Knowledge that Transforms

Abstract The use of preference or discriminatory auctions, where one class of bidders is offered favored treatment over another class, has received mixed attention in the literature. Research has shown there is often an economic benefit of such policies in procurement auctions thanks to lower costs for buyers as incentives are offered to disadvantaged sellers. In this paper, we study one type of preferential procurement auction: the subsidy. Using a set of controlled experiments, we compare actual bidder behavior to what is predicted in equilibrium and find consistent (but overly aggressive) patterns overall. By testing a common bid strategy assumption, we also identify a behavioral framing bias that may trap sellers in these suboptimal strategies. Finally, we compare subsidies to another common discriminatory mechanism—the price preference—and find evidence that buyers interested in increasing the welfare of disadvantaged sellers should use subsidies instead of price preference auctions, thanks to a surprising difference in outcomes between preference types. Due to the wide use of bid preference auctions to support both policy and social aims, our findings have both financial and societal implications.

Abstract This study empirically examines how induced learning through adopting a set of best practices and learning‐by‐doing improved a hospital's care of ischemic stroke patients using ad hoc teams. While previous studies in healthcare operations management conducted in ad hoc team environments predominantly focused on volume‐based learning (learning by doing, team familiarity via interactions among team members), our study focuses on induced learning in ad hoc teams through best practice adoptions. The analysis uses secondary data (Data period: January 2009–March 2017) about stroke patients from a comprehensive stroke center (CSC) in a U.S. tertiary teaching hospital as it adopted the U.S. American Heart Association/American Stroke Association (AHA/ASA) Target:Stroke best practices. The ad hoc stroke teams provide the initial care and their performance is measured using “ Door‐to‐Needle (DTN) ” time and its sub‐time segments. The DTN time is measured as the time elapsed between the stroke patient's arrival at the hospital's emergency department (ED) and the appropriate infusion of “ Tissue Plasminogen Activator (TPA) ” (i.e., a thrombolytic medication informally referred to as a “clot buster”). We found that adopting these best practices improved ischemic stroke care beyond improvement due to repetition. We also found that the neurologist's recent experience providing stroke care for the prior patient is positively associated with meeting the time performance goal for the current patient. This study provides insights into the use of management mechanisms to adopt and sustain best practices in healthcare that are generalizable to other organizations with ad hoc team environments.

Abstract We examine the impact of vampire attack, a unique platform entry strategy in the blockchain ecosystem, on the operational performance of the incumbent platform. During the vampire attack period, the entrant (attacker) clones the incumbent platform and offers tokenized incentives to entice users away from the incumbent. Prior studies offer little insight into the impact of vampire attack strategy because of its uniqueness in platform cloning, tokenized incentives, and targeted attacks. We implement a quasi‐experimental design by leveraging the first and most famous vampire attack launched by SushiSwap (the attacker) against Uniswap (the incumbent). We examine both the deposit‐side and exchange‐side impacts of the vampire attack on the operational performance of the liquidity pools on Uniswap. Surprisingly, we find that the vampire attack has no significant effect on the liquidity provision on the deposit side. Even more surprisingly, the vampire attack significantly increases the incumbent's trading volume on the exchange side. We further uncover the underlying mechanisms contributing to these intriguing results. We also demonstrate the efficacy of the novel tokenized incentives strategy. We show the generalizability of our findings by examining an alternative vampire attack event in the context of non‐fungible token marketplaces. Our study offers significant contributions to the literature on the implications of blockchain on platform operations and platform competition in operations management.

Abstract In hog farming, optimizing hog sales is a complex challenge due to uncertain factors, such as hog availability, market prices, and operating costs. This study uses a Markov Decision Process (MDP) to model these decisions, revealing the importance of the final weeks in profit management. The MDP's intractability due to the curse of dimensionality leads us to employ Deep Reinforcement Learning (DRL) for optimization. Using real‐world and synthetic data, our DRL model outperforms existing practices. However, it lacks interpretability, hindering trust and legal compliance in the food industry. To address this, we introduce “managerial learning,” extracting actionable insights from DRL outputs using classification trees that would have been difficult to obtain otherwise. We leverage these insights to devise a smart heuristic that significantly beats the heuristic currently used in practice. This study has broader implications for operations management, where DRL can solve complex dynamic optimization problems that are often intractable due to dimensionality. By applying methods, such as classification trees and DRL, one can scrutinize solutions for actionable managerial insights that can enhance existing practices with straightforward planning guidelines.

Abstract The use of part‐time employees to support operations has been a contentious topic in the literature. While part‐time employees add cost‐effective flexibility to operations, their impact on operational outcomes has largely been documented as negative. However, there are a number of sectors (e.g., non‐profit) which rely heavily on part‐time employees, with anecdotal evidence supporting their role in improving outcomes. Through this research, we seek to shed light on these contradicting perspectives. We do so by investigating the impact of the percentage of part‐time employees in the workforce dedicated to the Supplemental Nutrition Assistance Program (SNAP) outreach efforts at United States (U.S.) food banks on the effectiveness of this initiative. SNAP is the largest domestic hunger program in the U.S., assisting over 42 million individuals, and food banks play a critical role in outreach and enrollment for SNAP. We utilize data on the operational characteristics and SNAP activities of food banks that are members of the Feeding America network and U.S. Census data on the demographic characteristics of their service area. We find that an increased percentage of part‐time FTEs (full‐time equivalent) in a food bank's workforce dedicated to SNAP outreach efforts increases its effectiveness, particularly in relation to operational and contextual factors that can benefit from a more flexible workforce. Based on these findings and our review of the literature, we propose a conceptual framework on the effectiveness of part‐time employees in different settings .

Abstract Research in healthcare suggests that repeated interaction between a provider and a patient can support better decision‐making, resulting in improved efficiencies. To date, these repeated interactions enabling continuity of care have not been studied in hospital inpatient settings. During a hospital stay, decisions related to patient treatment are usually made by two key decision‐makers: the attending physician (AP) and the operating physician (OP). Under the single decision‐making approach (S‐DMA), the AP and OP are the same; in contrast, under the dual decision‐making approach (D‐DMA), the AP and OP are different. In recent years, there has been an increasing trend toward the use of D‐DMA over S‐DMA across U.S. hospitals owing to scheduling conflicts. Although research outside healthcare operations management has argued for benefits from both approaches, their impacts on a patient's hospital stay are unclear. In this study, we address this gap by investigating the effects of S‐DMA and D‐DMA on patient care outcomes in terms of patient length of stay (LOS), treatment cost, and mortality. Data for our study come from the state of Florida and involve 520,554 cardiology patients treated by 9483 APs and 18,398 OPs at 241 hospitals between 2014 and 2016. We account for both patient and physician selection issues when choosing a particular decision‐making strategy. Our results suggest that, on average, using S‐DMA is associated with reduced patient LOS and treatment cost but has no effect on mortality. We also find that S‐DMA is more beneficial for patients with low comorbidity and low process uncertainty, whereas D‐DMA is more beneficial for patients with high comorbidity and high process uncertainty. Our results are robust to alternative explanations. We demonstrate that a single decision‐maker offers benefits in the context of healthcare delivery, but dual decision‐makers may yield benefits when caring for patients with high comorbidity and high process complexity. We discuss the implications of these findings for appropriately deploying S‐DMA and D‐DMA in inpatient services.

Abstract Brand retailers are increasingly offering product personalization to accommodate the requests of individual customers. While prior research has examined product personalization from a customer's or a manufacturer's perspective, this study focuses on product personalization from a brand retailer's perspective. Under product personalization, brand retailers that outsource production to upstream suppliers face two significant challenges: product returns from downstream customers, and personalization costs charged by upstream suppliers. Employing an analysis of transaction data from an online women's wedding dress firm, this study finds that as product requests increase (i.e., more product features must be modified), personalization costs increase, but the likelihood of product returns decreases. Additionally, as time requests increase (i.e., delivery is more urgent), the likelihood of product returns increases. Furthermore, relationship‐specific process control exacerbates (i) the effect of product requests on personalization costs, and (ii) the effect of time requests on the likelihood of product returns. Finally, relationship‐specific outcome control and transaction‐specific control alleviate the impact of time requests on the likelihood of product returns while exacerbating the effect of product requests on personalization costs. The findings suggest that retailers can realize greater benefits from product personalization by selecting control mechanisms tailored to personalization requests.

Abstract Drug shortages have been persistent in the United States for over a decade, posing serious threats to public health and the healthcare system. While previous research has investigated the causes and effects of drug shortages, there is a dearth of research exploring potential solutions to mitigate this problem. Using a system dynamics model of the US generic drug market, we evaluate the long‐term effectiveness of two existing policy interventions (expediting drug approvals and nudging manufacturers to ramp up their production) and the “quality reward” initiative that is being actively explored by the FDA and industry. Our results indicate that while the existing interventions can be helpful in addressing shortages, their long‐term effect seems limited. In contrast, quality reward can mitigate drug shortages in a sustainable way. However, a caveat of quality reward is the potential emergence of a monopolistic supply market with negative consequences. We suggest that a carefully designed quality disclosure mechanism can address this issue. To the best of our knowledge, this is the first study to quantitatively and comparatively evaluate the long‐term effectiveness of quality reward and other interventions on drug shortages and provide structural explanations for their performance.

Abstract Health promotion and disease prevention requires health systems address the patients' social needs using new care delivery models. Yet, research in this area has stalled for several reasons. We study a partnership model of care that couples clinical care delivered by primary care providers and social services delivered by community‐based organizations, and its impact on patients' preventive health outcomes and behaviors. We use data from the Mid‐Ohio Farmacy, which is a collaboration across the Mid‐Ohio Food Collective (MOFC), a network of 650+ affiliated food pantries, and a large federally qualified health center (FQHC). The FQHC offers primary and preventative healthcare services across eight free clinics, which are co‐located with the MOFC‐affiliated food pantries. Patients were screened for food insecurity during their clinic visit and, if positive, were referred to the Farmacy. Compliers made at least one visit to the food pantry after referral, while noncompliers did not. Using difference‐in‐differences, we find that compliers had no discernible change in their body mass index (BMI, kg/m 2 ), which we refer to as a BMI stabilization effect. Noncompliers' BMI increased after referral. High comorbid and high pantry use compliers experienced a significant reduction in their BMI and a marginally significant reduction in glycated hemoglobin (HbA1c, %). These patients had unique compliance behaviors, including greater search, frequency, and consistency of food pantry use. Travel costs suggests that high comorbid patients ascribed a greater value to the Farmacy program. In terms of primary care utilization, we find that compliers' clinic visit patterns after referral were consistent with the visit patterns observed in the food secure cohort, suggesting that the Farmacy program may have helped compliers address competing demands that are known to inhibit health behaviors.

Abstract Millions of mom‐and‐pop nanostores dominate the grocery retail landscape in emerging markets. As nanostores are cash and storage space constrained, their suppliers tend to visit them with a high frequency, causing high operational costs. It is unclear if credit could mitigate such costs by allowing for a lower visit frequency. To investigate this, we conduct a randomized field experiment on nanostores that have not made recent use of supplier credit, with the objective to uncover how trade credits interact with the visit frequency and with the available storage space to shape shopkeepers' ordering behavior. We find that trade credits moderate the positive relationship between visit frequency and bi‐weekly order size, with credit effects being salient under low frequency visits. Storage space, by contrast, does not directionally shape shopkeepers' ordering behavior. While trade credits may more than offset the negative effect of low frequency visits among adopting nanostores, credit adoption remains challenging, with only 24% of nanostores assigned to the credit condition actually adopting the credit. Remarkably, not a single credit line was defaulted over the entire duration of the intervention. In terms of assortment, trade credit is mostly used to acquire nanostores' core assortment of popular, low‐price items with low physical volume. We contribute to the extant literature by showing that the gesture by the supplier to extend trade credit may only partly legitimize a reduction of the visit frequency.