Models and Forecasts of Inflation in a Developing Economy

A striking feature of inflation in developing countries is the high variability from one period to another. This feature would seem to make forecasting of the inflation rate a difficult task. This article applies three different forecasting techniques to predict the monthly rate of inflation in India over the period June, 1971 to May, 1980: The three methods used include the more conventional regression method and the newer time‐series and combined regression‐time‐series methods. A subsidiary objective is to examine the empirical applicability of the monetarist model of inflation in a developing economy. It is found that the combined regression‐time‐series model and the regression model are good predictors of the monthly rate of inflation in India. The results also indicate that the monetarist model performs well in predicting the monthly inflation rate in a regression framework.