HBS Theses and Dissertations

This dissertation provides empirical evidence from high-stakes field settings of how productivity and quality are affected by workers' deviations from prescribed processes. The first essay of the dissertation explores the role of experimentation in field settings to investigate the drivers of performance and how to implement this methodology to answer relevant operational questions rigorously. This dissertation then uses field data from proprietary sources to investigate the behavioral drivers of process variation and their effects on productivity and quality. In particular, the next two essays of this dissertation consider the effects of (1) how workers' decisions are influenced by task schedules and (2) how workers' decisions exert influence on task schedules. Many tasks are decisions, which are thus subject to human decision errors. How does scheduling affect how humans-in contrast to machines-perform these tasks? To explore this question, the second essay focuses on one critical task: quality evaluations. The accuracy of quality evaluations is critical to their being a useful input to key managerial decisions, to penalize compliance failures, and to motivate quality improvements. Yet, task-scheduling factors that are related to the workers' work structure (but unrelated to the task itself) could shape workers' predisposition toward the task and subsequent performance. We explore how inspection scheduling can affect inspection quality by influencing bias. Analyzing thousands of food safety inspections, we find that inspection results are affected by when the inspection occurs within an inspector's daily schedule and by inspectors' experience at their prior inspection of a different establishment. For example, the more compliance deterioration found in an inspector's prior inspected establishment, the more violations cited in the inspector's next inspection (of a different establishment). Consistent with negativity bias, this effect is asymmetric, applying when compliance at the inspector's prior establishment deteriorates but not when it improves. Overall, by identifying factors that bias inspections, our work contributes to the literature on monitoring, quality improvement, and scheduling. Our work also suggests a cost-effective lever: exploiting the behavioral effects of the organization of work. Task scheduling is not always a managerial decision. Those who execute tasks often have discretion over the order in which to perform them. How do these choices affect productivity and quality? The third essay of this dissertation focuses on the drivers and consequences of exercising discretion to "deviate" from a prescribed task sequence. Analyzing 2.4 million decisions, we find that radiologists prioritize similar tasks (grouping tasks into batches) and those tasks they expect to complete faster (shortest expected processing time). Exploiting random assignment of tasks to doctors' queues, instrumental variable estimates reveal that both of these types of deviations erode productivity. Actively grouping similar tasks reduces productivity, in stark contrast to productivity gains from exogenous grouping, indicating deviation costs outweigh benefits from repetition. We also find learning-by-doing in exercising discretion, with doctors deviating more often and more productively over time. Our results highlight the tradeoffs between the time required to exercise discretion and the potential gains from doing so, which has implications for managers deciding task sequence assignments and system design. Together, these essays generate new scholarly insights regarding the connections between operational factors, decision-making, and performance by analyzing data from high-stakes field settings. In doing so, this research seeks to contribute to theory while also improving management practice.

While multiple literatures suggest that emotion shapes behavior and satisfaction in experiences, there has yet to be a concerted effort to explicitly consider human emotion – whether from customer or employee -- as a source of variability in operations management (Karmarkar, 2015; Dasu and Chase, 2013). Environmental emotions, those that are endemic to the service, or emotions stimulated by operational design choices may each exert influences on service outcomes that providers and scholars alike have yet to consider. As scholarly work to arrive at an agreed upon theoretical foundation for the study of service operations has continued, open empirical questions remain about 1) the magnitude and predictability of emotion’s influence on service outcomes, 2) the opportunity to affect emotional experience through service design and 3) the role that technology plays – particularly in self-service contexts (Berry et al, 2015).
This body of work uses both laboratory and field experimentation to understand the impacts of emotional sources of variability to operational performance and to investigate the potential for more empathic service design to improve customer engagement while preserving sought-after efficiencies. Across three investigations, set in the domains of financial services and ride-sharing, I show that anxiety, whether it is directly related to the decision at hand or not, is a source of variation that exerts a costly influence on choice satisfaction and decision-making that spills over to affect service relationships. I also find that these previously ignored effects can be mitigated through relatively low-cost service design choices.

Managing human capital is one of the most critical activities for any organizations in knowledge-intensive industries. Increasingly, it requires detailed understandings on the interplay between individuals and environments as these individuals respond to various opportunities and challenges that arise outside the organizational boundary. My dissertation examines how individual innovators and knowledge workers from three different organizations respond to rapidly changing incentives, opportunities, and contexts. The first chapter focuses on user-innovators in an online community and how they respond to commercial opportunities by pursuing entrepreneurship. The second chapter studies knowledge workers in an IT company and investigates how distance from hometown affect their productivity and learning over time. The final chapter analyzes workers in a rental car company to understand how they allocate efforts in different tasks in response to increased career concerns from potential corporate restructuring.

As firms increasingly rely on data and algorithms to create value, new strategic management challenges arise which lead to revenue loss and user attrition. My dissertation examines the strategic and managerial puzzles that hinder digital firms’ value creation, from the perspectives of online community design, disintermediation, and platform-based strategies. The first chapter focuses on platform design and user segregation. The second and third chapters focus on disintermediation and its relationship to user trust and technological development. This dissertation aims to direct attention to strategic dilemmas that firms face as they embrace digital business models and provide theoretical and managerial contributions to platform studies in strategic management literature.

The use of electronic money transfer through cellular networks (“mobile money”) is rapidly increasing in the developing world. The resulting electronic currency ecosystem could improve the lives of the estimated two billion people who live outside the formal financial infrastructure by facilitating more efficient, accessible, and reliable ways to store and transfer money than are currently available. Sustaining this ecosystem requires a healthy network of agents to conduct cash-for-electronic value transactions and vice versa. This thesis consists of three chapters formatted as stand-alone papers that explore how these agents compete, how they should manage their inventories, and how they might benefit from inventory pooling. Chapter I, Service Quality and Competition: Empirical Analysis of Mobile Money Agents in Africa, explores how service quality, competition, and poverty are related to demand and inventory. Chapter II, Inventory Management for Mobile Money Agents in the Developing World, describes various approaches to modeling the agent’s inventory problem, including a simple analytical heuristic that performs well relative to actual agent decisions. Finally, Chapter III, Inventory Pooling for Mobile Money Agents in the Developing World, proposes an inventory pooling framework that has the potential to increase channel profitability by simultaneously reducing inventory requirements and increasing service levels system-wide.

This dissertation examines how the management of human capital factors impacts firm innovation. The first chapter produces causal estimates of the asset value of inventors to firms following inventors' unexpected deaths. The second chapter studies how multinational entities manage patent innovation in context of reforms to migration policy that shift the ease of business-related travel. The third chapter explores how the breadth and depth of expert reviewers' knowledge impacts their preferences for financing novel innovation in individual R&D review settings through examining the case of scientific grant review.