The direction of causation between fluctuations in the money supply and fluctu-ations in output is hard to establish. The only sure way to determine cause andeffect would be to conduct a controlled experiment. Imagine that the Fed set themoney supply according to some random process. Every January, the Fed chair-man would flip a coin. Heads would mean an expansionary monetary policy forthe coming year; tails a contractionary one. After a number of years we wouldknow with confidence the effects of monetary policy. If output and employmentusually rose after the coin came up heads and usually fell after it came up tails,then we would conclude that monetary policy has real effects.Yet if the flip of theFed’s coin were unrelated to subsequent economic performance, then we wouldconclude that real business cycle theorists are right about the neutrality of money.

Unfortunately for scientific progress, but fortunately for the economy, econo-mists are not allowed to conduct such experiments. Instead, we must glean whatwe can from the data that history gives us.

One classic study in the history of monetary policy is the 1963 book by Mil-ton Friedman and Anna Schwartz, A Monetary History of the United States,1867–1960. This book describes the historical events that shaped decisions overmonetary policy and the economic events that resulted from those decisions.Friedman and Schwartz claim, for instance, that the death in 1928 of BenjaminStrong, the president of the New York Federal Reserve Bank, was one cause ofthe Great Depression of the 1930s: Strong’s death left a power vacuum at the Fed,which prevented the Fed from responding vigorously as economic conditionsdeteriorated. In other words, Strong’s death, like the Fed’s coin coming up tails,was a random event leading to more contractionary monetary policy.6

A more recent study by Christina Romer and David Romer follows in thefootsteps of Friedman and Schwartz. The Romers carefully read through the

C A S E S T U D Y

5 Robert G. King and Charles I. Plosser, “Money, Credit, and Prices in a Real Business Cycle,’’American Economic Review 74 ( June 1984): 363–380.6 Milton Friedman and Anna J. Schwartz, A Monetary History of the United States, 1867–1960(Princeton, NJ: Princeton University Press, 1960).


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

minutes of the meetings of the Federal Reserve’s Open Market Committee,which sets monetary policy. From these minutes, they identified dates when theFed appears to have shifted its policy toward reducing the rate of inflation. TheRomers argue that these dates are, in essence, the equivalent of the Fed’s coincoming up tails.They then show that the economy experienced a decline in out-put and employment after each of these dates. Thus, the Romers’ evidenceappears to establish the short-run nonneutrality of money.7

Interpretations of history, however, are always open to dispute. No one can besure what would have happened during the 1930s had Benjamin Strong lived.Similarly, not everyone is convinced that the Romers’ dates are as exogenous as acoin’s flip: perhaps the Fed was actually responding to events that would havecaused declining output and employment even without Fed action.Thus, althoughmost economists are convinced that monetary policy has an important role in thebusiness cycle, this judgment is based on the accumulation of evidence from manystudies.There is no “smoking gun” that convinces absolutely everyone.

The Flexibility of Wages and PricesReal business cycle theory assumes that wages and prices adjust quickly to clearmarkets, just as Crusoe always achieves his optimal level of GDP without anyimpediment from a market imperfection. Advocates of this theory believe thatthe market imperfection of sticky wages and prices is not important for under-standing economic fluctuations.They also believe that the assumption of flexibleprices is superior methodologically to the assumption of sticky prices, because itties macroeconomic theory more closely to microeconomic theory.

Critics point out that many wages and prices are not flexible. They believethat this inflexibility explains both the existence of unemployment and the non-neutrality of money. To explain why prices are sticky, they rely on the variousnew Keynesian theories that we discuss in the next section.8

19-2 New Keynesian Economics

Most economists are skeptical of the theory of real business cycles and believethat short-run fluctuations in output and employment represent deviations fromthe natural levels of these variables. They think these deviations occur becausewages and prices are slow to adjust to changing economic conditions. As we


7 Christina Romer and David Romer, “Does Monetary Policy Matter? A New Test in the Spiritof Friedman and Schwartz,” NBER Macroeconomics Annual (1989): 121–170.8 To read more about real business cycle theory, see N. Gregory Mankiw,“Real Business Cycles:A New Keynesian Perspective,’’ Journal of Economic Perspectives 3 (Summer 1989): 79–90; BennettT. McCallum,“Real Business Cycle Models,’’ in R. Barro, ed., Modern Business Cycle Theory (Cam-bridge, MA: Harvard University Press, 1989), 16–50; and Charles I. Plosser,“Understanding RealBusiness Cycles,’’ Journal of Economic Perspectives 3 (Summer 1989): 51–77.


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

discussed in Chapters 9 and 13, this stickiness makes the short-run aggregatesupply curve upward sloping rather than vertical. As a result, fluctuations inaggregate demand cause short-run fluctuations in output and employment.

But why exactly are prices sticky? New Keynesian research has attempted toanswer this question by examining the microeconomics behind short-run priceadjustment. By doing so, it tries to put the traditional theories of short-run fluc-tuations on a firmer theoretical foundation.

Small Menu Costs and Aggregate-DemandExternalitiesOne reason prices do not adjust immediately in the short run is that there arecosts to price adjustment.To change its prices, a firm may need to send out newcatalogs to customers, distribute new price lists to its sales staff, or, in the case ofa restaurant, print new menus. These costs of price adjustment, called menucosts, lead firms to adjust prices intermittently rather than continuously.

Economists disagree about whether menu costs explain the short-run stickinessof prices. Skeptics point out that menu costs are usually very small. How can smallmenu costs help to explain recessions,which are very costly for society? Proponentsreply that small does not mean inconsequential: even though menu costs are smallfor the individual firm, they can have large effects on the economy as a whole.

According to proponents of the menu-cost hypothesis, to understand whyprices adjust slowly, we must acknowledge that there are externalities to priceadjustment: a price reduction by one firm benefits other firms in the economy.When a firm lowers the price it charges, it slightly lowers the average price leveland thereby raises real money balances. The increase in real money balancesexpands aggregate income (by shifting the LM curve outward). The economicexpansion in turn raises the demand for the products of all firms. This macro-economic impact of one firm’s price adjustment on the demand for all otherfirms’ products is called an aggregate-demand externality.

In the presence of this aggregate-demand externality, small menu costs canmake prices sticky, and this stickiness can have a large cost to society. Supposethat a firm originally sets its price too high and later must decide whether tocut its price. The firm makes this decision by comparing the benefit of a pricecut—higher sales and profit—to the cost of price adjustment.Yet because of theaggregate-demand externality, the benefit to society of the price cut wouldexceed the benefit to the firm. The firm ignores this externality when makingits decision, so it may decide not to pay the menu cost and cut its price eventhough the price cut is socially desirable. Hence, sticky prices may be optimal forthose setting prices, even though they are undesirable for the economy as a whole.9


9 For more on this topic, see N. Gregory Mankiw,“Small Menu Costs and Large Business Cycles:A Macroeconomic Model of Monopoly,’’ Quarterly Journal of Economics 100 (May 1985): 529–537;George A. Akerlof and Janet L.Yellen,“A Near Rational Model of the Business Cycle,With Wageand Price Inertia,’’ Quarterly Journal of Economics 100 (Supplement 1985): 823–838; and Olivier JeanBlanchard and Nobuhiro Kiyotaki, “Monopolistic Competition and the Effects of AggregateDemand,’’ American Economic Review 77 (September 1987): 647–666.


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate


How Large Are Menu Costs?

When microeconomists discuss a firm’s costs, they usually emphasize the labor,capital, and raw materials that are needed to produce the firm’s output.The costof changing prices is rarely mentioned. For many purposes, this omission is a rea-sonable simplification. Yet the cost of changing prices is not zero, as was estab-lished by a study of price changes in five large supermarket chains.

In this study, a group of economists examined a unique store-level data set todetermine how large menu costs really are.They found that price adjustment “isa complex process, requiring dozens of steps and a nontrivial amount ofresources.” These resources include the labor cost of changing shelf prices, thecosts of printing and delivering new price tags, and the cost of supervising theprocess.The data included detailed measurements of these costs; when necessary,a stopwatch was used to measure the labor input.

The study reported that for a typical store in a supermarket chain, menu costsadd up to $105,887 a year.This amount equals 0.70 percent of a store’s revenue,or 35 percent of net profits. If that total is divided by the number of pricechanges that a store institutes in a given year on all of its products, the result isthat each price change costs $0.52.

A notable finding from this research is that the cost of changing prices dependson the legal environment. One supermarket chain examined in the study was oper-ating in a state with an “item-pricing law,” which required that a separate price tagbe placed on each individual item sold (in addition to the price tag on the shelf ).The law raised the estimated cost of a price change from $0.52 to $1.33. As onewould expect, the higher cost of changing prices reduced the frequency of priceadjustment: the supermarket chain operating under the item-pricing law changed6.3 percent of product prices per week, compared to 15.6 percent for other chains.

These findings apply to only a single industry, so one should be cautious aboutextrapolating the results to the entire economy. Nonetheless, the authors of thestudy conclude that “the magnitude of the menu costs we find is large enoughto be capable of having macroeconomic significance.”10

C A S E S T U D Y

10 Daniel Levy,Mark Bergen, Shantanu Dutta, and Robert Venable,“The Magnitude of Menu Costs:Direct Evidence From Large Supermarket Chains,” Quarterly Journal of Economics 112 (August 1997):791–825. All dollar figures in this case study are expressed in 1991 dollars, because that is the yearwhen the data were collected. If expressed in 2005 dollars, they would be about 40 percent larger.

Recessions as Coordination FailureSome new Keynesian economists suggest that recessions result from a failure ofcoordination among economic decisionmakers. In recessions, output is low,workers are unemployed, and factories sit idle. It is possible to imagine alloca-tions of resources in which everyone is better off: the high output and employ-ment of the 1920s, for example, were clearly preferable to the low output andemployment of the 1930s. If society fails to reach an outcome that is feasible and


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

that everyone prefers, then the members of society have failed to coordinate theirbehavior in some way.

Coordination problems can arise in the setting of wages and prices becausethose who set them must anticipate the actions of other wage and price set-ters. Union leaders negotiating wages are concerned about the concessionsother unions will win. Firms setting prices are mindful of the prices otherfirms will charge.

To see how a recession could arise as a failure of coordination, consider thefollowing parable. The economy is made up of two firms. After a fall in themoney supply, each firm must decide whether to cut its price, based on its goalof maximizing profit. Each firm’s profit, however, depends not only on its pric-ing decision but also on the decision made by the other firm.

The choices facing each firm are listed in Figure 19-2, which shows how theprofits of the two firms depend on their actions. If neither firm cuts its price, realmoney balances are low, a recession ensues, and each firm makes a profit of only$15. If both firms cut their prices, real money balances are high, a recession isavoided, and each firm makes a profit of $30. Although both firms prefer to avoida recession, neither can do so by its own actions. If one firm cuts its price andthe other does not, a recession follows. The firm making the price cut makesonly $5, while the other firm makes $15.

The essence of this parable is that each firm’s decision influences the set ofoutcomes available to the other firm. When one firm cuts its price, it improvesthe position of the other firm, because the other firm can then act to avoid therecession. This positive impact of one firm’s price cut on the other firm’s profitopportunities might arise from an aggregate-demand externality.

What outcome should we expect in this economy? On the one hand, if eachfirm expects the other to cut its price, both will cut prices, resulting in the pre-ferred outcome in which each makes $30. On the other hand, if each firmexpects the other to maintain its price, both will maintain their prices, resultingin the inferior outcome in which each makes $15. Either of these outcomes ispossible: economists say that there are multiple equilibria.


figure 19-2

Firm 1 makes $30Firm 2 makes $30



Firm 1

Firm 2

CutPrice

Cut Price Keep High Price

KeepHighPrice


Price Setting and CoordinationFailure This figure shows a hypothetical “game’’ between twofirms, each of which is decidingwhether to cut prices after a fall in the money supply. Each firmmust choose a strategy withoutknowing the strategy the other firmwill choose. What outcome wouldyou expect?


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

The inferior outcome, in which each firm makes $15, is an example of acoordination failure. If the two firms could coordinate, they would both cuttheir price and reach the preferred outcome. In the real world, unlike in ourparable, coordination is often difficult because the number of firms setting pricesis large. The moral of the story is that prices can be sticky simply because people expectthem to be sticky, even though stickiness is in no one’s interest.11

The Staggering of Wages and PricesNot everyone in the economy sets new wages and prices at the same time.Instead, the adjustment of wages and prices throughout the economy is stag-gered. Staggering slows the process of coordination and price adjustment. In par-ticular, staggering makes the overall level of wages and prices adjust gradually, even whenindividual wages and prices change frequently.

Consider the following example. Suppose first that price setting is synchro-nized: every firm adjusts its price on the first day of every month. If the moneysupply and aggregate demand rise on May 10, output will be higher from May10 to June 1 because prices are fixed during this interval. But on June 1 all firmswill raise their prices in response to the higher demand, ending the boom.

Now suppose that price setting is staggered: half the firms set prices on thefirst of each month and half on the fifteenth. If the money supply rises on May10, then half the firms can raise their prices on May 15. But these firms willprobably not raise their prices very much. Because half the firms will not bechanging their prices on the fifteenth, a price increase by any firm will raise thatfirm’s relative price, causing it to lose customers. (By contrast, if all firms are syn-chronized, all firms can raise prices together, leaving relative prices unaffected.)If the May 15 price setters make little adjustment in their prices, then the otherfirms will make little adjustment when their turn comes on June 1, because theyalso want to avoid relative price changes. And so on. The price level rises slow-ly as the result of small price increases on the first and the fifteenth of eachmonth. Hence, staggering makes the overall price level adjust sluggishly, becauseno firm wishes to be the first to post a substantial price increase.

Staggering also affects wage determination. Consider, for example, how a fallin the money supply works its way through the economy. A smaller money sup-ply reduces aggregate demand, which in turn requires a proportionate fall innominal wages to maintain full employment. Each worker might be willing totake a lower nominal wage if all other wages were to fall proportionately. Buteach worker is reluctant to be the first to take a pay cut, knowing that this means,at least temporarily, a fall in his or her relative wage. Because the setting of wagesis staggered, the reluctance of each worker to reduce his or her wage first makes


11 For more on coordination failure, see Russell Cooper and Andrew John,“Coordinating Coor-dination Failures in Keynesian Models,’’ Quarterly Journal of Economics 103 (1988): 441– 463; andLaurence Ball and David Romer,“Sticky Prices as Coordination Failure,’’ American Economic Review81 ( June 1991): 539–552.


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

the overall level of wages slow to respond to changes in aggregate demand. Inother words, the staggered setting of individual wages makes the overall level ofwages sticky.12


12 For more on the effects of staggering, see John Taylor, “Staggered Price Setting in a MacroModel,’’ American Economic Review 69 (May 1979): 108–113; and Olivier J. Blanchard,“Price Asyn-chronization and Price Level Inertia,’’ in R. Dornbusch and Mario Henrique Simonsen, eds., Infla-tion, Debt, and Indexation (Cambridge, MA: MIT Press, 1983), 3–24.

This table is based on answers to the question:How often do the prices of your most importantproducts change in a typical year?

Frequency Percentage of Firms

Less than once 10.2Once 39.31.01 to 2 15.62.01 to 4 12.94.01 to 12 7.512.01 to 52 4.352.01 to 365 8.6More than 365 1.6

Source: Table 4.1, Alan S. Blinder, “On Sticky Prices:Academic Theories Meet the Real World,’’ in N. G.Mankiw, ed., Monetary Policy (Chicago: University ofChicago Press, 1994), 117–154.

The Frequency of Price Adjustment

table 19-1

If You Want to Know Why Firms Have Sticky Prices, Ask Them

How sticky are prices, and why are they sticky? As we have seen, these questionsare at the heart of new Keynesian theories of short-run economic fluctuations(as well as of the traditional model of aggregate demand and aggregate supply).In an intriguing study, economist Alan Blinder attacked these questions directlyby surveying firms about their price adjustment decisions.

Blinder began by asking firm managers how often they change prices. Theanswers, summarized in Table 19-1, yielded two conclusions. First, stickyprices are quite common. The typical firm in the economy adjusts its pricesonce or twice a year. Second, there are large differences among firms in thefrequency of price adjustment. About 10 percent of firms change prices moreoften than once a week, and about the same number change prices less oftenthan once a year.

C A S E S T U D Y


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

Blinder then asked the firm managers why they don’t change prices moreoften. In particular, he explained to the managers 12 economic theories of stickyprices and asked them to judge how well each of these theories describe theirfirms. Table 19-2 summarizes the theories and ranks them by the percentage ofmanagers who accepted the theory. Notice that each of the theories wasendorsed by some of the managers, and each was rejected by a large number aswell. One interpretation is that different theories apply to different firms,depending on industry characteristics, and that price stickiness is a macroeco-nomic phenomenon without a single microeconomic explanation.


Theory and Percentage of ManagersBrief Description Who Accepted Theory

Coordination failure: 60.6Firms hold back on price changes, waiting for others to go first

Cost-based pricing with lags: 55.5Price rises are delayed until costs rise

Delivery lags, service, etc.: 54.8Firms prefer to vary other product attributes, such as delivery lags, service, or product quality

Implicit contracts: 50.4Firms tacitly agree to stabilize prices, perhaps out of “fairness” to customers

Nominal contracts: 35.7Prices are fixed by explicit contracts

Costs of price adjustment: 30.0Firms incur costs by changing prices

Procyclical elasticity: 29.7Demand curves become less elastic as they shift in

Pricing points: 24.0Certain prices (e.g., $9.99) have special psychological significance

Inventories: 20.9Firms vary inventory stocks instead of prices

Constant marginal cost: 19.7Marginal cost is flat and markups are constant

Hierarchical delays: 13.6Bureaucratic delays slow down decisions

Judging quality by price: 10.0Firms fear customers will mistake price cuts for reductions in quality

Source: Tables 4.3 and 4.4, Alan S. Blinder, “On Sticky Prices: Academic Theories Meet the Real World,’’ in N. G. Mankiw,ed., Monetary Policy (Chicago: University of Chicago Press, 1994), 117–154.

Theories of Price Stickiness

table 19-2


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

Among the dozen theories, coordination failure tops the list. According toBlinder, this is an important finding, because it suggests that the theory ofcoordination failure explains price stickiness, which in turn explains why theeconomy experiences short-run fluctuations around its natural rate. He writes,“The most obvious policy implication of the model is that more coordinatedwage and price setting—somehow achieved—could improve welfare. But ifthis proves difficult or impossible, the door is opened to activist monetary pol-icy to cure recessions.”13

19-3 Conclusion

Recent developments in the theory of short-run economic fluctuations remindus that we do not understand economic fluctuations as well as we would like.Fundamental questions about the economy remain open to dispute. Is the stick-iness of wages and prices a key to understanding economic fluctuations? Doesmonetary policy have real effects?

The way economists answer these questions affects how they view the role ofeconomic policy. Economists who believe that wages and prices are sticky, suchas those pursuing new Keynesian theories, often believe that monetary and fiscalpolicy should be used to try to stabilize the economy. Price stickiness is a typeof market imperfection, and it leaves open the possibility that government poli-cies can raise economic well-being for society as a whole.

By contrast, real business cycle theory suggests that the government’s influenceon the economy is limited and that even if the government could stabilize theeconomy, it should not try to do so. According to this theory, the ups and downsof the business cycle are the natural and efficient response of the economy tochanging technological possibilities.The standard real business cycle model doesnot include any type of market imperfection. In this model, the “invisible hand”of the marketplace guides the economy to an optimal allocation of resources.

To evaluate alternative views of the economy, research economists bring tobear a wide variety of evidence, as we have seen in this chapter’s five case stud-ies.They have used micro data to study intertemporal substitution, macro data toexamine the cyclical behavior of technology, the minutes of Fed meetings to testmonetary neutrality, business studies to measure the magnitude of menu costs,and surveys to judge theories of price stickiness. Economists differ in whichpieces of evidence they find most convincing, and so the theory of economicfluctuations remains a source of debate.

Although this chapter has divided recent research into two distinct camps, not alleconomists fall entirely into one camp or the other. Over time, more economists


13 To read more about this study, see Alan S. Blinder,“On Sticky Prices:Academic Theories Meet theReal World,’’ in N. G. Mankiw, ed., Monetary Policy (Chicago: University of Chicago Press, 1994),117–154; or Alan S. Blinder, Elie R.D. Canetti, David E. Lebow, and Jeremy E. Rudd, Asking AboutPrices:A New Approach to Understanding Price Stickiness, (New York: Russell Sage Foundation, 1998).


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

have been trying to incorporate the strengths of both approaches into their research.Real business cycle theory emphasizes intertemporal optimization and forward-looking behavior, whereas new Keynesian theory stresses the importance of stickyprices and other market imperfections. Increasingly, theories at the research frontiermeld many of these elements to advance our understanding of economic fluctua-tions. It is this kind of work that makes macroeconomics an exciting field of study.14

Summary

1. The theory of real business cycles is an explanation of short-run economic fluc-tuations built on the assumptions of the classical model, including the classicaldichotomy and the flexibility of wages and prices.According to this theory, eco-nomic fluctuations are the natural and efficient response of the economy tochanging economic circumstances, especially changes in technology.

2. Advocates and critics of real business cycle theory disagree about whetheremployment fluctuations represent intertemporal substitution of labor,whether technology shocks cause most economic fluctuations, whethermonetary policy affects real variables, and whether the short-run stickiness ofwages and prices is important for understanding economic fluctuations.

3. New Keynesian research on short-run economic fluctuations builds on thetraditional model of aggregate demand and aggregate supply and tries to pro-vide a better explanation of why wages and prices are sticky in the short run.One new Keynesian theory suggests that even small costs of price adjustmentcan have large macroeconomic effects because of aggregate-demand exter-nalities. Another theory suggests that recessions occur as a type of coordina-tion failure. A third theory suggests that staggering in price adjustment makesthe overall price level sluggish in response to changing economic conditions.


14 For some research that brings together the different approaches, see Marvin Goodfriend andRobert King,“The New Neoclassical Synthesis and the Role of Monetary Policy,” NBER Macro-economics Annual (1997): 231–283; Julio Rotemberg and Michael Woodford,“An Optimization-BasedEconometric Framework for the Evaluation of Monetary Policy,” NBER Macroeconomics Annual(1997): 297–346; Richard Clarida, Jordi Gali, and Mark Gertler,“The Science of Monetary Policy:A New Keynesian Perspective,” Journal of Economic Literature 37 (December 1999): 1661–1707.Thesepapers examine models in which both forward-looking optimizing behavior and sticky prices playcentral roles in explaining the business cycle and the short-run effects of monetary policy.

K E Y C O N C E P T S

Real business cycle theory

New Keynesian economics

Intertemporal substitution of labor

Solow residual

Labor hoarding

Menu costs

Aggregate-demand externality

Coordination failure


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate


1. According to real business cycle theory, perma-nent and transitory shocks to technology shouldhave very different effects on the economy. Usethe parable of Robinson Crusoe to compare theeffects of a transitory shock (good weatherexpected to last only a few days) and a permanentshock (a beneficial change in weather patterns).Which shock would have a greater effect on Cru-soe’s work effort? On GDP? Is it possible that oneof these shocks might reduce work effort?

2. Suppose that prices are fully flexible and that theoutput of the economy fluctuates because ofshocks to technology, as real business cycle theo-ry claims.

a. If the Federal Reserve holds the money supplyconstant, what will happen to the price level asoutput fluctuates?

b. If the Federal Reserve adjusts the money sup-ply to stabilize the price level, what will hap-pen to the money supply as output fluctuates?

c. Many economists have observed that fluctua-tions in the money supply are positively corre-lated with fluctuations in output. Is thisevidence against real business cycle theory?

3. Coordination failure is an idea with many appli-cations. Here is one:Andy and Ben are running abusiness together. If both work hard, the businessis a success, and they each earn $100 in profit. Ifone of them fails to work hard, the business is lesssuccessful, and they each earn $70. If neitherworks hard, the business is even less successful,and they each earn $60 in profit. Working hardtakes $20 worth of effort.

1. How does real business cycle theory explain fluc-tuations in employment?

2. What are the four central disagreements in thedebate over real business cycle theory?

Q U E S T I O N S F O R R E V I E W

P R O B L E M S A N D A P P L I C A T I O N S

3. How does the staggering of price adjustments byindividual firms affect the adjustment of the over-all price level to a monetary contraction?

4. According to surveys, how often does the typicalfirm change its prices? How do firm managersexplain the stickiness of their prices?

a. Set up this “game” as in Figure 19-2.

b. What outcome would Andy and Ben prefer?

c. What outcome would occur if each expectedhis partner to work hard?

d. What outcome would occur if each expectedhis partner to be lazy?

e. Is this a good description of the relationshipamong partners? Why or why not?

4. (This problem uses basic microeconomics.) Thechapter discusses the price-adjustment decisionsof firms with menu costs. This problem asks youto consider that issue more analytically in thesimple case of a single firm.

a. Draw a diagram describing a monopoly firm,including a downward-sloping demand curveand a cost curve. (For simplicity, assume thatmarginal cost is constant, so the cost curve is ahorizontal line.) Show the profit-maximizingprice and quantity. Show the areas that representprofit and consumer surplus at this optimum.

b. Now suppose the firm has previouslyannounced a price slightly above the opti-mum. Show this price and the quantity sold.Show the area representing the lost profit fromthe excessive price. Show the area representingthe lost consumer surplus.

c. The firm decides whether to cut its price bycomparing the extra profit from a lower priceto the menu cost. In making this decision, whatexternality is the firm ignoring? In what senseis the firm’s price-adjustment decision ineffi-cient from the standpoint of society as a whole?


© Wor

th Pub

lishe

rs, D

o Not

Duplic

ate

advances in business cycle theory - …bcs.worthpublishers.com/webpub/economics/mankiw7/6th...

Documents