Improved Estimation of Regional Input-Output Tables Using Cross-Regional Methods

Xuemei Jiang (1) , Erik Dietzenbacher (2, 3) , Bart Los (2)

1 Center for Forecasting Science, Academy of Mathematics and Systems Science
2 Faculty of Economics and Business
3 Graduate University

If a survey to construct a regional input-output table is too costly, nonsurvey methods are needed to estimate such an 'object table'. Often, information on the margins of the object table can be complemented by full information on intermediate inputs from tables for other regions. This paper compares the estimation performance of four 'cross-regional' methods. Two of these have been proposed in the literature before, while the other two are novel and based on recent advances in linear regression analysis. The methods are not only tested against each other, but also against popular methods that do not employ any regression technique. To this end, 27 regional input-output tables for China in 1997 and 2002 are used.

Mots clés: Social Sciences & Humanities

Xuemei Jiang, Erik Dietzenbacher, Bart Los. Improved Estimation of Regional Input-Output Tables Using Cross-Regional Methods. Regional Studies, 2010, pp.1. ⟨10.1080/00343404.2010.522566⟩. ⟨hal-00648528⟩