Research

Trade credit insurance (TCI) is a risk management tool commonly used by suppliers to guarantee against payment default by credit buyers. TCI contracts can be either cancelable (the insurer has the discretion to cancel this guarantee during the insured period) or noncancelable (the terms cannot be renegotiated within the insured period). This paper identifies two roles of TCI: the (cash flow) smoothing role (smoothing the supplier’s cash flows) and the monitoring role (tracking the buyer’s continued creditworthiness after contracting, which enables the supplier to make efficient operational decisions regarding whether to ship goods to the credit buyer). We further explore which contracts better facilitate these two roles of TCI by modeling the strategic interaction between the insurer and the supplier. Noncancelable contracts rely on the deductible to implement both roles, which may result in a conflict: a high deductible inhibits the smoothing role, whereas a low deductible weakens the monitoring role. Under cancelable contracts, the insurer’s cancelation action ensures that the information acquired is reflected in the supplier’s shipping decision. Thus, the insurer has adequate incentives to perform its monitoring function without resorting to a high deductible. Despite this advantage, we find that the insurer may exercise the cancelation option too aggressively; this thereby restores a preference for noncancelable contracts, especially when the supplier’s outside option is unattractive and the insurer’s monitoring cost is low. Noncancelable contracts are also relatively more attractive when the acquired information is verifiable than when it is unverifiable.

Trade credit is an important source of short-term financing and an integrated part in supply contracts. Although a number of theories have been proposed on how trade credit could improve supply chain efficiency, causal studies on the impact of trade credit on operational decisions are scarce. In this study, we examine the impact of trade credit on inventory decisions using an empirical strategy that leverages (i) an exogenous shock imparted by the French government’s intervention to impose a ceiling on trade credit duration, (ii) a triple difference-in-differences identification strategy, and (iii) synthetic controls (SCs). By considering the 60-day ceiling coverage and SC construction requirements, we identify four French retail sectors as our main sample. Among them, in the postregulation period, the hardware retail sector firms on average exhibited a significant 16% decline in trade credit usage. Correspondingly, these firms also displayed a significant 11% decline in inventory level. In the remaining three sectors, we found mixed results in the main sample. All the four sectors, however, show consistent support for a causal link between trade credit and inventory in a subsample compiled using a stringent 90-day ceiling criterion. Collectively, our findings offer direct evidence that trade credit is an indispensable financing source for inventory procurement. Finally, in the postregulation period, the hardware retailers exhibited a 15.5% decline in revenue and 3.5% reduction in gross profit. This cautions policy makers that regulations limiting the use of trade credit may have unintended consequences on downstream firms, and may harm overall supply chain efficiency.

Large-scale disruptions have exposed vulnerabilities in global supply chains, particularly around the domestic production of essential goods. However, we lack systematic evidence on how government policies and infrastructure affect suppliers' ability to pivot to producing new products during crises. Understanding how government structures and policies influence supplier entry and contributes to supply chain resilience by encouraging domestic production capacity during emergencies. Using data from North America's largest online platform for supplier sourcing, we analyze supplier responses to demand for COVID-19 products during the pandemic and study how fiscal support, procurement practices, and regulatory adjustments affect suppliers' likelihood of offering new products. Leveraging heterogeneous, exogenous exposures to COVID-19, along with instrumental variables for COVID-19-induced demand, we estimate moderating effects of government on entry decisions. Firms in states with more fiscal support or higher government agency engagement show greater responsiveness to demand, particularly among firms with prior government contracting experience. States with decentralized procurement practices and better IT infrastructure see increased supplier offerings, as well as those that reduced licensing requirements and implemented supportive procurement measures. A one standard deviation increase in demand is associated with a 60% increase in the baseline probability of entry in supportive policy environments. Policymakers may consider (1) decentralizing procurement processes to enhance flexibility, (2) investing in government IT infrastructure to reduce transaction costs, (3) implementing targeted fiscal support during crises, and (4) reducing regulatory barriers to entry to build more resilient domestic supply chains for essential products during future emergencies.

At the onset of the COVID-19 pandemic, hospitals were in dire need of data‐driven analytics to provide support for critical, expensive, and complex decisions. Yet, the majority of analytics being developed were targeted at state- and national-level policy decisions, with little availability of actionable information to support tactical and operational decision-making and execution at the hospital level. To fill this gap, we developed a multi-method framework leveraging a parsimonious design philosophy that allows for rapid deployment of high-impact predictive and prescriptive analytics in a time-sensitive, dynamic, data-limited environment, such as a novel pandemic. The product of this research is a workload prediction and decision support tool to provide mission-critical, actionable information for individual hospitals. Our framework forecasts time-varying patient workload and demand for critical resources by integrating disease progression models, tailored to data availability during different stages of the pandemic, with a stochastic network model of patient movements among units within individual hospitals. Both components employ adaptive tuning to account for hospital-dependent, time-varying parameters that provide consistently accurate predictions by dynamically learning the impact of latent changes in system dynamics. Our decision support system is designed to be portable and easily implementable across hospital data systems for expeditious expansion and deployment. This work was contextually grounded in close collaboration with IU Health, the largest health system in Indiana, which has 18 hospitals serving over one million residents. Our initial prototype was implemented in April 2020 and has supported managerial decisions, from the operational to the strategic, across multiple functionalities at IU Health.

U.S. health system reform simultaneously promotes alternative payment models (APMs) and provider organizations through integration, which is assumed to better enable providers to coordinate and manage the risks introduced by APMs. While the interaction of these two changes has implications for both APM design and regulation of provider integration, it is not well studied. In this paper, we offer empirical evidence on the role of horizontal and vertical integration of orthopaedic surgeons in driving heterogeneity in the impact of the Comprehensive Joint Replacement (CJR) APM. Using a quasi-experimental generalized difference-in-differences approach, we find that, relative to similar control hospitals, CJR hospitals with fully horizontally and vertically integrated orthopaedic surgeons receive a 5.83% increase in CMS payments to hospitals but do not differ in CJR quality measures. Further analysis shows that the increase in payments stems from changes in physicians' clinical decision-making that reduce penalties (increase CMS payments) to hospitals for low-value care and reduces CMS payments to PAC providers. Consequently, these hospitals also receive greater CJR reward payments. Overall, our findings suggest that integrated providers are likely to perform better under APMs, providing novel insights for policymakers, payors, and healthcare providers as they evaluate and implement value-based reforms.

This paper examines the impact of the Hospital Readmissions Reduction Program (HRRP) on hospitals' admission decisions. HRRP provides incentives for hospitals to reduce readmissions by imposing penalties to hospitals that exhibited higher than average 30-day risk-adjusted readmission rates. Indeed, after HRRP was introduced in 2012, readmission rates have fallen by 1.2% on average across monitored conditions. In this paper, we show that part of this reduction may be due to hospitals admitting more patients for observation as opposed to inpatient admission. Under this classification, patients receive hospital-level care but do not count towards admissions for HRRP purposes. To show that this is the case, we combine four hospital-level datasets. We exploit variation in hospitals' financial exposure to HRRP due to: i) readmission performance relative to national average; and ii) financial constraints. We find that hospitals exposed to HRRP penalties increased observation bed usage by 16.9% more compared to non-penalized hospitals, and by as much as 40.6% if they were also financially constrained. A back-of-the-envelope calculation suggests that if just 10% of the change in observation admissions was done to avoid readmissions, the observed reduction in readmissions following HRRP would be 21% smaller than reported. Furthermore, if hospitals are indeed using observation admissions to avoid readmissions then one would expect that post-HRRP readmissions of patients originally discharged from the same hospital would have decreased at a higher rate compared to readmissions to hospitals other than the discharging hospital. Using a detailed patient-level dataset, we confirm that this is indeed the case.

Problem Definition: While more than 90% of medications are industrially produced and distributed through retail pharmacies, when patients require personalized medications or medications that are otherwise unavailable, e.g., due to patients’ allergies or drug shortages, retail pharmacies must rely on compounding pharmacies that produce medications to order. Inspired by discussions with the management team of a compounding pharmacy, this project aims to reduce production delays for patients requiring compounded medications through the development of a dynamic production control algorithm.

Methodology/results: We propose and evaluate a novel and theory-driven implementable heuristic policy based on a fluid model to capture production dynamics, paired with an infinite-horizon State-Dependent Riccati Equation controller, lightly modified for use as a forward value function approximation. Numerical studies suggest that for large systems this heuristic is within 10% of a theoretical lower bound on waiting cost per unit time within a fluid setting. Within a data-driven simulation, a hybrid policy combining an age-limit with the proposed heuristic can reduce the average cost of waiting time by almost 30% compared to the status-quo oldest request sequencing policy without increasing worst case delays.

Managerial Implications: This contributes a novel approach to challenging multi-product production management problems that can be generalized to other settings where optimal policies are intractable.

Since 2015, global progress in improving energy efficiency (EE) has been lagging behind the targets set by the UN Sustainable Development Goals, in part due to the behavioral barriers to improving EE. The rising renewable energy (RE) adoption may impact EE improvement by raising or lowering the barriers to EE. The well-documented RE rebound effect may hinder EE improvement, while RE adoption may also heighten overall awareness of energy usage, thereby driving EE improvement. This paper examines whether and how RE adoption influences EE improvement. We leverage data from 183 sites of an multinational industrial conglomerate from 2015 to 2020 to estimate the impact of changes in RE usage and acquisition methods (e.g., sourcing off-site or generating on-site) on EE improvement. We find that using RE to satisfy an additional 10% of a site’s energy demand caused an additional 2.8% improvement in EE. However, the impact varies significantly depending on the acquisition approach. While sourcing off-site RE led to EE improvement, installing on-site RE generation had either no or a negative effect. To understand the mechanism behind these findings, we surveyed site managers and conducted further analysis by leveraging variations in on-site generation costs. Our results indicate that the rebound effect prevailed when sites adopted RE with low ongoing cost. For corporations expanding RE usage, we offer evidence of additional EE benefits, but capitalizing on these benefits requires careful consideration of the RE acquisition approach. For policymakers, we provide guidance on prioritizing certain approaches to accelerate the progress toward net-zero.

Firms often elicit judgemental adjustments to an algorithm-generated demand forecast. This process aims to utilize humans’ private information that is inaccessible to the algorithm. However, humans are vulnerable to systematic biases when making such adjustments. Thus, it is challenging to integrate these adjustments with algorithm-generated forecasts to improve forecast accuracy (measured as the absolute deviation from realized sales). We propose a novel classification-based solution to address this challenge. First, we predict an adjustment’s quality (likelihood to improve forecast accuracy) using predictors of humans’ private information advantage (e.g., product and store characteristics) and systematic biases (e.g., recent forecast errors and adjustment characteristics). Next, we apply a simple heuristic -- based on the predicted quality -- to classify (accept/reject) each adjustment integration in the final forecast. We collaborate with a European retailer to test our approach using a large dataset (~ 1.1mn transactions) of algorithm-generated forecasts. Humans adjusted 38% of these forecasts, and nearly 51% of these adjustments improved forecast accuracy. In out-of-sample tests, we benchmark our approach against the strategy of using historical evidence of humans’ private information advantage to determine when to always accept or reject adjustments at the product-store level. We find that our approach improves benchmark forecast accuracy by 12%. Moreover, a substantial share (46%) of this improvement is associated with the inclusion of predictors of systematic biases. Our paper expands practitioners' toolkit for managing judgemental adjustments by showcasing an intuitive and easy-to-implement adjustment-level integration solution.

Volunteers play a pivotal role in the daily operations of nonprofit organizations (NPOs). However, retaining volunteers remains a critical challenge for NPOs due to declining volunteering rates and high turnover. We explore a cost-effective approach to improving volunteer retention by leveraging the potential of experienced volunteers. Using data from 49,378 volunteers at a food bank over a five-year period, we explore how interactions between experienced and new volunteers affect the latter's retention. Our findings indicate that the presence of experienced volunteers in a shift increases new volunteer retention by 52%, with a substantially stronger effect (125% increase) for group volunteers. However, we find that increased familiarity among experienced volunteers negates this positive effect. This negative effect depends on the uncertainty inherent in the tasks undertaken by the volunteers. Based on our findings, we propose a strategic reassignment policy for NPO managers. This approach involves willing experienced volunteers to be reassigned to concurrent shifts that lack experienced volunteer presence. With volunteer consent, this reassignment policy could mitigate the potential negative effects of strong social bonds among experienced volunteers, while improving new volunteer retention by 14% on shifts that initially lacked an experienced volunteer at conservative estimates.

We study how assessment team composition affects the implementation and effectiveness of continuous process improvement programs in manufacturing. Focusing on World Class Manufacturing (WCM) assessments, we analyze the influence of team members' language proficiency and experience on assessment scoring and subsequent manufacturing performance. Utilizing data from 224 WCM assessments across 98 manufacturing sites, we explore the impacts of assessment team experience (industry, product category, and prior site visits) and language proficiency (languages spoken and language alignment) on assessment outcomes. Our analysis reveals that assessment teams proficient in multiple languages assign higher managerial and technical scores. Experience contributes differently, with overall professional experience positively affecting technical scores, while prior site and product category experience show varied effects. Despite standardized assessment scoring, mediation analysis finds that scores do not fully mediate the effect of team characteristics on manufacturing performance, indicating that language proficiency directly influences manufacturing performance and suggesting that nuanced feedback is key to improvement. Our findings highlight the critical role of assessment team selection and training in enhancing operational performance. For multinational companies, the study underlines the importance of communication skills in assessment teams.

A key operations management choice involves the variety of tasks a manufacturing plant performs, i.e., its scope. We examine how scope changes affect production-line performance within a plant. Past research presents conflicting arguments on the influence of scope on operational performance, and empirical studies on scope typically study the effects of differences across operating units, instead of the effects of changes in scope over time within operating units. Leveraging over 3,400 scope changes across 158 plants of a global FMCG firm, we employ a dynamic panel system GMM estimator with fixed effects and instrumental variables. We find strong evidence that broadening scope initially improves production line performance, subject to diminishing returns, reconciling these opposing views. Crucially, this effect is significantly moderated by process heterogeneity: the benefits of scope diminish substantially, and may even reverse, when lines differ technologically. While capacity utilization and volume heterogeneity also play a role, their impact is less consistent in our main specifications. We find primarily positive effects of increasing scope across the typical operating range, with little evidence that broad scope negatively influences line performance for most plants. Findings from mechanism analyses suggest knowledge transfer, facilitated by process similarity, drives the observed economies of scope. Our results indicate that while complexity eventually tempers scope's gains, well-managed operations can reap substantial performance benefits by incrementally broadening scope, particularly when strategically managing the technological heterogeneity of their production lines.