R&D policy evaluation: a case study on Law 46/1982 in Italy

Abstract

Even in the context of neo-classical theory, characterized by a strong trust in the virtues of the free market, the need for a policy intervention in the field of research and innovation is theorized, because of the presence of market failures. In the contemporary context of economic and public finance crisis, on one side there is a strong need to boost the industrial productivity through investment in research and technology; on the other side the public budget constraints call for prudency. In this context the need for systematic evaluations of public incentives to firms is particularly strong. In this context, the paper offers an evaluation exercise on the major instruments used to promote R&D and innovation activities of Italian firms. The analysis concentrates in particular on the incentives offered by Law 46/1982 (and revisions) and their effects on firms expenditures and employment in R&D. The Law represents the main national funding program as well as the longest lasting single instrument for technology and innovation promotion, although revisions of the Law have occurred. In addition to Law 46/1982, Italian firms have the chance to benefit from a larger span of public subsidies, particularly at the regional and local level. This makes the counterfactual question of “what would have happened without the policy” particularly interesting since it is likely, and this is confirmed by the data, that firms that do not access to the incentives of Law 46/82 benefit from other sources of public financing. Therefore it is particularly important to consider the effects of Law 46/82 not just in the hypothetical situation of complete absence of policy intervention, but also when other similar laws are at work. For this reason besides the difference-in-difference estimation, the paper analyses the effects of Law 46/1982 through a difference-in-difference-indifference model, which allows to verify whether the interaction between different kinds of incentives has a multiplicative or a substitutive effect. The paper also addresses another typical aspect of evaluation studies: the average effect of the policy normally retrieved seems to be not sufficiently informative, especially in a country characterized by a marked territorial economic dualism (Centre-North vs. South) and by a strong presence of small-medium firms, with profoundly different characteristics from large firms. In other words the effects of the policy instrument might vary substantially among firms. Therefore the paper takes explicitly into consideration the different effect of the incentives on different sectors (according to Pavitt classification), on different size of firms and in different zones of the country. The database used for the analysis is the Capitalia Survey (former Mediocredito Centrale). Data are obtained from three consecutives surveys for a total of nine years starting from 1995 to 2003.