Symmetrical Approaches for the Non-Survey Regionalization Techniques: Ameliorating the Flegg’s Location Quotients

Abstract

In most countries, policy planners face a lack of published primary regional and local input-output (I-O) data for analysing productive networks, which has led researchers to develop various non-survey techniques for the secondary estimation of regional and local intersectoral direct requirements coefficients, serving as the basis for calculating sectoral multipliers. This study seeks to improve non-survey regionalization techniques to better capture regional and local sectoral specializations and to produce more accurate sectoral multipliers for subnational development planning. The hypothesis is that a symmetrical and unrestricted use of the simple location quotient (SLQ), as part of the adjusted Flegg’s location quotient (aFLQ), such as the proposed KFLQ variation, can provide a more reliable database for modelling regional development. Under this approach, regional and local coefficients are allowed to surpass national averages. For the empirical analysis, the productive network of the West Greece region was simulated. Weighted and non-weighted type I backward sectoral employment multipliers were estimated to illustrate the differences resulting from the application of various regionalization techniques. The hypothesis was tested using the assumption that the parameter δ should be set so that KFLQ approaches 1 when the regional-to-national size of a sector approaches its average national allocation across regions. For SLQ, this occurs for each sectoral indicator at approximately 1.5. This assumption resolves the problem of the previously arbitrary definition of the exponent δ.