As industry-university collaborations are promoted to commercialize university research and foster economic growth, it is important to understand how companies benefit from these collaborations, and to ensure that resulting academic discoveries are developed for the benefit of all stakeholders: companies, universities and public. Lock up of inventions, and censoring of academic publications, should be avoided if feasible. This case-study analysis of interviews with 90 companies in Canada, Japan, the UK and USA assesses the scope of this challenge and suggests possible resolutions. The participating companies were asked to describe an important interaction with universities, and most described collaborative research. The most frequently cited tensions concerned intellectual property management and publication freedom. IP disagreements were most frequent in the context of narrowly-focused collaborations with American universities. However, in the case of exploratory research, companies accepted the IP management practices of US universities. It might make sense to let companies have an automatic exclusive license to IP from narrowly defined collaborations, but to encourage universities to manage inventions from exploratory collaborations to ensure development incentives. Although Canada, the UK and US have strong publication freedom guarantees, tensions over this issue arose frequently in focused collaborations, though were rare in exploratory collaborations. The UK Lambert Agreements give sponsors the option to control publications in return for paying the full economic cost of a project. This may offer a model for the other three countries. Uniquely among the four countries, Japan enables companies to control exclusively most collaborative inventions and to censor academic publications. Despite this high degree of control, the interviews suggest many companies do not develop university discoveries to their full potential. The steps suggested above may rebalance the situation in Japan. Overall, the interviews reveal the complexity of these issues and the need for flexibility on the part of universities and companies.

In some US states, policy makers, pressed by local and regional industrial interests, are debating how to ‘reform’ technology transfer at public universities. ‘Reform’ in this context is generally understood to mean redirecting university technology transfer activities to increase the benefits of state-funded research to local industries. Progress towards this goal is often constrained by federally mandated laws applicable to technology transfer at universities (such as the Bayh–Dole Act) and by university policies that have been placed by state legislatures outside the purview of policy making state officials. Calls for change have also been countered by the view of many universities that the system is not broken. Suggested reforms range from the abolition of the Bayh–Dole Act, which gives universities the flexibility to transfer ownership of federally funded inventions to local industries, to structural or management changes in universities that will promote innovation and/or expedite the licensing of new ideas. This article proposes a new paradigm: instead of measuring the success of technology transfer by counting numbers of patents and licensing deals, the authors suggest measuring knowledge flows between state universities and their localities. This approach should produce a more accurate picture of the full impact of universities on their regions.