Informed by the learning processes in the MG-FN relationship, we now discuss the findings of our longitudinal study of this buyer-supplier relationship, describing the case findings, presenting conclusions, and identifying future directions.
5.1. Case Findings
Our study of the MG-FN relationship shows that MG took the lead in designing the management control (MC) of this relationship. It also demonstrates that MG was unable to develop an overall MC package at the outset of the relationship. Since MG’s corporate managers had no prior experience in developing an outsourcing contract for managing and executing over 750 FM activities worldwide, they enlisted the help of a legal firm to develop the controls to be included in the contract. The contractual controls advised by the legal experts focused primarily on mitigating expected appropriation concerns, with little attention given to controlling the daily FM activities, which required adequate communication and information exchange between the partners’ operating managers, as well as clarity about their respective responsibilities.
Once the FM activities were underway, MG’s operating managers learned by trial and error how to develop controls for managing these daily activities. Clearly, both MG’s corporate and operating managers were unable to foresee all appropriation concerns and coordination requirements related to the MG-FN relationship, as assumed in a major stream of the management accounting literature drawing on Transaction Cost Economics (TCE) (e.g., Dekker, 2004, 2008; Langfield-Smith and Smith, 2003; Phua et al., 2011; Vélez et al., 2008). Both sets of managers needed to learn throughout the relationship how to develop and adapt the MC package. This finding supports literature suggesting that learning is a key mechanism for shaping and adapting controls in inter-organizational relationships (e.g., Bingham and Davis, 2012; Mayer and Argyres, 2004; Vanneste and Puranam, 2010).
In contrast to the management literature claiming that trial-and-error learning is the main driver of MC design, our study shows that MG’s managers also learned to design and adapt the MC package through other learning processes, such as advice from third parties, including the partner, experimentation, and cross-level learning. Experimental learning in our case took place in real situations, rather than offline as described in the organizational learning literature (Argote and Miron-Spektor, 2011; Zollo and Winter, 2002). We also observed that existing controls could stimulate learning processes, such as experimental learning in our case. This supports the conclusion of other management accounting researchers that MCs can promote curiosity and experimentation (e.g., Free, 2007).
A major distinction between our study and similar studies in the management literature, which describe learning processes, is that we focus on the design of MC packages, including both contractual and non-contractual controls, whereas other studies investigate the design of separate contractual controls. By examining separate controls, it is impossible to understand how partners control their buyer-supplier relationships, since the controls as a whole—and not separate controls—determine the control of these relationships.
Another finding is that studying learning processes at two management levels (corporate and operating) allowed us to gain an in-depth understanding of the focus of various learning processes and their contribution to the design of the MC package. Our case study shows that corporate managers focus on strategic aspects of the relationship, particularly reducing appropriation concerns, while operating managers focus on the daily execution of FM activities and the coordination challenges involved. As role theory suggests (e.g., Floyd and Lane, 2000; Janowicz-Panjaitan and Noorderhaven, 2009; Janowicz-Panjaitan and Krishnan, 2009), this difference in focus aligns with the roles the two types of managers fulfill. We observed that both types of managers learned different lessons, which led to the design of different controls: controls to mitigate appropriation concerns, and controls to solve coordination problems. All these controls were necessary to create an MC package capable of effectively controlling the buyer-supplier relationship.
Furthermore, by studying learning processes at two management levels, we identified various processes of cross-level learning. One example is when MG’s operating managers observed that the many acquisitions, which substantially increased the scope of FN’s activities, negatively impacted FN’s performance. The “performance-upon-request” clause required FN’s operating managers to manage the FM activities of newly acquired plants and meet the KPIs described in the contract. They discussed their concerns with MG’s corporate managers and suggested introducing a 90-day grace period before the KPIs had to be met. MG’s corporate managers agreed and incorporated the grace period into the KPI dashboard.
Another example is when MG’s operating managers realized that existing MCs hindered FN’s innovation over time. They questioned the long-term impact of some contractual controls, which led FN’s operating managers to focus on short-term cost-reduction activities instead of innovating their FM practices for long-term improvements. Scholars studying new product development projects argue that controls need to provide room for maneuver, or they risk stifling creativity (Dunk, 2011; Kamoche and Cunha, 2001; Van der Meer-Kooistra and Scapens, 2015). This was the lesson MG’s operating managers learned and discussed with MG’s corporate managers. After a successful experiment with an adapted budget allocation system suggested by FN’s corporate managers, MG’s corporate managers ultimately decided to change this system.
