# Long-Wave Rhythms in Economic Development and Political Behavior.

By Brian J. L. Berry. Baltimore and London: The John Hopkins Press,
1991. Pp. xiv, 242.

The value of this book depends heavily on the interests of the reader. The book can be approached in several different ways: a summary of long-wave theories; as an application of chaos theory to economics; as a merging of long-wave theories; as a new attempt to interrelate economic, social and political events; or as a challenge to conventional economic wisdom. The interplay and skilled weaving of these various themes along with Professor Berry's clear and engaging writing style make this book eminently readable. Despite these strengths in style, the method and arguments fail to convince this reader of the validity of his long-wave theory of economic and political behavior.

The first chapter opens with the question of the appropriate tools to be used in analyzing and presenting evidence for long-waves in economic variables. Chaos theory is the main tool employed in the book to analyze data for long wave propensities. Chaos theory contends that the relationships between variables can change as the variables themselves change, resulting in seemingly random, or chaotic, behavior by the variables. Chaos theory does not rely on new mathematical techniques, but instead is based on analysis of data patterns developed using conventional techniques. The seemingly random data of a chaotic system will often reveal itself by operating around a strange attractor, that is a point of collection of points around which the variables will oscillate. In this exposition on long-wave theory, as with previous long wave theory presentations, the argument relies on the use of smoothed data to show the long term trends. Professor Berry notes Slutzky's argument that a moving average applied to random numbers can create cyclical fluctuations where none existed. In reply, the author argues that by using chaos theory the critique is no longer valid because ". . . a chaotic system lies somewhere between one that is perfectly periodic and one that is a random walk" [p. 14]. While this conclusion could be valid, the data set he employs to analyze by long-wave economic activity is not appropriate to chaos theory. Professor Berry chooses to use moving averages to "smooth out the year-to-year oscillations" [p. 12], or eliminate the behavior in the annual data he cannot explain. This elimination of "year-to-year oscillations" eliminates seemingly random variations that chaos theory attempts to explain.

A second difficulty in the application of chaos theory is the use of a "strange attractor". A strange attractor is a point or a collection of points around which a chaotic variable will oscillate. The strange attractor is not obvious in the raw data, but can be revealed in the graphical presentation of the data. Berry plots a strange attractor for both the price and output changes by plotting this year's change against last year's, resulting in the strange attractor of a positive relationship. This positive relationship is not the kind of hidden pattern in the seemingly random data that a strange attractor is suppose to reveal. The best direction prediction for the smoothed change in this year's prices or output is that will be the same as last year's. The graphing of the strange attractor of a positive relationship between lagged values provides no new insights into long-wave activity. The difficulties in the use of chaos theory, though, do not invalidate the body of his presentation because Professor Berry's use of chaos theory fails to enhance or clarify his long-wave theory. The only role of chaos theory is to avoid Slutzky's critique, which it fails to do.

A second disproof of the Slutzky technique presented in the book is the comparison of four year moving averages to four year moving averages computed every fifth year. The use of the four year moving averages computed every fifth year is to decouple the averages, and show that the trends are from the data rather than the technique. Berry argues that the overlaid graphs show ". . . quite clearly that the decoupled averages track very closely the four year moving averages. . . . I therefore conclude that nothing is being propagated by the smoothing technique; the rhythms are a quality of the data, not of how the data is being handled [p. 18; author's italics]." The evidence would not seem to warrant this conclusion. No goodness of fit analysis is undertaken to statistically determine the similarity of the two curves or data sets. The visual similarity of the two graphs may be the product of the scale of the graphs. Two hundred years and 28% range in growth change are shown on a graph that is a 4 1/2 inches by 3 inches. The reduced size can make large variations look insignificant in absolute distance. Finally, the graphs can be interpreted to suggest the opposite of the conclusion. The four-year averages computed every fifth year changes in 15 times during the two hundred years, while the four-year moving average changes sign 23 times, a 53% difference in the number of sign changes. The evidence does not support a rejection of the Slutzky critique.

Ignoring the method issues, the long-wave arguments are laid out systematically. After presenting the evidence for long waves in wholesale prices at the end of Chapter 1, in Chapter 2 he reviews and integrates the various long wave theories that followed Kondratiev. In Chapters 3 and 4 he does the same for Kuznets long wave cycles of economic growth, and argues for a double peak output cycle. In Chapter 5 he brings together the double peak output cycle with the Kondratiev price cycle, presenting a summary cartesian model showing the path of changes in both output and prices. The path travels through all four quadrants, making deflationary growth and stagflation crises part of the normal path of the long wave cycle. The "long-wave clock" that results from the cartesian map is then coordinated with stock market cycles (Chapter 7), "critical" elections (Chapter 8), war and world leadership cycles (Chapter 9), and finally climate cycles (Chapter 10). Chapter 11 places the contemporary economy in the context of the long wave clock. The final chapter looks at the ability of modern economics to deal with the crisis inherent in the long-wave process, and concludes optimistically that institutional innovations have reduced and can reduce some of the oscillations of the long wave clock.

The value of this book depends heavily on the interests of the reader. The book can be approached in several different ways: a summary of long-wave theories; as an application of chaos theory to economics; as a merging of long-wave theories; as a new attempt to interrelate economic, social and political events; or as a challenge to conventional economic wisdom. The interplay and skilled weaving of these various themes along with Professor Berry's clear and engaging writing style make this book eminently readable. Despite these strengths in style, the method and arguments fail to convince this reader of the validity of his long-wave theory of economic and political behavior.

The first chapter opens with the question of the appropriate tools to be used in analyzing and presenting evidence for long-waves in economic variables. Chaos theory is the main tool employed in the book to analyze data for long wave propensities. Chaos theory contends that the relationships between variables can change as the variables themselves change, resulting in seemingly random, or chaotic, behavior by the variables. Chaos theory does not rely on new mathematical techniques, but instead is based on analysis of data patterns developed using conventional techniques. The seemingly random data of a chaotic system will often reveal itself by operating around a strange attractor, that is a point of collection of points around which the variables will oscillate. In this exposition on long-wave theory, as with previous long wave theory presentations, the argument relies on the use of smoothed data to show the long term trends. Professor Berry notes Slutzky's argument that a moving average applied to random numbers can create cyclical fluctuations where none existed. In reply, the author argues that by using chaos theory the critique is no longer valid because ". . . a chaotic system lies somewhere between one that is perfectly periodic and one that is a random walk" [p. 14]. While this conclusion could be valid, the data set he employs to analyze by long-wave economic activity is not appropriate to chaos theory. Professor Berry chooses to use moving averages to "smooth out the year-to-year oscillations" [p. 12], or eliminate the behavior in the annual data he cannot explain. This elimination of "year-to-year oscillations" eliminates seemingly random variations that chaos theory attempts to explain.

A second difficulty in the application of chaos theory is the use of a "strange attractor". A strange attractor is a point or a collection of points around which a chaotic variable will oscillate. The strange attractor is not obvious in the raw data, but can be revealed in the graphical presentation of the data. Berry plots a strange attractor for both the price and output changes by plotting this year's change against last year's, resulting in the strange attractor of a positive relationship. This positive relationship is not the kind of hidden pattern in the seemingly random data that a strange attractor is suppose to reveal. The best direction prediction for the smoothed change in this year's prices or output is that will be the same as last year's. The graphing of the strange attractor of a positive relationship between lagged values provides no new insights into long-wave activity. The difficulties in the use of chaos theory, though, do not invalidate the body of his presentation because Professor Berry's use of chaos theory fails to enhance or clarify his long-wave theory. The only role of chaos theory is to avoid Slutzky's critique, which it fails to do.

A second disproof of the Slutzky technique presented in the book is the comparison of four year moving averages to four year moving averages computed every fifth year. The use of the four year moving averages computed every fifth year is to decouple the averages, and show that the trends are from the data rather than the technique. Berry argues that the overlaid graphs show ". . . quite clearly that the decoupled averages track very closely the four year moving averages. . . . I therefore conclude that nothing is being propagated by the smoothing technique; the rhythms are a quality of the data, not of how the data is being handled [p. 18; author's italics]." The evidence would not seem to warrant this conclusion. No goodness of fit analysis is undertaken to statistically determine the similarity of the two curves or data sets. The visual similarity of the two graphs may be the product of the scale of the graphs. Two hundred years and 28% range in growth change are shown on a graph that is a 4 1/2 inches by 3 inches. The reduced size can make large variations look insignificant in absolute distance. Finally, the graphs can be interpreted to suggest the opposite of the conclusion. The four-year averages computed every fifth year changes in 15 times during the two hundred years, while the four-year moving average changes sign 23 times, a 53% difference in the number of sign changes. The evidence does not support a rejection of the Slutzky critique.

Ignoring the method issues, the long-wave arguments are laid out systematically. After presenting the evidence for long waves in wholesale prices at the end of Chapter 1, in Chapter 2 he reviews and integrates the various long wave theories that followed Kondratiev. In Chapters 3 and 4 he does the same for Kuznets long wave cycles of economic growth, and argues for a double peak output cycle. In Chapter 5 he brings together the double peak output cycle with the Kondratiev price cycle, presenting a summary cartesian model showing the path of changes in both output and prices. The path travels through all four quadrants, making deflationary growth and stagflation crises part of the normal path of the long wave cycle. The "long-wave clock" that results from the cartesian map is then coordinated with stock market cycles (Chapter 7), "critical" elections (Chapter 8), war and world leadership cycles (Chapter 9), and finally climate cycles (Chapter 10). Chapter 11 places the contemporary economy in the context of the long wave clock. The final chapter looks at the ability of modern economics to deal with the crisis inherent in the long-wave process, and concludes optimistically that institutional innovations have reduced and can reduce some of the oscillations of the long wave clock.

Printer friendly Cite/link Email Feedback | |

Author: | Schaniel, William C. |
---|---|

Publication: | Southern Economic Journal |

Article Type: | Book Review |

Date: | Apr 1, 1992 |

Words: | 1052 |

Previous Article: | Evolving Technology and Market Structure: Studies in Schumpeterian Economics. |

Next Article: | Competitive Advantage on the Shop Floor. |

Topics: |