We consider a policy regime allowing academic institutions to grant industry the intellectual property rights (IPRs) over invention resulting from collaborations. If a firm plays an important role in generating an invention, the researcher offers the IPRs to the firm, as an incentive to collaborate. However, he retains certain domains where he can exploit an invention without having to apply for a license. The choice of these domains involves a tradeoff. In fact, the researcher either induces the firm's effort, by assigning a broad field of use, or he ensures that he can use an invention in other applications, by granting a narrow field of use. The reverse occurs if it is the researcher who plays an important role in generating an invention. The main difference, however, is that if effort were contractible, the firm could reward the researcher for supplying the first best level of effort, because, unlike the researcher, it is not cash constrained. An empirical analysis, based on École Polytechnique Fédérale de Lausanne research contracts, supports the role of broad fields in bolstering a firm's effort, when the latter is important for generating an invention.