STATISTICAL PERSPECTIVE OF OILSEED CROP PRODUCTION IN INDIA: TREND, INSTABILITY, DECOMPOSITION AND ARIMA FORECASTING

Abstract

The present study was conducted to analyse the long-term trends and variability in the area, production and productivity of total oilseed crops in India, to estimate the contribution of area, productivity, and their interaction to production growth, and to forecast future area, production, and productivity using time series analysis. The study was based on secondary time-series data for the period 1950-51 to 2024-25, collected from the IIOR, Hyderabad and Directorate of Economics and Statistics, Government of India, New Delhi. Descriptive statistics, compound growth rate, coefficient of variation, the Mann–Kendall trend test, decomposition analysis, and the Box–Jenkins ARIMA methodology were used for data analysis. Stationarity was examined using the Augmented Dickey–Fuller (ADF) test, and the analyses were carried out in R. The fitted models were evaluated using MAPE, AIC, and BIC. The results showed an increasing trend in the area, production and productivity of total oilseed crops India. Decomposition analysis revealed that productivity contributed the most to production growth, accounting for 70.3% in India, followed by the area effect (18.6%) and the interaction effect (11.0%), respectively. For forecasting ARIMA (0,1,0) with drift, ARIMA (1,1,0) with drift, and ARIMA (1,1,1) with drift model were selected for oilseed area, production and productivity for India respectively. The selected models showed satisfactory forecasting accuracy, with MAPE values ranging from 3.91 % to 10.84 %. The forecasts indicated a gradual increase in the area, production, and productivity of total oilseed crops during the forecast period. The findings suggest that ARIMA models are useful for analysing historical trends and generating reliable short-term forecasts, providing valuable support for planning and policy decisions related to oilseed development in India.