Endogeneous Matching in University-Industry Collaboration: Theory and Empirical Evidence from the UK

Do top academics collaborate with top firms, whereas less productive researchers collaborate with less productive firms? Do they collaborate because they have similar preferences? Do partners choose each other because of individual or institutional characteristics? Are other, less productive and more applied academics, more likely to stay independent?

In 2007, Professor Sir Colin John Humphreys of Cambridge University, who specializes in electron microscopy and analysis, formed a partnership with FEI, a leading company in the production and distribution of electron microscopes. In this case, a prolific researcher of a top university, whose research is considered basic, collaborated with a research-intensive firm, heavily oriented toward basic research.

The full transformation of modern societies from industrial into knowledge- and science-based economies requires collaboration not only within academia or firms but also across research and production. In their recent contribution to the Barcelona Graduate School of Economics working paper series (No. 704) (July 2013), Albert Banal-Estañol, Inés Macho-Stadler, and David Pérez-Castrillo study collaboration between academics and firms. They analyze which academic researchers and firms engage in collaboration and which two-sided market partnerships are actually formed.

The market model

The authors are the first to provide a one-to-one two-sided matching market model of academic researchers and firms developing research projects. Participants on each side of the market prefer different types of project, i.e. in terms of the degree of “appliedness,” and in terms of their scientific ability. Each participant can either develop a project on her own or search for an appropriate partner on the other side of the market to develop a collaborative project. In a collaborative project, the value to each participant depends on the agreed type of project and on the investments of both participants.

The theoretical model makes three predictions concerning the partnerships formed:

  • First, due to the complementarity in abilities top academics collaborate with top firms and less able academics with less able firms.
  • Second, the more basic (applied) researchers collaborate with more basic (applied) firms.
  • Third, the higher the ability of the academics, the closer they are to partners of their ideal type.

In addition, the model predicts that if the costs due to collaboration are not too large, the most able and applied researchers and the most able and basic firms engage in collaboration, whereas the others stay independent.

Data and findings from UK firms and institutions

These theoretical predictions are tested on the teams of academic researchers and firms that have proposed research projects to the Engineering and Physical Sciences Research Council (EPSRC), the main government agency for research funding for the engineering departments of the UK universities. Publications of the five years prior to the project allow the authors to construct variables that measure the ability of academics and firms and the type of projects they like.

Matching is found to occur amongst the best in terms and over the whole distribution in terms of the type of research. Assessing the relative importance of the horizontal (type of research) versus the vertical (ability) characteristics, the authors find that the horizontal characteristics dominate, both in terms of magnitude and significance. Also, the individual characteristics are found to be more relevant than those of the institution. A top firm tends to collaborate with a top researcher, independently of her institution.

Concerning the characteristics of the academics that submitted collaborative instead of non-collaborative projects, the authors find that academics that are above the median in terms of ability and those who are more applied than the median are 9.1% and 39.5% more likely to propose collaborative projects, respectively. In terms of institutional characteristics, academics in larger universities, in terms of number of projects, are also more likely to submit collaborative projects.

Why is this matching process important?

Understanding the two sided matching process is important because, as the authors show in their previous Barcelona GSE Working Paper (March 2011), the characteristics of the matched partners determine collaboration outcomes. Indeed, they use the same dataset to study the effects of the composition of the two-sided market partnerships on the outcomes of the project in terms of publication results. And they show, for example, that research projects in collaboration with firms produce more scientific output than those without them when the firms in the project are research-intensive.


“Research Output from University-Industry Collaborative Projects” by Albert Banal-Estañol, Inés Macho-Stadler, and David Pérez-Castrillo. Barcelona GSE Working Paper Series (539).