Organizational routines are persistent sequential interaction patterns among the members of an organization. Such routines are believed to be responsible for the efficiency advantages of organizations as structures for collective decision-making. Nevertheless, to date there has been little formal modeling or evidence collected as to how organizational routines emerge, endure and adapt to change. This project proposes an agenda for studying the emergence, maintenance, adaptability and optimality of organizational routines that combines game theoretic analysis with experimental evidence. The baseline model is a novel but simple "harvesting game" wherein teams of agents repeatedly harvest a grid over several periods with full information about all past choices. The team's goal is to uncover as many novel outcomes on the grid as possible in the time allowed but without the ability to communicate with one another. The absence of communication proxies for specialization by team members that makes communication difficult or for settings where communication is impractical, for example during an emergency response. In such environments there are easily identifiable optimal routines and this study examines both theoretically and experimentally which routines emerge, how long it takes them to emerge, whether they are stable once they emerge, and if agents make short-run sacrifices in the interest of promoting the emergence of optimal routines.
This research will build upon and extend the baseline model by studying routine formation in variants where 1) there is no information about past choices, 2) the complexity of the decision task is increased, 3) more players are added, 4) some players have private, payoff-relevant information that may naturally position them as team leaders and 5) players have the ability to communicate with one another prior to making their harvesting decisions. The new framework will broaden understanding of how members of organizations learn to form routines in a variety of different environments that are all amenable to theoretical and experimental evaluation. Thus will bring game-theoretic analysis and rigor to the organizational routine literature which to date has been studied using the methods of psychology and organizational behavior researchers. The game-theoretic framework allows one to assess the extent and source of deviations in routine formation from the rational choice ideal so that there is a better understanding of the role played by bounded rationality in routine formation. In addition, data on organizational routines are difficult to measure or observe -- the experimental tests proposed using the methods of experimental economics, where subjects earn payments based on their own choices and the choices of other members of their same team, provides an excellent framework in which to study organizational routines. The experimental designs, programs, data and findings resulting from this proposal will be made freely available so that other researchers can build upon the new framework and instructors can teach this new approach to organizational routines in the classroom.
