Measuring competitiveness differentials inside the same region: a propensity-score matching approach

This paper analyses regional competitiveness at the subregional level through a novel methodological approach that adopts a matching design. By comparing the performance of similar firms in different parts of the region, it is possible to detect whether different places provide different competitive territorial assets. Using data for Lombardy, a large and competitive European region, the analysis shows that the different territories of the region are differently competitive in different industries, even when they are similar in terms of total GDP per capita or specialization. The paper also confirms that measuring competitiveness on different indicators (Labour Productivity, TFP, Profitability) can provide different results, and this especially happens when comparing static and dynamic indicators. The methodology presented here is especially relevant to the design of regional policies, that are mostly deployed at the NUTS-2 level but would benefit from accounting for the presence of strongly dis-homogeneous territories inside the same region.