Preliminary

This paper is aimed at exploring the e¤ects of R&D commercial subsidies by mod-elling the …rm decisions about performing R&D when some government support can be expected. To estimate the parameters of interest we use an unbalanced panel sample of 2,000 Spanish manufacturing performing and non-performing …rms. For the non-performing …rms, we compute the trigger subsidies required to induce R&D spending (a measure of the market failure). Among the performing …rms, we compute the pro-portion that would cease to perform R&D if subsidies were suppressed (a measure of subsidy efectiveness). In addition, we assess subsidy e…ciency for all the performing …rms by looking at the e¤ective cost reduction obtained by means of the grant. Results show, among other things, that subsidies turn out to be decisive in inducing to perform R&D among 20 % of the performing small …rms and only 4 % among the biggest. ¤We thank comments and suggestions by Isabel Busom. We acknowledge …nancial support from CICYT, projects SEC97-1368 and SEC1999-1236-C02-01