Economic fluctuations.

Hall, Robert E.

Economic Fluctuations

Robert E. Hall

The research activities of the NBER's Program on Economic Fluctuations are arranged through a number of small working groups as well as in individual projects. This report surveys the activities of the working groups and highlights several studies presented at the program's major research meetings.

Another important responsibility of the program is the preparation of the Bureau's business cycle chronology, a function carried out by the seven-member Business Cycle Dating Committee.

The 1990 Business Cycle Peak

The Business Cycle Dating Committee conferred twice in the past year. In December 1990, the committee reviewed the evidence on the decline in real activity in the U.S. economy starting during the previous summer. At that time, the committee concluded that the economy probably had reached a peak at some point in the summer, but there was insufficient evidence of a deep enough contraction to enter the peak into the NBER's chronology. In other words, the committee's opinion was that the economy probably was in a recession, but the evidence was not strong enough to make a definitive pronouncement.

Traditionally, the NBER has not made an announcement on a business cycle peak or trough until there was almost no doubt that the date would not be revised in the light of subsequent availability of data. A number of previous episodes have challenged the Bureau's dating process. In 1967, the economy paused dramatically during a period of otherwise strong growth. Many economists considered the period a recession. But the Bureau concluded that the depth of the decline of the economy was considerably less than its standard for a recession.

In 1973 and 1974, the economy stopped growing and real activity remained almost constant. In late 1974, real activity plunged and the economy entered what was then the most severe contraction since the Great Depression. The Bureau placed the peak of the cycle in November 1973, following the principle that once an episode is identified as a recession, the peak is when real activity reached its peak, even if the sharp plunge characteristic of a recession occurs many months later.

The recession of 1980 presented a rather different challenge. Real activity rebounded strongly from the trough in July 1980, but began contracting again in mid-1981. Was 1980 a separate recession, or was it part of the beginning of the severe recession of 1981-2?

The National Bureau of Economic Research is a private, nonprofit research organization founded in 1920 and devoted to objective quantitative analysis of the American economy. Its officers and board of directors are:

Chairman - George T. Conklin, Jr.

Vice Chairman - Paul W. McCracken

Treasurer - Charles A. Walworth

President and Chief Executive Officer - Martin Feldstein

Executive Director - Geoffrey Carliner

Director of Finance and Administration - Sam Parker

DIRECTORS AT LARGE

John Herron Biggs Paul W. McCracken
Andrew Brimmer Leo Melamed
Carl F. Christ Michael H. Moskow

George T. Conklin, Jr. James J. O'Leary

Kathleen B. Cooper Robert T. Parry
Jean A. Crockett Peter G. Peterson
George C. Eads Robert V. Roosa
Morton Ehrlich Richard N. Rosett
Martin Feldstein Bert Seidman
George Hatsopoulos Eli Shapiro
Lawrence R. Klein Donald S. Wasserman

Franklin A. Lindsay

DIRECTORS BY UNIVERSITY APPOINTMENT

Jagdish W. Bhagwati, Columbia William C. Brainard, Yale Glen Cain, Wisconsin Franklin Fisher, Massachusetts Institute of Technology Jonathan R. T. Hughes, Northwestern Saul Hymans, Michigan Marjorie B. McElroy, Duke James L. Pierce, California, Berkeley Andrew Postlewaite, Pennsylvania Nathan Rosenberg, Stanford Harold Shapiro, Princeton Craig Swan, Minnesota Michael Yoshino, Harvard Arnold Zellner, Chicago

DIRECTORS BY APPOINTMENT OF OTHER ORGANIZATIONS

Rueben C. Buse, American Agricultural Economics Association Richard A. Easterlin, Economic History Association Gail Fosler, The Conference Board Ronald Gallant, American Statistical Association Robert S. Hamada, American Finance Association David Kendrick, American Economic Association Ben Laden, National Association of Business Economists Rudolph A. Oswald, American Federation of Labor and Congress

of Industrial Organizations Dean P. Phypers, Committee for Economic Development Douglas D. Purvis, Canadian Economics Association Charles A. Walworth, American Institute of Certified Public

Accountants

The committee concluded that 1980 truly was a separate recession, because real activity reached a level at the next peak, July 1981, above its level of the previous peak, January 1980. The committee has reviewed all three of these decisions well after the fact, and has concluded that the original decisions were correct.

In April, the committee had additional data showing that real activity had contracted sharply at the end of 1990. Earlier doubts about the depth of the contraction were resolved. The monthly series considered most seriously by the committee were total employment, real manufacturing and trade sales, industrial production, and real personal income. All of these series turned down during the summer and showed sufficiently large contractions by early 1991 to meet the committee's criterion for depth. The indicators reached peaks in different months, with employment peaking first and industrial production last. The committee selected July 1990 as the overall peak in real activity.

As this report is being written, there are indications that the economy may have passed the trough of activity and may have entered the recovery phase of the business cycle. Consistent with its earlier practices, the committee will wait until the recovery is fully evident in the data before entering a date for the trough into the chronology.

Research Meetings

The program holds research meetings in February in Palo Alto, in July at the Bureau's Summer Institute in Cambridge, and in October in Cambridge. The papers discussed at research meetings are at an intermediate stage of development, roughly at the point of distribution as NBER Working Papers.

At the research meeting in July 1989, organized by Andrei Shleifer and Lawrence H. Summers, Steven Davis and John C. Haltiwanger presented the first results of a major research project that has attracted a great deal of subsequent attention. Davis and Haltiwanger carry out their research within the Bureau of the Census, where they have access to data on employment by firm. Confidentiality requirements prohibit the distribution of the data to outside researchers. The authors have examined the ways that employment rises and falls at the plant level over the business cycle. They find that the sharp decline in employment that occurs during a recession is the result of large drops in employment at a minority of firms. Many firms continue to expand employment at normal rates during recessions. One implication of their findings is that many more workers move through the labor market during contractions than during booms.

At the same meeting, Lawrence Ausubel, an industrial organization economist, presented a challenging paper on the profitability of the credit card business. He found that competition has not depressed the profit on an incremental credit card customer at a bank; on the contrary, credit card accounts change hands among banks at a considerable premium over the amount owed by the customer. His results are important for macro-economists who need to choose between strictly competitive models of financial markets and those with information limitations and other departures from the competitive model.

At the research meeting in February 1990, Joseph G. Altonji, Fumio Hayashi, and Laurence J. Kotlikoff discussed the association between the consumption levels of related households. According to a famous hypothesis advanced by Robert J. Barro, related families may make transfers between themselves to offset misfortunes in one family or windfalls in another. If so, only the joint income of two related families should matter for the consumption of either family - the family's own income should have no influence beyond that of joint income. Altonji, Hayashi, and Kotlikoff test this implication and reject it strongly.

The program's research meeting in July 1990, organized by Russell Cooper and Steven N. Durlauf, featured papers on complementaries. Research in this area of macroeconomics pursues the idea that positive interactions among firms or other economic units can amplify the effects of external driving forces. Macro-economists starting with Keynes have looked for amplification mechanisms, because the driving forces seem weak in comparison to the strength of the business cycle. A paper by Marianne Baxter and Robert G. King showed that within the type of business cycle model pioneered by Edward Prescott, the introduction of complementarities can resolve some of the failures of earlier models to reproduce the actual behavior of the U.S. economy.

At the research meeting in October 1990, Robert B. Barsky and Elizabeth Warner discussed research on retail pricing over the seasonal cycle. Their work pursues the idea developed by Jeffrey A. Miron and others, including Barsky, that there is a useful analogy between seasonal cycles and business cycles. Many of the same features of the movements of important variables over the business cycle also show in the seasonal cycle. One of these is the lack of cyclical movements in prices, in the presence of large changes in demand over both the seasonal cycle and the business cycle. Barsky and Warner examine the behavior of the prices of about two dozen products that are popular Christmas presents, from the beginning of the shopping season in mid-November until the conclusion of the post-Christmas season in January. They find little tendency for the prices of standardized products, such as consumer electronics, to rise during the period of high demand or to fall in January. The phenomenon of sales is important in their findings: sales are most common on the Friday after Thanksgiving, but remain important throughout the Christmas season. (Barsky and Warner were motivated to examine this issue by an interchange between two participants in the program's research meeting in July 1987.)

In February 1991, Jess Benhabib, Richard Rogerson, and Randall Wright presented an analysis of labor supply with a focus on changes in the level of employment over the business cycle. The traditional view of labor supply sees the worker dividing time between the labor market and nonwork activities, often called leisure. These authors look at the problem in a different way. They picture the individual as deciding between work in the market and work at home. The analysis leads to the conclusion that market work is more likely to be volatile when the alternative activity is work at home rather than leisure.

In the past decade, macroeconomists have given little attention to the important topic of the effect of monetary policy on real economic activity. Many of the models considered at research meetings of the Program on Economic Fluctuations imply that monetary policy has little or no influence on output or employment, while at the same time most of the participants believe that monetary policy in the actual U.S. economy is a major influence on these variables. Two of the papers scheduled for the program's July 1991 meeting, one by King and one by Lawrence J. Christiano and Martin S. Eichenbaum, make up for this neglect.

Research Groups

Much of the activity of the economic fluctuations program occurs in its research groups. A typical group has about 10 members and meets twice a year, once at the Summer Institute and once during the academic year, sometimes in conjunction with a research meeting. A group is created and convened by an organizer or a pair of coorganizers. Groups tend to be focused on relatively narrow research topics of current interest and usually last about two or three years. Some group members are affiliated with the NBER but many others are not. Groups often include advanced graduate students and new faculty members.

Cooper is the organizer of the oldest active group in the program, dating from 1986 and dealing with issues of macroeconomic complementarities. The current research topics in his group include overhead costs and economic fluctuations (Mohamad Hammour), growth and unemployment (Peter Howitt), international currency (Nobuhiro Kiyotaki), and business cycles in a model with multiple steady states (Peter Klenow).

This year Ricardo J. Caballero and Andrew Caplin have created a new group on nonconvexities in macroeconomics. The common idea in this research is that fixed costs and increasing returns can help explain certain macroeconomic phenomena not explained by the standard neoclassical model with constant returns. Caballero, and also Janice Eberly, are studying consumer purchases of durables. The nonconvexity in durables arises because of their "lumpiness." A family cannot buy a fraction of a car when it needs more transportation. Caballero's model, which carefully aggregates individual decisions about the purchase of lumpy durables, shows us why modern models of aggregate consumption fail to account for the time-series properties of durables purchases. Eberly has developed methods for dealing with nonconvexities in studying microdata on durables purchases from individual families. Boyan Jovanovic is studying the Great Depression with the framework of models with nonconvexities in which a low-level equilibrium is possible. Roland Benabou is developing models of urban development - where transportation costs create a nonconvexity - and thus pursuing the analogy between the agglomeration of economic activity in space and its agglomeration in time (the business cycle).

Since 1989, Christiano and Eichenbaum have been in charge of a research group on impulses and propagation mechanisms. As part of his role in the group, Eichenbaum has been working on labor hoarding over the business cycle. John Donaldson has been considering whether aggregate fluctuations can be caused by random shocks hitting individual sectors of the economy. Anton Braun and Charles Evans have been investigating the relationship between seasonal and cyclical activity within real business cycle or equilibrium macro models. Also within this group, King has developed his work on money and the business cycle.

Davis has organized a group on labor market dynamics and aggregate fluctuations. Within the group, Jeremy Greenwood, Rogerson, and Wright are studying housing and durables in a real business cycle model. Harry J. Holzer is analyzing data on job vacancies at the firm level. Tito Boeri, pursuing ideas developed by Davis and Haltiwanger, is using data on employment changes for firms in Germany. Ana Aizcorbe is studying cyclical labor issues, including productivity and labor hoarding, in the U.S. auto business.

Frank Diebold and Durlauf lead a group on common elements of growth and fluctuations. Research in this group pursues the idea that growth and fluctuations involve similar mechanisms. A recession is, in effect, a time when the economy becomes a little less developed. Much of the work in the group focuses on measurement and econometric issues. Danny Quah and Thomas J. Sargent are working on time-series models that study many countries simultaneously. Paul Johnson is studying the time-series implications of the basic Solow growth model. Andrew Bernard and Durlauf have created a statistical framework for testing the idea of convergence - that differences in income per capita among countries tend to disappear over time. Kenneth D. West is studying Japanese monetary policy within a growth-and-cycles framework. Jordi Gali and Hammour are considering the long-run effects of business cycles.

Louis Maccini and Valerie Ramey have just organized a group on inventories. The swing in inventory investment is well known to be the dominant change in the components of GNP during a recession - typically, the decline in GNP from peak to trough is of about the same magnitude as the reduction in inventory investment. Timothy F. Bresnahan and Ramey are looking at plant-level data in the auto industry in order to understand this process better. Durlauf and Maccini are looking at the large amount of unexplained movement in inventories; this noise in inventory investment appears to be an important driving force in the business cycle. West is continuing his research on inventories by looking at the Japanese case. James A. Kahn is using inventory behavior to make inferences about cost conditions. Barsky, Miron, Stephen G. Cecchetti, Anil Kashyap, David Wilcox, Cooper, and Haltiwanger are all studying the various aspects of seasonal movements of the economy as they influence inventory accumulation.

For the past year, Rogerson and Wright have been responsible for a research group on micro and macro perspectives on the aggregate labor market. The group brings together labor economists and macroeconomists. John Kennan is carrying out research on wage setting when there is private information. Kenneth Burdett, Wright, and Eric Smith are considering various aspects of job search models. Davis and Haltiwanger are doing research on the determinants of the cyclical job flows revealed in their earlier research. Gary Hansen and Sargent are studying variations over time in straight time and overtime employment.

Related Activities

The NBER's overall research includes several activities related to the Program on Economic Fluctuations. These include the International Seminar on Macroeconomics, arranged by Robert J. Gordon and Georges de Menil, the Macroeconomics Annual Conference organized by Stanley Fischer and Olivier J. Blanchard, and the Macroeconomic History Workshop under the direction of N. Gregory Mankiw and Christina D. Romer. Separate reports on these activities appear in this and other issues of the NBER Reporter.