Can we reveal the growth potentials with input-output model at sub-national level?

The regional growth potential, which frequently appears in economic analyses, is somewhat difficult to handle in practice. Estimation of growth potentials for development forecasting and strategic planning remains actual and meaningful, which is especially true for long-term predictions of structural shifts in the economy, both at national and regional level. The Regional Input-Output (RIO) model has been a powerful tool in applications of the industrial interdependence, impact analysis and policy strategies. Also the RIO model can be attempted to identify leading industries through linkage effects. The backward and forward linkages for the sectors in a single economy provide a mechanism for identifying key or leading sectors. In this study, we try to identify the leading industries by the RIO model along with non-survey techniques to explore growth potentials. The estimated multipliers which represent the backward linkage are compared with the survey based benchmarks through the rank correlation coefficients (Spearman's Rho and Kendall's Tau). From the empirical results, it appears that non-survey techniques may distort the information of regional industrial structure and be incapable for identifying leading industries. This suggests that more cautions might be needed while using LQs to identify the regional leading industries and reveal the growth potentials.