Testing the Permanent Income and Random Walk Hypotheses for Turkey

Abstract

There have been ongoing researches and debates on the dynamics of developing economies with structural changes within literature of economics since 1770s. Economics investigated some basic themes during Mercantilism and Physiocracy period and, however, shaped its scientific approaches together with related techniques of philosophy/mathematics/statistics through the models of Classical, Keynesian, Neo-classical, Neo-Keynesian and Monetarist. Among these approaches, Permanent Income Hypothesis (PIH) and Random Walk Hypothesis (RWH) have been searching/analyzing the alternative consumption functions with related possible significant parameters since 1930s. This paper, first observes different income-consumption relations through decomposing them by using Autoregressive (AR) process to test the validity of Permanent Income Hypothesis, later, considers testing the Random Walk Hypothesis by decomposing permanent income into expected and unexpected parts throughout AR processes. Thereby, paper intends to explore, if exists, excessive smoothness and/or excessive sensitivity of consumption in Turkey by employing Turkish quarterly data for consumption and income, spanning from 1998:1 to 2012:1. Outcome of this paper indicates that consumption is found sensitive to changes in unexpected income as well as changes in expected income. Findings reveal overall that Absolute Income Hypothesis is confirmed in Turkish economy.

Keywords

• Consumption, Income Hypothesis, Auto-regressive Process, Excessive Sensitivity