Traditionally, empirical and theoretical studies of coopetition have focused their attention on the interorganizational level (Brandenburger and Nalebuff 1996; Lado et al. 1997; Padula and Dagnino 2007). More recently, some studies have stressed the existence of simultaneously competitive and cooperative relationships between different organizational units (Tsai 2002; Luo et al. 2006). A recurrent rationale accounting for this surging interest is represented by the lack of a straightforward distinction between inter-and intraorganizational work settings. In this sense, the analysis of intrafirm coopetitive behavior becomes especially crucial in all organizations which involve elements such as temporary teams and independent (freelance) workers, as is the case in many knowledgeintensive industries. Faced with increasing organizational changes such as the blurring of organizational boundaries and employment of professionals working autonomously (see, for an empirical application, the problem of swift trust in temporary teams studied by Meyerson et al. 1996), the examination of the peculiarities of intrafirm relationships might provide a sort of “micro-foundation” in the study of coopetition and might contribute to the state-of-the-art in the field of organizational theories of the firm. While in the above-mentioned literature, intraorganizational coopetitive behavior has been investigated at the horizontal level only (e.g. cross-functional mixed-motive behavior), in this chapter we propose a simple hierarchical model which makes it possible to study coopetitive motives within the horizontal and the vertical dimension simultaneously. More precisely, we aim to observe the evolution and the determinants of coopetitive behavior in the smallest possible group, constituted by a principal and two agents. The principal might be an individual (e.g. a team leader), or an organizational unit supervising the production process involving different subunits. The agents might be two professionals (e.g. two consultants) or even two subunits that have been given the task of implementing some activities that are needed to reach a challenging goal/target. The adoption of such a model makes it possible to enrich the standard coopetition framework, in which mixed-motive relationships are modeled among peers and in terms of a social dilemma, by imposing other mixed motive elements

which are typical of hierarchical settings, such as incentive schemes and rent division mechanisms. The originality of the present work lies in the methodological approach, which is based on the use of laboratory economic experiments, and on the focus on the role of fairness and reciprocal concerns as behavioral determinants of intraorganizational coopetition. Equity and fairness in organizations have been widely recognized as key issues for many years by human-resource practitioners (see National Research Council 1991 for a review). However, only in the last decade, have they been put in the spotlight in empirical research, typically in the field of agency theory and contract design. Such developments have been fostered mostly by experimental studies investigating, under laboratory conditions, the behavioral consequences of alternative types of compensation schemes and the effectiveness of economic and non-economic contract enforcement devices on work effort levels. Much of the evidence gathered by experiments has shown that the behavior of subjects in agency relationships is significantly affected by relative and distributive concerns. Agents seem to take into account how principals behave and perform systematic intersubjective comparisons of utilities. Agents’ concerns for principals’ fairness leads to reciprocating behavior (costly punishment of principals’ unfair behavior or costly reward for principals’ fair behavior). In turn, principals also seem affected by fairness considerations, often suggesting “fair contracts” to the agents. In most of the previous analyses, the focus was usually restricted to bilateral, “vertical” relationships between a principal and an agent, reflecting the dyadic orientation of both organizational economics and theories of fairness. However, most organizational contexts imply pyramid-shaped, multi-agent structures. In such contexts, “vertical” fairness considerations often become inextricably intermingled with concerns for “horizontal” equity between agents. Studies of team compensation and peer-to-peer working relationships show that such considerations may be of crucial importance in affecting job performance (consider, for instance, the impact of relative evaluation or group incentive schemes, or the effects on task performance of information about peers’ remuneration). Fairness issues in hierarchies are thus at the crossroads of both horizontal and vertical relations. Nevertheless, very little research has addressed both of these two dimensions of interaction. Empirical research on “triangular” principalagent relationships is still substantially incomplete and gaps are also to be found in the theoretical background. To our knowledge, only a few theoretical studies (Mookherjee 1984; Itoh 1994) have developed the principal-agent framework in a multi-agent setting. Triangular features are similarly overlooked by economic theories of reciprocity. This chapter may be regarded as an exploratory attempt to blend vertical agency relationships with horizontal, peer interactions between agents. In contrast with previously existing three-person experimental games (such as Camerer and Knez 1995; Güth and van Damme 1998; Kagel and Wolfe 1999; McCabe et al. 2000), we introduce a hierarchical structure (one principal and two agents).1