Foundation of Consumption Behavior, Knowledge and Information, and Modeling Consumption Function in Iran

This article investigates private consumption behavior in Iran using annual data over the period 1959  2003. There are two approaches in modeling consumption function. First approach is based on the Euler equation while the second approach aims to identify the main determinants of consumption. One interpretation of the two approaches is that Euler equation test just theories, while the second approach is designed to answer different questions, primarily about the role of different variables in the consumption function. Derivation of Euler equation is based on rather strange assumptions when it comes to explaining process of knowledge, information, and long-term change. Since this approach takes no account of imperfect information and unstable preferences, the second approach is considered to model the consumer behavior in Iran. A linear model is specified and estimated and then due to the rejection of linearity, a nonlinear process is used as a plausible specification of consumption function. Smooth transition regression (STAR) models are used to model consumption function. The null hypothesis of linearity is rejected in favor of the logistic STAR (LSTAR) specification. The results show that the long run determinants of the private consumption in Iran are disposable income, wealth, and the rate of inflation. Moreover, findings show that the lack of sufficient knowledge and information of consumer is one of the determinants of inertia behavior of consumer in Iran. Moreover, the estimate of transition parameter in nonlinear model indicates that the speed of transition from one regime to another is quite smooth and gradual. While the assumption of the equilibrium term makes the linear model unbalanced, a LSTAR model specification yields a balance equation. Therefore, the empirical evidence suggesting that modeling of consumption is in favor of nonlinear rather than linear.