Pré-publication, Document de travail: This research studies the effects of a R&D tax credit and a R&D subsidy in Mexico.The Mexican tax credit removed the usual market oriented traits that define most taxcredits. It essentially acted as a “deferred” subsidy, as firms got a discount on theircorporate tax at the end of the fiscal year. Whereas the subsidy granted the funds atthe start of the R&D project. My estimates show that both policies had a positive im-pact on innovation personnel, but the subsidy’s impact was larger. As for patents, theimpacts are less clear but favor the subsidy over the tax credit. The subsidy appearsto have allowed less profitable firms to take on their R&D projects. This might havedriven the larger subsidy effects. The awarding procedure in both programs is similar.Firms submitted their R&D projects to a non tax collecting institution. The projectswere evaluated according to detailed guidelines. The awarded projects were selectedbased on the evaluations. The guidelines allow to construct a set of conditioning vari-ables in a matching estimation approach. In addition, I use the difference-in-differencematching method to purge time-invariant unobservables.