Regional Industrial Clusters : A Note on the Flaws of MFA (Minimal Flow Analysis)

MFA (Minimal Flow Analysis) is a method of qualitative input-output analysis used for identifying national of regional industrial clusters. It is based on the analysis of layers (in an input-output model, flow matrices generated at each iteration toward equilibrium). We show through theory that all normalized layers (column-coefficient matrices computed from each layer) are equal in MFA. This is illustrated by an application to France's 2009 input-output table. Therefore, the information brought by one layer is identical to those of another layer: layers bring no specific information, which unfortunately implies that MFA is not a good tool to study intersectoral linkages, despite the interesting character of the topic