Control and Benefits in a Research Collaborations

In many sectors such as biotech, large firms contract and collaborate with small research units on single projects. Designing an optimal contract to govern such collaborations, however, turns out very often to be a problem, as it entails deciding on issues such as the unit's research input, its monetary compensation, and so forth. Recent results, however, suggest that the role played by the allocation of control over innovation is more significant than previously thought, and it could be fruitfully used to design better contracts by shifting the benefits obtained through the collaboration from one party to the other.

In a recent working paper titled Control and Contract Design in Research Collaborations: A Complete Contract Perspective, forthcoming at International Journal of Industrial Organization (doi: 10.1016/j.ijindorg.2012.03.001), Claudio Panico (Department of Management and Technology) makes an original contribution to extant theory using a complete contracts viewpoint to investigate the allocation of control. The premise is that control can be allocated ex ante (i.e. during the contracting stage) according to the information reported by the unit to the incumbent. Additionally, control does not have to be completely centralized or delegated; it can also be fractional, meaning that each party holds a fraction of it.

To investigate the contracting implications the above conditions imply, the author develops a formal model of a research collaboration in which a large incumbent company contracts with a cash-constrained research unit. The model shows how to use control to shift the private benefits from one party to the other. Unlike previous models that focus on moral hazard issues, the framework reveals the allocation of control under adverse selection. Additionally, the conditions in which it makes sense for the incumbent to pick delegation, centralization or fractional control emerge clearly, as do the cases in which it is convenient to retain full control and reward the unit with a monetary payment. It is shown that the incumbent can generate countervailing incentives and reduce the unit's information rents, even to the point of overcoming the well-known limitations due to adverse selection.

The model also has implications for practice. First, the incumbent may find it optimal to delegate control when the unit is a research-intensive firm that runs several projects and activities in parallel, as larger technological capabilities imply a larger opportunity cost. And second, the incumbent may decide to retain more control if it owns critical resources that allow to leverage the unit's technological skills, allowing for significant marginal benefits from research. Overall, the analysis highlights under which conditions control sharing is more likely to occur, i.e. when the unit's marginal opportunity cost is relatively small or when it obtains significant marginal benefits from the project.