Integration, Causality and Prediction of Agricultural Commodity Outputs in India

Identification of major sources of growth from agricultural and allied sector would provide the way forward for strategic framework for enhancing farmers’ income. Since gross domestic product data are available for the disaggregated forms of crops, livestock, fisheries and forestry sub-sectors, the value of production (VoP) from various commodity groups (at 2004–05 prices) has been used as the most appropriate indicator for commodity outputs. The commodity VoP series from 1950–51 to 2014–15 have been used to study integration, causality and projections. Various linear and non-linear techniques along with their combinations have been applied for prediction of VoP in India till 2022–23. The VoP is projected separately for cereals, pulses, oilseeds, sugar, fibres, condiments & spices, fruits & vegetables, fisheries, and livestock. The hybrid models, based on wavelet, have been used along with usual ARIMA model for projecting the VoP. The integration among VoP from different sources is investigated along with the existence of causality. It is found that VoP from crop, fisheries and livestock are cointegrated among themselves. The pairwise cointegration was also found to be significant and there existed bidirectional causality in the long-run in most of the cases, indicating the complementarities and growth linkages.