This paper considers pricing and remanufacturing strategy of a firm that decides to offer both new and remanufactured versions of its product in the market and is concerned with demand cannibalization. We present a model of demand cannibalization and a behavioral study that estimates a key modeling parameter: a fraction of consumers who switch from new to remanufactured product. As we show, this fraction has an inverted‐U shape, and, thus, the underlying consumer behavior cannot be modeled using the standard methodologies that rely on consumers' willingness to pay (WTP). We find that by incorporating the inverted‐U‐shaped consumer behavior, the firm remanufactures under broader conditions, charges a much lower price, and typically remanufactures more units—leading to an increase of profits from remanufacturing by up to a factor of two as compared with making decisions based on the WTP only. Lastly, we find that the behavior of the low‐price market segment plays an important role because the firm reacts to it differently than the WTP‐based logic would suggest.

We study several important aspects of using environmental taxes to motivate the choice of innovative and “green" emissions‐reducing technologies as well as the role of fixed cost subsidies and consumer rebates in this process. In our model, a profit‐maximizing monopolistic firm facing price‐dependent demand selects emissions control technology, production quantity, and price in response to the tax, subsidy, and rebate levels set by the regulator. The available technologies vary in environmental efficiency as well as in the fixed and variable costs. Both the optimal policy for the firm and the social‐welfare maximizing policy for the regulator are analyzed. We find that the firm's reaction to an increase in taxes may be non‐monotone: while an initial increase in taxes may motivate a switch to a greener technology, further tax increases may motivate a reverse switch. For the regulator, we compare the social welfare achievable in the centralized system (which serves as an upper bound) to the highest level achievable under different classes of environmental policies. If the regulator is limited to a tax‐only policy, then when the regulator is moderately concerned with environmental impacts, the tax level that maximizes social welfare simultaneously motivates the choice of clean technology and closes the gap to the upper bound; however, both low and high levels of societal environmental concerns may lead to the choice of dirty technology and significant welfare losses as compared to the centralized case. Supplementing the environmental taxation with fixed cost subsidies and consumer rebates can eliminate this effect, expanding the range of parameters over which the green technology is chosen and often closing the welfare gap to the centralized solution.

We study the interaction between the design of a premium-status loyalty program, revenue management, and strategic consumer behavior. Specifically, we consider a contemporaneous change where firms across several industries switch their loyalty programs from quantity-based toward spending-based designs. This change has been met with fierce opposition from the media and consumers. Building on the microfoundations of strategic, forward-looking, and status-seeking consumer behavior, we endogenize strategic consumer response to firms’ pricing and loyalty program design decisions, and we characterize conditions under which, by coordinating these decisions, firms can benefit from strategic consumer behavior. We further show that by switching to a spending-based design, firms can benefit from strategic behavior even more, under broader conditions, and in a Pareto-improving way. Finally, we also analyze combined designs, which utilize a combination of quantity and/or spending requirements, and show how they can be used to better manage the transition toward spending-based designs, possibly minimizing negative consumer reactions. This paper was accepted by Serguei Netessine, operations management.

This paper discusses the interaction between revenue management and customer relationship management for a firm that operates in a customer retention situation but faces limited capacity. We present a dynamic programming model for how the firm balances investments in customer acquisition and retention, as well as retention across multiple customer types. We characterize the optimal policy and discuss how the policy changes depending on capacity limitations. We then contrast the modeling results with those of a behavioral experiment in which subjects acted as managers making acquisition and retention decisions. In the modeling part of the paper, we introduce a concept of the value of an incremental customer (VIC), and show that when capacity is unlimited, VIC equals customer lifetime value (CLV), but when capacity is limited, VIC is much smaller and changes dynamically depending on the number of customers and their mix. As a result, the optimal spending is constant and depends on CLV for the firms with unlimited capacity, but changes dynamically and is generally unrelated to CLV when capacity is limited. In the experimental part, we introduce a concept of conditional optimality for the analysis of state-dependent decisions. Applying this concept to our data, we document a number of decision biases, specifically the subjects' tendency to overspend on retaining high-value customers and underspend on lower-value customers retention and acquisition. We show that providing CLV information exacerbates these biases and leads to a loss of net revenue when capacity is limited, but providing information about the marginal costs of acquisition and retention eliminated these biases and increases net revenue. Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2013.1842 . This paper was accepted by Yossi Aviv, operations management.

Problem definition: We use a realistically large, publicly available data set from a global fintech lender to simulate the impact of different antidiscrimination laws and their corresponding data management and model-building regimes on gender-based discrimination in the nonmortgage fintech lending setting. Academic/practical relevance: Our paper extends the conceptual understanding of model-based discrimination from computer science to a realistic context that simulates the situations faced by fintech lenders in practice, where advanced machine learning (ML) techniques are used with high-dimensional, feature-rich, highly multicollinear data. We provide technically and legally permissible approaches for firms to reduce discrimination across different antidiscrimination regimes whilst managing profitability. Methodology: We train statistical and ML models on a large and realistically rich publicly available data set to simulate different antidiscrimination regimes and measure their impact on model quality and firm profitability. We use ML explainability techniques to understand the drivers of ML discrimination. Results: We find that regimes that prohibit the use of gender (like those in the United States) substantially increase discrimination and slightly decrease firm profitability. We observe that ML models are less discriminatory, of better predictive quality, and more profitable compared with traditional statistical models like logistic regression. Unlike omitted variable bias—which drives discrimination in statistical models—ML discrimination is driven by changes in the model training procedure, including feature engineering and feature selection, when gender is excluded. We observe that down sampling the training data to rebalance gender, gender-aware hyperparameter selection, and up sampling the training data to rebalance gender all reduce discrimination, with varying trade-offs in predictive quality and firm profitability. Probabilistic gender proxy modeling (imputing applicant gender) further reduces discrimination with negligible impact on predictive quality and a slight increase in firm profitability. Managerial implications: A rethink is required of the antidiscrimination laws, specifically with respect to the collection and use of protected attributes for ML models. Firms should be able to collect protected attributes to, at minimum, measure discrimination and ideally, take steps to reduce it. Increased data access should come with greater accountability for firms. History: This paper has been accepted for the Manufacturing & Service Operations Management Special Section on Responsible Research in Operations Management. Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2022.1108 .

Recent advances in information technology, advanced manufacturing (robotics, 3D printing, etc.), and logistics have allowed firms to customize their products to the specifications of individual consumers, who, in turn, prefer these products to standard ones. In the unlikely event that customized products do not match expectations, however, consumers often feel entitled to a return. Should firms offer returns on customized products? We examine this question via a Stackelberg game model, in which the firm (leader) decides the prices and returns policies for its customized and standard products; consumers (followers) decide which product to buy, given the initial noisy valuations and, upon experiencing the product, whether to return it. Both parties act strategically: Forward-looking consumers incorporate the real option value of possible returns into their initial purchasing decisions, and the firm incorporates consumers’ best purchase and return response into its pricing and returns policy decisions. Our model produces three key insights. First, firms can use customized products to induce some consumers who otherwise would buy and return a standard product to switch to lower-return-rate customized products. Second, it may be optimal to offer returns on customized products, despite their lower salvage value. Third, firms can increase profits and reduce (total) returns by offering returnable customized products. This paper was accepted by Duncan Simester, marketing.