Knowledge workers are often assigned creative tasks that consist of an originality goal and a usefulness or feasibility requirement or constraint. While it is often difficult to create ideas that are both original and useful, it is unclear how varying usefulness constraints affect the usefulness and originality of a knowledge worker's creative output. We conduct a lab experiment that asks participants to create images using a set of drawing materials and emoji and pays them based on the originality of their image on condition that it meets a certain usefulness constraint. We measure an image's originality based on how different it is to other submitted images in the experiment and its usefulness based on how recognizable it is to a set of inexperienced raters. Our experimental treatments vary the usefulness constraint from no usefulness constraint in treatment T0 to low, medium, and high usefulness constraints in treatments T10, T40, and T80. Our results show that, as expected, the T10 treatment results in more useful and less original images compared to T0 with a similar originality-usefulness trade-off. Surprisingly, our T40 treatment results in images with the same recognizability as T10, but with worse originality and a worse originality-usefulness trade-off. Finally, we find that participants in the T80 treatment effectively ignore the high usefulness constraint and submit images with similar image recognizability, originality, and originality-recognizability trade-off as the T10 treatment. Our paper suggests that low usefulness constraints can effectively "nudge" employees to factor in usefulness in their creative output, but that high constraints can cause employees to ignore the constraint or to factor it in unsuccessfully. We show that managers, in such cases, may benefit from "artificially" lowering the usefulness constraint they set for their employees or making the task goal more focused on usefulness.

Problem definition: This paper studies how teams make operational decisions in two canonical settings: standalone Newsvendor inventory decisions (tactical decision-making) and Newsvendor under information sharing (strategic decision-making).

Academic / Practical Relevance: Team decision-making, while being prevalent in practice, has been rarely studied in operational contexts. Past research in behavioral economics and psychology uses simple, abstract decision tasks and show that teams generally outperform individuals. However, it is less clear whether the results extend to practical yet more complex operational settings. Answering this will help researchers and practitioners understand how teams will behave in operational settings and, subsequently, whether team decision-making should be employed.

Methodology: We employ behavioral and experimental studies, including text chat analysis to explore teams' decision-making mechanism.

Results: We find that whether teams will perform differently from individuals depends critically on the team decision mechanism, as reflected in what team members find compelling in intra-team chats. When playing the role of the supplier, both in the standalone Newsvendor setting and the information sharing game, teams find multiple (potentially conflicting) arguments compelling which pull team decisions in different directions, and overall, the teams do not outperform individuals. By contrast, as the retailer in the information sharing game, the compelling argument points to the direction of self-interest and team retailers outperform individual retailers in terms of earning profits.
Managerial Implications: Our findings suggest that when companies consider whether to implement team decision-making, it is important to think carefully about whether and which compelling argument(s) will emerge during team discussions, and whether they promote the outcome the company wants. Careful behavioral studies are useful to study this, and chat analysis proves to be a powerful tool in studying teams' decision mechanism. 

The desire to uphold promise-keeping norms greases the wheels of interaction by creating trust. Norms establish a set of mutual expectations which parties rely on to interact in the presence of uncertainty and renegotiation. We present a model of social norm compliance in a risky trust game. We establish a set of assumptions about the norm that characterize how promises affect the norm to fulfill an agreement, how the norm is changed once unforeseen contingencies are resolved and is changed if a renegotiation request is accepted or rejected. Using these assumptions, the model makes predictions about behavior patterns in the risky trust game. We conduct an experiment to test predictions both of the behavior patterns and the assumptions of the model. We show that behavior is consistent with the norms model and that our assumptions about the norms are supported. Using this model, we explain why most subjects make promises, why promises are largely fulfilled even when it is costly, how renegotiation success or failure affects the propensity to fulfill the promise and why nearly half of subjects do not request costless renegotiation even if it is available. This work sheds light on the impact of norms to influence renegotiation and extends the promise-keeping literature. For policies written against the backdrop of strong norms, we address implications and guidelines.

Many Operations Management (OM) models assume that people act as forward-looking optimizers in

dynamic environments. We experimentally examine this assumption. To cover a wide range of settings we

look at several common classes of dynamic decision problems, and characterize behavior as either optimal, or

as consistent with one of several non-optimal policies. At a high level, our results suggest that behaviors are

not uniform and depend on the features of the problem. Specifically we find that: (1) Decisions are generally

forward-looking (though not always optimal) in Technology Adoption and Capacity Allocation problems,

but not in Search/Stopping problems. (2) The optimal policy is a good, but not the best, representation of

behavior in both Technology Adoption and in Capacity Allocation; in both tasks simpler forward-looking

heuristics achieve a better fit. (3) Performance (payoff) is correlated within-subject for different dynamic

tasks but the specific policy usage may vary even within-subject. Together, these results provide new microfoundations

for researchers interested in building more descriptive models of dynamic behavior.