Introduction

The Grand Crisis that started in fall of 2008 shocked the confidence in market fundamentalism. Fundamental debates on the nature of economic crisis are going back to the Keynesian revolution during the Great Depression. This book of collected essays will address basic issues from five aspects. Part I is a methodological review on complex evolutionary economics. It includes two review articles. Readers could easily start from there for a basic knowledge in contested issues and competing alternatives. Chapter 2 on equilibrium illusion, economic complexity, and evolutionary foundation is a review talk on competing methodologies in economic analysis, which was a two-hour invited speech at the Japanese Association for Evolutionary Economics in September 2007, on the eve of the Grand Crisis. This chapter serves as a beginner’s guide on major differences between equilibrium and evolutionary perspective, between linear and complex thinking. There is a widespread misperception that neo-classical economics is imitating classical physics in a frictionless world (Mirowski 1989). This is not quite true. Neo-classical economics did borrow the optimization approach from Hamiltonian mechanics. However, some influential equilibrium models, such as the Frisch model of noise-driven cycles, the Lucas model of rational expectations,

and the Coase world of zero transaction costs, simply violate basic laws in physics including the uncertainty principle and the second law in thermodynamics. The role of the FD (first differencing) filter in macro econometrics is similar to that of the geocentric system in astronomy. Mainstream economics felt the fundamental impact of a physics revolution in the 1970s, when discoveries of dissipative structure and deterministic chaos changed the Newtonian paradigm into an evolutionary perspective in physics, chemistry, and biology. The manybody problem is fundamentally different from the one-body and two-body problems in mathematics. That is why methodological individualism and regression analysis is incapable of understanding complex economic dynamics, which is nonlinear and non-integrable in most of cases. This chapter gives a detailed comparison between a linear equilibrium approach and a complex evolutionary approach in micro, macro, finance, and institutional economics, including their policy differences and social outcome in dealing with transition economies and economic crisis. The development from linear Hamiltonian economics to complex evolutionary economics could learn historical lessons from classical mechanics to relativity theory, which is a theoretical progress toward a general theory with a more consistent framework and a wider empirical base (Galbraith 1994). Chapter 3 is a review article on evolutionary economic dynamics, invited by Kurt Dopfer in his Cambridge volume of Evolutionary Foundation of Economics in 2005, a project started in 1995. This chapter is a systematic review on evolutionary economic dynamics for readers with more mathematical background. It mainly discusses two basic issues: the endogenous nature of business cycles and diversity trends in social evolution. There are several significant findings in economic theory: the Frisch model of noise-driven cycles in macro econometrics is a perpetual motion machine; the geometric Brownian motion behind the Black-Scholes model in option pricing is explosive in nature; there is a weak microfoundation but a strong meso foundation for generating business cycles. The 1987 model of the division of labor in Chapter 8 was applied to address Adam Smith’s dilemma in classical economics and the complexity puzzle in theoretical biology. Evolutionary dynamics has no convergence trend, which is contrary to the claim by the optimization approach. The diversified pattern observed in economic organizations can be explained by a trade-off between stability and complexity under environmental constraints. Market resilience and economic complexity can be better understood by nonlinear population dynamics, which is the pertinent feature for biological and social dynamics. Part II on macro vitality was our starting base in studying economic complexity. The main issues in business-cycle theory are trend-cycle separation, economic chaos, and persistent cycles. Chapter 4 on empirical and theoretical evidence of economic chaos in 1988 marked the methodological departure of nonlinear dynamics from econometric analysis. The chapter begins from a brief comparison between linear stochastic and nonlinear deterministic models in a time series analysis. The salient feature

of many macro indexes is their growing trend. How to separate long-term trend and business cycles is a critical issue in business-cycle theory. A random image of business cycles is obtained by the FD (first differencing) filter, i.e., the observed variable is the rate of changes in a time unit. The first empirical evidence of low dimensional strange attractors is found in several empirical monetary aggregates by a log-linear detrending (LLD filter). A continuous-time deterministic model with delayed feedback is proposed to describe the monetary growth. Phase transition from periodic to chaotic motion occurs in the model. A one-dimensional nonlinear delay-differential equation is proposed to describe a nonlinear oscillator with soft-boundaries in target control. The soft-bouncing oscillator model is a generalization of hard boundaries, such as the Goodwin limit cycle model with investment floor and ceiling (1951). Time delay and overshooting are the causes of monetary chaos. Chaotic and multi-periodic regions are resilient under shocks in parameter space. This study is the most comprehensive analysis of economic chaos from empirical to theoretical aspects. Its implication in monetary policy is strong evidence of the Austrian theory of endogenous money, but a challenge to Milton Friedman’s theory of exogenous money. Chapter 5 on searching for economic chaos in 1993 was presented at the Austin Symposium in 1989. Methodological differences in testing economic chaos have induced an intensive debate between physicists and econometricians since 1985. Econometricians routinely used a regression analysis to a discrete time model of white chaos such as a logistic or Henon model, while physicists only found empirical evidence of continuous time deterministic color chaos. This chapter demonstrates basic differences between a discrete time and a continuous time chaos model and pitfalls of statistic techniques in testing strange attractors with fractal dimensionality. The role of the time arrow, time scaling, and long-term correlations in econometric testing and modeling is examined. The fundamental problems of econometric inference, modeling, and forecasting in studies of chaotic economic movements are discussed. Chapter 6 on random walk or color chaos on Wall Street introduced a new method in testing economic chaos by means of time-frequency analysis. The HP (Hodrick-Prescott) filter was suggested by Victor Zarnowicz in Chicago, in addition to FD and LLD filters tested in Chapter 4. The deterministic component from noisy data can be recovered by a time-variant filter in a two-dimensional time-frequency Gabor space. The characteristic frequency is calculated from the Wigner decomposed distribution series. The Wigner-Gabor-Qian (WGQ) spectrogram shows a strong capability in revealing complex cycles from noisy and nonstationary time series. It is found that about 70 percent of HP cyclic fluctuations in Standard & Poor stock price indexes, such as the FSPCOM and FSDXP monthly series can be explained by deterministic color chaos. The characteristic period of persistent cycles is around three to four years. Their correlation dimension is about 2.5. The color-chaos model of stock market movements provides a better alternative in business-cycle theory than a white noise model or Brownian motion. Time-frequency analysis is more powerful than nonlinear dynamics

algorithms in analyzing short and noisy economic time series. The analytic program of WGQ transform was provided by Shie Qian in Austin. Chapter 7 on trends, shocks, and persistent cycles in an evolving economy in 1996 gave a systematic report on business-cycle measurement in time-frequency representation. The term of persistent cycles was suggested by Finn Kydland when he was in Austin. The new method in analyzing business cycles developed in Chapter 6 was systematically applied to a wide range of macro aggregate indexes. Competing detrending methods, including the first differencing (FD) and Hodrick-Prescott (HP) filter, are tested with the mixed case of cycles and noise. The FD filter does not produce a consistent picture of business cycles. The HP filter provides a unified reference in explaining business cycles. Discovery of stable characteristic frequencies from wide ranged economic aggregates provides strong evidence of endogenous cycles and valuable information about structural changes. Economic behavior is more like an organism instead of random walks. A remarkable resilience of a market economy can be seen from a frequency stability compared to an amplitude variability of index levels. The role of time scale and preferred reference from economic observation is discussed. Part III on population dynamics with micro interaction provides a dynamic approach in studying social behavior including learning, communication, and its outcome in market share competition. Micro dynamic models provide fresh insights to collective behavior and culture diversity in social evolution. Chapter 8 on origin of division of labor and stochastic mechanism of differentiation in 1987 is a short paper with three basic models in micro dynamics and social evolution. The first model introduced a culture factor in risk orientation, which is observed by the remarkable difference between Western and Oriental cultures. Population dynamics of information diffusion and learning competition is developed for modeling division of labor. Its stability condition shows a tradeoff between stability and diversity. The second model is a stochastic birth process for multi-staged development. Emergence of multi-humped distribution, a typical feature of far from equilibrium dynamics, is observable from a theoretical model and empirical evidence. The Gaussian-type distribution breaks down during transitions between different birth rates. The third is a thought experiment of bifurcation mechanism in social behavior. It was a response to skeptics, who questioned the relevance of nonlinear dynamics in social science. Chapter 9 on imitation, learning, and communication in 1991 is a nonlinear stochastic model for social psychology such as sudden changes of fashion or polarized orientation, which is often called animal spirits by Keynes or irrational behavior in behavioral economics. Their probability distribution is a central or U-shaped pattern. The weakness of the Ising model in equilibrium physics is the concept of social temperature, which is hard to define for non-Hamiltonian systems. A modified population model with social interaction is developed. Imitation, learning, and communication play important roles in understanding complexity in social behavior including a rapid swing in social modes and market variability. An equilibrium perspective based on a mean-variance approach

breaks down for U-shaped or bi-modular distribution, since its mean is least likely and its variance is explosive. Chapter 10 on Needham’s question and China’s evolution in 1990 was a historical and philosophical consideration of nonequilibrium social transition in the shadow of the June 4 event in 1989 during China’s transition from a command economy to a market economy. Joseph Needham’s question of why did capitalism emerge in the West and not in China is discussed. There are other related puzzles, such as the Chaunu-Wallerstein puzzle of why land-rich Europe lacks space while population-heavy China lacks population during the civilization division around the fifteenth century; and why a centralized state emerged in pre-modern China 2000 years ago, but division of labor was hard to develop during China’s modernization process. The labor-saving resource-intensive technology developed in Europe and resource-saving labor-intensive agriculture developed in China was shaped by ecological constraints, climate change, and patterns of war. The trade-off between the stability and complexity of socio-ecological systems is studied. The evolutionary perspective of nonequilibrium thermodynamics and nonlinear dynamics is helpful in understanding China’s transition and economic reform. Chapter 11 on China’s challenge to economic orthodoxy is a dialogue with American economists on the issue of shock therapy in 1993. From an evolutionary perspective, the Asian experience and China’s reform can be considered as a self-organizing process, not an equilibrium process. The success of China’s economic reform, contrasted with the difficulties of EEFSU, stems mainly from China’s willingness to tolerate decentralized experimentation and a gradual evolution of new institutions, whereas in EEFSU a belief of institutional convergence led to the wholesale importation of foreign institutions. This contrast highlights the advantages of a nonlinear nonequilibrium paradigm in evolutionary economics over a linear equilibrium paradigm in neo-classical economics. Part IV on meso foundation and equilibrium illusions discusses the source of business cycles and false beliefs in equilibrium theories, such as the perpetual motion machine, representative agents, and the driving force of institutional changes. Chapter 12 on the Frisch model of business cycles in 1999 was a first examination of the theoretical foundation in equilibrium economics. Surprisingly, the most influential model in business-cycle theory and macro econometrics was a spurious doctrine, but a mysterious success, which shared the first Nobel Prize in economics. Frisch once claimed that persistent cycles could be maintained by external shocks, but never formally published its analytical proof. Physicists had already proved that a harmonic oscillator under Brownian motion would quickly cease its harmonic oscillation. An empirical study of US real GDP cycles further reveal that the linear stochastic model is not capable of explaining observed persistent cycles. Theoretically, the Frisch model is a perpetual motion machine, which violates the law of thermodynamics. Persistent business cycles can only be generated by nonlinear economic dynamics. This paper was rejected by a mainstream economic journal, but its main findings were disclosed in Chapter 3.