China has huge differences among its regions in terms of socio-economic development, industrial structure, natural resource endowments, and technological advancement. These differences have created complicated linkages between regions in China. In this study, building upon gravity model and location quotient techniques, we develop a sector-specific model to estimate inter-provincial trade flows, which is the base for making a multi-regional input-output table. In the model, we distinguish sectors with less intra-sector input from those with larger intra-sector input, and assume that the former sectors tend to compete among regions while the latter tend to cooperate among regions. Then we apply this new method of inter-regional trade estimation to three sectors: food and tobacco, metal smelting and processing, and electrical equipment. The results show that selection of bandwidth has a significant impact on the assessment of inter-regional trade. Trade flows are more scattered with the increase of bandwidths. As a result, bandwidth reflects the spatial concentration of geographical activities, which should be distinguishable for different industries. We conclude that the sector-specific spatial model can increase the credibility of estimates of inter-regional trade flows.
- multi-regional input-output analysis
- sector-specific statistical model
- trade flows