Supply response with mix of stationary and nonstationary data: Case study in pulses, India

Supply response studies in the past were based on traditional econometric techniques (classic linear regression) and the nerlovian framework. Results of traditional econometric techniques are reliable when the time series data are stationary. However, there can be a possibility of some macroeconomic time series data are non-stationary, thereby results and conclusion drawn from using those techniques are having the risk of invalidity. This paper specifically attempted to quantify the relationship between pulses production and price and non price factors viz., land productivity, annual rainfall, irrigated area and revenue difference between cereals and pulses, when the variables in the data expressed in levels are neither stationary [I(0)] nor non-stationary [I(1)], and do not have the same order of integration. Finding of this study suggests rainfall and revenue difference between the cereals and pulses are major determinants of pulses production.