M. R. Singariya^1,, S. C. Naval¹

The objective of the present paper is to examine casual relationship between GDP, agricultural, industrial and service sector output in India using time series data from 1950-51 to 2011-12.The study conducts an econometric investigation by applying methodologies, viz., Stationary tests, and Johansen’s Cointegration test, Johansen’s Vector Error Correction Model (VECM) in VAR and Impulse Response Function and Variance Decomposition Analysis. With all the variables in log terms being I(1), Johansen’s co-integration test confirms two long run relationships among the variables at 5% significance level. It reveals that there exists bidirectional causality among the agriculture, industry, service sector and GDP and agriculture and industrial sector with services sector, while there is a unidirectional causality between agriculture and industry sector. However, results based on vector error correction model indicate a weak association between the sectors in the short run. Dynamic causality results show that contribution GDP forecast error by the services sector is the highest, followed by agriculture and industry sectors, while the contribution to the agriculture sector forecast error by GDP is the highest, followed by the service sector and industry. In the case of the industry sector, the explanatory power of one standard deviation innovation in the agriculture sector and the services sector to forecast error variance is quite high (33.38% and 5.38%). Further, results of decomposition variance analysis and impulse response suggest that the agriculture sector plays the main role in determining the overall growth rate of the economy through its linkages to other sector. The analysis of inter – sectoral linkages identify agriculture as the main economic activity that controls most economic activities in India.

GDP, agriculture, industry, services sector, Cointegration and Vector Error Correction Model, Impulse Response Function and Variance decomposition analysis