The private order of innovation networks.

• Author: Jennejohn, Matthew
• Position: II. Multivalent Contracting B. The Implications of Multidimensional Exchange Hazards through Conclusion, with footnotes, p. 323-366

2. The Implications of Multidimensional Exchange Hazards

   The discussion in this Part has so far laid the groundwork for the claim that multiple types of transaction costs affect the design of relational contracts. Including spillover and entropy concerns in our theory of economic organization introduces an intriguing possibility: rather than just responding to holdup threats, collaborators may design their agreements to address an array of innovation problems. In other words, both sides of the theoretical equation--how exchange hazards are understood and how governance tools are fashioned--can be pluralistic. The result is a highly complex decision landscape. Parties not only have to navigate more than one type of transaction cost, but they must also choose how to combine different types of governance tools into a coherent portfolio--the multivalent contract.

   Multidimensionality raises two important possibilities. First, it expands our understanding of the ends a governance mechanism can serve. Consider, for example, how the management committees at the core of Gilson et al.'s braiding theory might serve purposes other than addressing opportunistic holdup. Perhaps the most obvious explanation for consensus-based committees is that they give parties veto rights over the trajectory of the collaboration, not that they foster informal constraints. (201) As Gordon Smith notes, consensual decisionmaking allows parties to put the brakes on a joint discovery process without breaching the terms of the deal. (202) A multidimensional perspective illuminates why those rights may be so important. Consensus-based committees give a party a veto right at three points in the collaborative relationship where spillover issues arise. (203)

   First, in the collaboration's early stages, a unanimity requirement gives a party a chance to steer the collaboration in a direction that avoids outcomes that may, for example, conflict with adjacent areas of that party's patent portfolio or with other collaborations that it might have underway. In that way, the party can "design away" from fraught areas, and target "white space" in the patent landscape where some freedom of movement is available. Potential externalities are nipped in the bud, as it were. Delineation of discrete project boundaries is consistent with this reading--the committee is a tool for policing those boundaries and ensuring that the collaboration does not experience mission drift that interferes with adjacent projects.

   Second, the unanimity requirement gives a party a veto over definitional decisions relating to the boundaries of intellectual property. (204) The committee is a forum for negotiating the relationship between background IP and foreground IP as well as for drawing dividing lines based on those relationships. (205) Each party has a tool for protecting its interest when decisions are being made as to whether an invention will fall within one party's exclusive ownership or whether it will be owned jointly. Relatedly, the unanimity requirement allows a party to prevent its counterparty from making decisions with respect to patent prosecution or enforcement that might affect those boundary decisions or other aspects of a party's background IP. (206)

   Finally, where foreground IP is jointly owned, the veto is a means for controlling spillovers that could result from the choices of the other co-owner. Because U.S. patent law allows a joint owner to license and otherwise exploit a jointly owned asset without requiring approval of or accounting to the other owner(s), a party could find itself in a situation where, for example, its collaborator is licensing foreground IP to one of the party's competitors. Contract provisions requiring collective decisionmaking on licensing jointly owned foreground IP to third parties allows a party to address such situations, (207) and the veto right approximates through contractual means the right to exclude that arises from sole ownership.

   But veto rights responding to spillover concerns cannot be the entire explanation for consensus-based committees. An explanation focused on veto rights struggles to explain why multiple committees are used in some collaborations. If a consensus-based committee is simply a general-purpose veto mechanism, then why go to the trouble of establishing more than one such committee?

   Perhaps governance committees are also part of a modularity strategy. Where interactions between parties are thick, collaborators delegate management of those issues to a committee bound by a unanimity rule, which creates partial separation between the committee and the constituent firms, thereby compartmentalizing the decision environment. That separation allows the committee representatives to specialize in managing the discovery process. On the flip side, delegating management to a dedicated committee allows the parent firms to focus on systemic coordination issues while the committee focuses on coordination issues at the bilateral level. As the task environment's complexity increases, there is greater separation of and devolution to subcommittees. Finally, a modularity strategy allows disputes to be cabined, preserving executives' time and preventing dispute resolution in the context of a continuing relationship from transforming into litigation.

   The possibility of more than one exchange hazard affecting economic organization is attracting the sustained interest of management researchers, who are beginning to explore how governance choice correlates with indicia of problems other than holdup. (208) For example, Oxley and Sampson examine how spillover concerns and collaborators' ability to absorb one another's know-how affects the scope of electronics and telecommunications alliances, finding that greater spillover concerns and the parties' relative capacities to absorb information influence scope decisions. (209) In a study examining over (400) service contracts entered into by a single firm, Mayer and Salomon find evidence of interdependence between coordination problems and holdup problems: greater differences in parties' technological capabilities influence the intensity of the contractual hazards (such as holdup problems) the firms face, which in turn affects contractual governance choice. (210) Kale, Singh, and Perlmutter find that informal governance mechanisms can help address not only opportunism concerns but also joint learning problems. (211) Reuer and Devarakonda find evidence that a combination of exchange hazards affects the design of steering committees in alliance contracts. (212)

   That expanded understanding of contracts' ends naturally leads to the second implication: governance mechanisms are interdependent. A contract provision that addresses holdup problems may also affect, either positively or negatively, the mitigation of spillover and/or entropy concerns. The possibility of governance mechanisms complementing or substituting for one another occurs at a much more fine-grained level than the formal/informal debate envisions: the provisions of the formal contract may be complements to or substitutes for one another, and then we must ask how those formal provisions interact with any informal constraints that may affect the relationship. The implication is that governance mechanisms can present multidimensional tradeoffs.

   To illustrate, consider the following hypothesis: committees' responses to spillover problems may be complementary to their responses to entropy problems, but they may exacerbate holdup problems. A committee may be employed when intellectual property rights are strong and the intellectual property landscape is organized tightly, as in the life sciences and chemical industries, to prevent spillovers by midwifing new asset boundaries and coordinating a modulated task structure, as outlined above. In that respect, the committee governance mechanism operates to address the two exchange hazards in a complementary fashion. But there is a countervailing risk: the committee itself invites holdup via the unanimous decisionmaking rule, which presents an opportunity for deadlock. Therefore, it is only in those collaborations where coordination problems and intellectual-property-boundary definition are critical enough that parties are regularly willing to invite the holdup risk that comes with committee governance. In other words, holdup does influence the use of administrative mechanisms, but not in the way Gilson et al. envision--the contract referee mechanism creates holdup risk in some circumstances rather than minimizing it. (213) It is the balance between spillover and entropy concerns on one hand and holdup problems on the other that explains why certain industries use committees, unanimity rules, and escalation procedures and some do not. (214)

   Recognizing the contingent relationships between governance devices is powerful because it provides a framework for understanding at a detailed level the diversity of design strategies we observe in practice. If only holdup problems motivate contract design, then, as we saw in Part I.C above, differences in governance strategy are befuddling unless there is reason to believe that the intensity of holdup threats differs materially between collaborations. By acknowledging the potential for tradeoffs between governance tools, multiple exchange hazards gives us a framework for understanding why one collaboration may employ different mechanisms than another collaboration. Variations between the intensity of exchange hazards collectively affecting one collaborative relationship and the next lead to different governance tradeoffs for different collaborations. By illuminating those tradeoffs, a multidimensional approach removes the straightjacket of the holdup paradigm.

   3. Demonstrating Proof of Concept

   This Part explores whether the approach suggested above provides an answer to the fundamental question why we see such diversity in alliance...