R&D efficiency of Mexican regions ? an output DEA approach-

Introduction: Performance of regions is commonly based on the comparison of economic indicators. This may help to understand particular situations and identify optimization alternatives. In order to adopt the most effective technological research, development and innovation policies there is a need to measure the results that firms obtain and the impact of performed projects in regional economy. The results of R&D projects of the productive subsystem impact the productive subsystem itself and also other stakeholders in the region, including the knowledge subsystem. This paper contributes to the debate on how to measure regional R&D performance. Different research lines argue about choosing global regions' measures or firms' ones. In this analysis, we compare both alternatives. Methods: The study applies a Data Envelopment Analysis (DEA) to the evaluation of the Innovation, Research and Development efficiency of regions, by comparing the innovation results of the firms, with the economic results of the regions they are located in. A 2 stages analysis is proposed, focusing firstly on the productive subsystems' efficiency and then on the comparison between the efficiency of the regions. The analysis is performed for Mexico federal regions, considering the ESIDET 2011 survey. This survey measures the behavior of Mexico as a whole, in R&D projects. The 2010 and 2011 edition of the survey include the results obtain in each region. Taking into account the variables analyzed in other R&D DEA studies, and keeping in mind the available indicators of the ESIDET survey, a Multivariate Analysis of Variance (MANOVA) will be performed, in order to identify the significant variables, which will turn into the DEA Input and Output variables. The objective of the analysis is to identify the most efficient federal region, according to the maximization or augmentation of the obtained benefits. The output DEA orientation aims to identify the factors that maximize the benefits of the R&D and innovation projects for the regions. Results: The main conclusion is that the most R&D involved regions - more R&D dedicated firms, higher R&D profits ? are not the more R&D efficient regions. Less R&D devoted regions result in more efficient ones eventually due to a better use of R&D resources by the few R&D firms located in those regions. Particular region analysis allow to identify how regions could maximize their results.