ABSTRACT
Time series data of social processes show not only random fluctuations around some invariant mean, but structural breaks, new periodic patterns, and sudden shifts away from what initially seemed to be trends. Inoculation programmes and population growth exhibit random spikes while the historical records of the Nile’s water level illustrate long periods of stable trends interrupted by random and sudden rapid changes.
This chapter studies the question of whether such ostensive randomness is caused by extrinsic stochastic shocks or whether these variations are inherent to the system at hand. It introduces readers to chaos theory and shows that deterministic chaos exhibits consistent properties that make it predictable. The chapter builds on the models of Chapter 3, Chapter 5, and Chapter 7 and shows that equilibrium convergence is not a universal characteristic of the evolutionary models discussed in these earlier chapters. Minor variations in the composition of a population or the rate of adoption of behavioural rules can lead to different historical trajectories. This chapter, therefore, offers a different perspective on history dependence and its impact on long-term institutions.
Among the concepts discussed are: the Feigenbaum constant, flip bifurcations, bifurcation diagrams, critical value curves, intermittency, the Ising model, and phase transition.
