The Business Cycle
GROWTH AND CRISIS UNDER
CAPITALISM
CHAPTER 1
The Waste of the Business Cycle
IN THE GREAT DEPRESSION of the 1930s, millions of people were involuntarily unemployed. The unemployed did not have enough money to buy the food, clothing, and shelter that they so badly needed. To the degree that this human misery repeats itself—to a much lesser extent— in the contraction phase of every business cycle, there is a major social problem arising from a seemingly irrational economic situation.
All capitalist economies suffer from business cycles. A business cycle may be defined as an expansion in economic activity (measured by such indicators as output, employment, and profits) followed by a contraction in economic activity (including declining production, massive unemployment, and business losses and bankruptcies). It has no regular periodicity, but the same sequence of economic events does take place time after time. Each cycle is different, but there are many regularities or similar sequences found in every business cycle. Although alleged long-cycle and long-run trends are discussed in this book, the focus is on the shortrun cyclical behavior of cycles ranging from two to ten years.
Since the business cycle includes a period of expansion, most economists of the neoclassical school emphasize the sunny side of the picture— that growth does occur through the business cycle. Neoclassical economists see recessions or depressions as merely a momentary, temporary problem. They view the contraction phase of the cycle as a necessary evil, which resolves some problems of the system, but opens the way to new and more vigorous growth. They contend that the present system is the only possible efficient system leading to growth, so cyclical downturns are a small price to pay. Moreover, neoclassical economists believe that people choose to be unemployed—that there is almost no involuntary unemployment.
Contrary to the neoclassical view, one hypothesis of this book is that the waste and misery of business contractions are not necessary in all efficient economic systems and could be totally eliminated in a more rational economic system. One problem of the business cycle is that it leaves workers, capitalists, and other citizens in a state of uncertainty much of the time. The problem of uncertainty is stressed as a business cycle problem throughout this book, but it must be emphasized that it is also a human problem. The other main problem is that cyclical contractions cause losses to society, business, and individual workers—especially involuntary unemployment. If the hypothesis of this book that capitalism is inherently unstable and generates cyclical unemployment is correct, then all of macroeconomics should be reconstructed around this focus. The dominant neoclassical view of equilibrium, clearing of all markets, and no involuntary unemployment should be replaced by a dynamic, historical, cycle-oriented view.
LOSSES TO SOCIETY
Society suffers many types of losses from the contractions that occur during business cycles. Thousands of factories stand idle, and millions of workers are unemployed, so society loses an enormous amount of potential output for current consumption. Society also loses because very little new plant and equipment are produced, so there is very little, if any, growth of productive potential for future expansion. For that reason, every recession or depression lowers the long-run rate of growth. Although the overall, long-run U.S. trend has been one of economic growth, the rate of growth has been lowered by these losses, according to the findings of this book. Society loses the new inventions that are not discovered because there is less motivation and less money for research and development. Society loses because millions of people are unable to work and to create to the best of their potential. Society loses because millions of people are frustrated and unhappy and the social atmosphere is poisoned.
LOSSES TO BUSINESS
In every contraction, many businesses cannot sell their goods at a profit. The number of bankruptcies skyrockets. The number of new businesses declines drastically. Millions of small businesses are forced out of business, and their owners are often left unemployed. Even a few large corporations go out of business, leaving all of their employees out of a job.
LOSSES TO INDIVIDUAL WORKERS
The greatest scourge of the business cycle, however, is the involuntary unemployment of millions of workers. Every one of these individuals suffers the disruption of a useful life. Heads of families cannot feed their children. The unemployed feel useless; each believes that he or she is a personal failure. There is a calculable increase in mental and physical sickness among the unemployed and their families. Increased unemployment causes increases in alcoholism, divorce, child abuse, crime, and even suicide.
A study for the Joint Economic Committee has documented the grim facts. A sustained 1 percent increase in unemployment is associated with the following statistically significant percentage increases: suicide, 4.1 percent; state mental hospital admissions, 3.4 percent; state prison admissions, 4.0 percent; homicide, 5.7 percent; deaths from cirrhosis of the liver, 1.9 percent; deaths from cardiovascular diseases, 1.9 percent (Brenner 1976, v).
THE CONTROVERSIES AND A FRAMEWORK FOR THEIR ANALYSIS
Chapter 2 presents a measuring system for the empirical description of the cycle, based on the approach of Wesley Mitchell. Chapter 3 argues the institutionalist position that the business cycle is uniquely caused by the institutions of capitalism. After contrasting capitalism with precapitalist institutions, the chapter demonstrates that the history of the cycle changes with the changing stages of capitalism.
No subject in economics is more controversial than unemployment. Chapter 4 explains the dominant view in U.S. economics: that involuntary unemployment does not exist, except for an irreducible frictional, or "natural," level. This view is based on the argument that the capitalist system automatically adjusts demand to supply so that there are only brief deviations from full employment equilibrium, due to factors that are external, or exogenous, to the system. This view is rejected in favor of a mainly endogenous approach. Chapter 4 then considers briefly the main hypothesis of this book: that internal, or endogenous, factors are the main cause of the business cycle of capitalism. Some endogenous theories emphasize the lack of consumer demand; others emphasize the cost of supply, including high wages, interest rates, or raw material prices. Both demand-side and supply-side theories have made major contributions to understanding, but it is shown that each is inadequate by itself. A synthesis is proposed, based on the theories of Marx, Mitchell, Keynes, and Kalecki.
Part Two of the book discusses the more detailed controversies involving the behavior of each important variable and the theories associated with the different aspects of the cycle. Thus, Chapters 5 and 6 discuss consumption and investment behavior, and Chapter 7 explicates the multiplier-accelerator theory based on the behavior of these two variables. Chapter 8 discusses the behavior of income distribution between labor and capital over the cycle, and Chapter 9 contains an exposition of demand-side theories, such as underconsumption, that build on the behavior of income distribution, consumption, and investment.
Chapter 10 details cost behavior of raw materials, plant, and equipment. These data set the stage for Chapter 11, which presents supply side theories based on the cost of capital (overinvestment) or the cost of labor (reserve army theory). Chapter 12 examines how profits and profit rates behave over the cycle, providing the foundation for a new type of profit squeeze (or nutcracker) theory in Chapter 13, which attempts a synthesis of the empirically supported elements of demand-side and supply-side theories.
Part Three of this book adds more complex reality to the theoretical framework. It does this by considering money and credit in Chapter 14, monopoly power in Chapter 15, international relationships in Chapter 16, and governmental behavior in Chapter 17. Some economists would argue for introducing each of these levels into the very first model of the economy, but that would mean an enormously complex model from the very start. If the model involved every important relationship from the start, it would be difficult, or impossible, to understandany of it. Using successive approximations starting from simple models and proceeding to more complex, realistic ones, makes the analysis both clearer and more rigorous. Finally, Part Four considers what changes in institutions and in policies are needed in order to ameliorate or totally eliminate the waste of the business cycle.
CHAPTER 2
Measuring the Business Cycle
THE PIONEER in empirical description of the business cycle was Wesley Clair Mitchell. Indeed, he helped develop many of our present national income accounts. With the help of Arthur Burns (see Burns and Mitchell 1946), he created a method specifically for measuring the business cycle. The method was used in several cycle studies of the National Bureau of Economic Research (NBER), which he founded. Alas, the NBER no longer follows Mitchell's method, but it is still usually called the NBER method.
Mitchell's NBER method depicts the exact path of a single variable over the average business cycle. Mitchell's method, the details of which are presented in this chapter, is still the best method for getting a clear picture of the business cycle. The NBER method may reveal a simple visual relationship of variables, which is helpful in suggesting a hypothesis for testing, but it should be stressed that it does not provide a statistical test of relationships. After the NBER method shows the typical cyclical behavior of a variable, then the next stage of analysis is often the use of econometric regression and correlation analysis to test its relation to other variables.
Before a phenomenon can be measured, it must be carefully defined. Wesley Mitchell presented the most useful definition of the business cycle; it is as follows:
Business cycles are a type of fluctuation found in the aggregate economic activity of nations that organize their work mainly in business enterprises; a cycle consists of expansions occurring at about the same time in many economic activities, followed by similarly general recessions, contractions, and revivals which merge into the expansion phase of the next cycle; this sequence of changes is recurrent but not periodic; in duration business cycles vary from more than one year to ten or twelve years; they are not divisible into shorter cycles of similar character with amplitudes approximately their own.
(Burns and Mitchell 1946, 3)
It is worth examining separately each of the points in Mitchell's definition. First, it is clear that the business cycle is a phenomenon found under capitalism, and not under other systems (as will be seen in the next chapter). Second, the business cycle is not limited to a single firm or industry, but is economywide, expected to show most clearly among aggregate indicators. It is widely diffused and is expected to show in most series. Third, one cycle follows after another; they are marked by regularities and similar sequences of events. Fourth, cycles differ, however, in many ways, including how long they are, so there is no regular periodicity.
Fifth, Mitchell mentioned time periods for a whole business cycle of anywhere from one to twelve years. Some authors have found shorter, mild three- or four-year cycles as well as longer, sharper ten-year cycles. Mitchell does not find that distinction in the evidence, nor does this author. Cycles vary in length from a year to ten years, each repeating roughly the same sequence of events, so that they are qualitatively similar in their pattern and relationships. There are, of course, some very mild cycles and some very severe cycles, but there is no evidence of two or three mild cycles within each severe, longer cycle. Mitchell proved in detail that cycles of shorter duration than those identified by the NBER would show no regular sequence of events, so their alleged patterns would be statistically insignificant.
On the other hand, some authors claim to have found long cycles of fifty to sixty years in length. The first person to argue this view was Kondratief, for whom they are named. Their most famous advocate was Schumpeter (1939). Little evidence was found for their existence, and discussion of them died away in the prosperous 1950s and 1960s. In the difficult times of the 1970s and 1980s, there has been a revival of interest in long cycles (see the sympathetic survey by Kotz [1987]).
Best known of the advocates of long waves in the present revival are Gordon, Weisskopf, and Bowles (1983), who argue: "The U.S. and world capitalist economies are currently in [the] midst of the third long swing crisis of the past century" (p. 152). How do they define long cycles? They admit that long cycles cannot be dated by total output or investment, but they claim long cycles can be dated by changes in the "social structure of accumulation." This concept is a multidimensional political-economic concept of great complexity, so their empirical estimates remain highly controversial.
Mitchell found no evidence of such long cycles, nor has this author in his own research. How could there be much scientific evidence of "long cycles" when even their advocates have discovered at most three of them? In the next chapter it will be shown that—rather than long waves—capitalism has passed through various stages and that the business cycle shows important differences in these stages.
We must distinguish between different types of movements over time. First, in the very long run, there is an evolution of economic systems from one mode of production to another, for example, from ancient Roman slavery to medieval feudalism in Europe. Second, each economic system evolves and goes through various stages, involving considerable changes, but still recognizably the same system. For example, the U.S. economic system was characterized by very small economic units at one time, but is now in a stage characterized by giant corporations with varying degrees of monopoly power. Third, we may identify many long-run trends, such as the increasing percentage of women in the labor force, generally under one stage of one system, but sometimes crossing over several stages. A long-term trend is almost always completely interrupted when there is an evolutionary or revolutionary change from one system to another. Fourth, there are the alleged long cycles. Fifth, there is the business cycle as defined by Mitchell. Sixth, many economic series have seasonal variations, such as higher growth of construction in warmer months. Seventh, there are also erratic movements of each economic variable, not directly connected to any of the above systematic movements. This book concentrates only on the business cycle, but does introduce long-run trends and stages of capitalism when necessary as a background.
DATES OF THE CYCLE
Mitchell's method begins by establishing the trough and peak dates of each cycle, using all available evidence, with heaviest reliance on the main aggregate series. Mitchell's work on dating the cycle was taken over by the NBER and then by the U.S. Department of Commerce, which publishes the dates in the Business Conditions Digest. The quarterly dates since reconversion from World War II are given in Table 2.1.
Table 2.1 reveals that cycle troughs (the lowest point of each cycle) were reached in 1949, 1954, 1958, 1961, 1970, 1975, 1980, and 1982. The most serious of these were in 1975 and 1982. Notice that the quarter of the year in which the trough occurs varies widely, with no pattern.
These dates are used throughout this book as the best available dates of the business cycle. There are many things that could be criticized about these dates (see Sherman 1986). For example, they do not distinguish in any way between a major depression and a minor recession. They simply record each case where aggregate business activity has continuously declined or continuously risen for some lengthy period. The exact criteria used for dating the peaks and troughs are complex, including a number of indicators; the criteria are clearly explained by Burns and Mitchell (1946, ch. 4) and by Moore (1983, ch. 1). These dates are used both because no better series is available and because they are accepted and used by most scholars in the field. The NBER dates for troughs and peaks go all the way back into the nineteenth century; these
earlier dates will be given in the next chapter, which deals with the history of the business cycle.
The quarterly dates are used throughout this book because this is probably the best time unit for cycle analysis. As Burns and Mitchell (1946) point out at great length, data given daily, weekly, or even monthly tend to have too much static; in a different metaphor, they lose the forest and show only the trees. On the other hand, annual data leave out many cyclical turning points and are not sufficiently detailed.
REFERENCE CYCLES VERSUS SPECIFIC CYCLES
The dates given in Table 2.1 determine what the NBER calls a reference cycle. A reference cycle is the average business cycle for all sectors of the U.S. economy. Unless stated otherwise, all empirical analyses of cycles in this book refer to reference cycles.
Each specific economic series, however, has slightly different peaks and troughs from the average cycle. Sometimes it is necessary to look at performance of an economic variable over its own particular cycle dates; this is called a specific cycle. It is used rarely, usually for a variable that differs considerably and systematically from the reference cycle. For example, profit rates almost always lead the reference cycle, that is, they turn down before the peak. Some interest rates usually iag after the reference cycle, that is, they turn down after the peak.
Note that a cycle may be measured from trough to trough or from peak to peak. Because it is the more common procedure, all cycles in this book are measured from trough to trough.
DIVISIONS OF THE CYCLE
Mitchell called the rising period of the business cycle, from the initial trough to the peak, the expansion period. The declining period of the business cycle, from the peak to the final trough, is called the contraction period.
In the business cycle, as defined by Mitchell, there are four phases: two in the expansion period and two in the contraction period. Starting from the low point, or initial trough, of the cycle, there is a rapid upturn, called a recovery (or revival). Next, there is a further expansion, called a prosperity. This is followed by a downturn, called the crisis. Finally, the crisis turns into a contraction, called a depression. Mild depressions are sometimes called recessions, but this book will use Mitchell's term of "depression" to describe the final phase of the cycle.
In a more detailed analysis, Mitchell then divides the cycle into nine stages. The number of stages is arbitrary, but has a logic to it. Stage 1 is the initial trough of the cycle, the low point from which it begins. One could measure peak to peak, but that is not as convenient for illustrating most theories of the business cycle, so this book uses trough-to-trough cycles exclusively. Stage 5 is the cycle peak, where most business activity reaches its highest point. Finally, stage 9 is the final trough, from which a new cycle begins. Stages 1, 5, and 9 are, by definition, just three months or one quarter long.
The expansion period lasts from stage 1 to stage 5. The whole expansion (excluding stages 1 and 5) is then divided up into three equal time periods. The three periods of equal length in expansion are called stages 2, 3, and 4. Thus, if the whole expansion is 15 quarters long (excluding stages 1 and 5), each of the three stages will be five quarters long.
Similarly, the contraction period lasts from stage 5 till stage 9. The whole contraction (excluding stages 5 and 9) is then divided up into three equal time periods. The three periods of equal length in contraction are called stages 6, 7, and 8. Thus, if the whole contraction is six quarters long (excluding stages 5 and 9), then each of the three stages will be two quarters long. Since expansions are normally longer than contractions, stages 2, 3, and 4 are normally longer than stages 6, 7, and 8.
The four phases may now be more precisely defined in terms of the nine stages. Thus, recovery is stages 1 to 3, prosperity is stages 3 to 5, crisis is stages 5 to 7, and depression is stages 7 to 9. The entire expansion period is stages 1 to 5, while the entire contraction period is stages 5 to 9. In other words, recovery is the first phase of expansion (stages 1-3),
while prosperity is the second phase of expansion (stages 3-5). Similarly, crisis is the first phase of contraction (stages 5-7), while depression is the last phase of contraction (stages 7-9). Mitchell considers that the task of business cycle theory is to explain how each phase leads to the next.