Coordination with Asymmetries in Highly Complex and Volatile Environments

Abstract

n the first chapter, I present an experimental analysis of weak-link coordination games in which subjects face specific time constraints in a volatile environment. In a basic 2×2 design, one factor involves situations in which all subjects face decision time (seconds to select a strategy) constraints. Another factor varies feedback time (seconds to review the outcome of the round) constraints. Overall, I find that stringent decision constraints lead to lower minimum effort levels, but increased feedback time seems to improve decision-making in anticipating others’ behaviors and aligning decisions with firm outcomes. Feedback constraints impact coordination outcomes more when decisions are made under time pressure. This finding suggests that if decisions are made under time duress, adequate recovery time can mean the difference between efficient and inefficient coordination. In the second chapter, I study the effect of heterogeneous groups on the efficiency of coordination in repeated weak-link coordination games. I develop an experimental environment to test compositional effects on organizational performance using peer effects to maximize productivity. Our experimental setup models a task environment with two different types of workers. In standard settings, subjects are given five effort choices: Ei ∈{0, 10, 20, 30, 40}. I introduce a second type of worker restricted to the highest effort choices: Ei ∈{30, 40}. I find that heterogeneous groups induce change to higher, more efficient equilibria. Despite increases in group size, firms with restricted workers report higher overall output. Several firms managed to coordinate at much higher levels, with two firms reaching the most efficient coordination level. Employees exert more effort when another employee(s) can be harmed in the firm. This paper provides a mechanism to alleviate coordination failure among large groups, demonstrating how social spillovers can overpower group-size effects. In the third chapter, we test whether participants exert more considerable effort when payoff-equivalent incentives are framed as losses rather than gains. I also intersect loss aversion and social preferences to model real-world situations when employees’ actions can cause another employee to bear a loss. I develop an experimental environment to test stake-size effects on organizational performance. Subjects are given an initial endowment at the beginning of the experiment, which allows us to assess decisions at different reference points and test the degree of loss aversion in each setting. We find significant treatment effects, such that employees in high-stake frames contribute significantly more effort than employees in low-stake frames. Loss aversion is efficiency-enhancing only in high-stake conditions. In other words, framing effects work if the stakes are high enough. We find that social preferences are not as influential in the loss domain compared to our previous findings in the gain domain. It appears that social concerns are influenced by the degree of loss framing. This paper contributes to the coordination literature on framing effects with different stakes and coordination outcomes when loss aversion intersects with social preferences. Indeed, the current COVID-19 pandemic underlines the importance of understanding how exposure to loss can shape decisions and whether this differs with social considerations when employees’ decisions can cause another to suffer the consequences.