This article is inspired by the observation of a massive change of opinion on the part of economists regarding the question of the patent as a mechanism allowing an economically acceptable solution to be found to the so-called “imperfect appropriability” problem. This change is due partly to empirical observation of the system’s disorientation, leading to situations where the ways in which intellectual property rights are used create more obstacles to innovation than they eliminate. On the other hand, this shift in opinion is also linked with empirical and theoretical analyses of situations where it seems that patents and other intellectual property rights can very well be dispensed with without innovation suffering. It is to these analyses that this article is devoted. It will first review works concerning historical economics and the economics of innovation and technology management that highlight innovation systems in which the absence of intellectual property rights or the weakening of the latter does not seem to undermine the capacity to invent and innovate. It will then move on to theoretical works that explore the conditions in which a competitive market can be characterised by optimal properties of resource allocation in R&D and innovation. Finally it will present works on experimental economics that attempt to reproduce an innovation process in a laboratory with a view to studying the respective merits of different institutions –patent or market– regarding their aptitudes to facilitate innovation.