Evaluating a Simple Method for Estimating Sub-National Trade Flows

Given the non-uniform distribution of industrial activity within a country, it is well-established that the effects from changes in trade policy have considerable sub-national heterogeneity. (Andriamananjara, Balistreri and Ross (2006), Balistreri, Bohringer, and Rutherford (2018), and Caceres, Cerdeiro, and Mano (2019)) For industry specific, partial equilibrium analysis, a gravity model based approach, as in Riker (2019), is potentially the most tractable method for time-sensitive analysis because it relies on readily available national accounts and import data to estimate sub-national flows. In this paper, we utilize Japanese inter-prefectural trade flow data from the Japanese inter-regional input-output table from 2005, the most recent year available, to evaluate the estimation method proposed in Riker (2019). In the first round results, we find that the gravity model based method approximates the observed inter-prefectural flow pat-tern reported in the 2005 Japan inter-regional input-output table, but only after tuning of the trade cost parameters. The main challenge is the lack of trade publicly available Japanese import trade margin data.