Search

The growing complexity of consulting, new product development, and information technology projects has led firms to increasingly adopt the strategy of collaborative value cocreation with their vendors. Although such value cocreation often involves a client firm engaging with more than one vendor, research in this field has primarily focused on single-vendor cocreation, whereas research on multivendor sourcing has rarely considered collaborative value cocreation. We bridge this gap by studying a cocreation environment involving a client, a primary vendor, and a potential secondary vendor. Examining the trade-offs that arise in the strategic interactions among the three firms, we derive novel insights about how these interactions differ from those under one-vendor cocreation, and from those under traditional multivendor sourcing. Specifically, we demonstrate that the client can sometimes be better off by adding a secondary vendor that is less efficient than the primary one; and a more expensive primary vendor can sometimes make the client less inclined to add the secondary vendor. The client’s decision to add the secondary vendor is driven by the value per unit of output rather than purely cost or efficiency parameters. We show that in such collaborative environments, the primary vendor benefits from the addition of a secondary vendor whenever it benefits the client. The conceptual and managerial insights drawn from our research contribute to a better theoretical and practical understanding of strategic decision making by firms in multivendor cocreation environments that are increasingly becoming prevalent. This paper was accepted by Jayashankar Swaminathan, operations management. Supplemental Material: Data and the e-companion are available at https://doi.org/10.1287/mnsc.2022.4490 .

We study direct load control contracts that utilities use to curtail customers’ electricity consumption during peak-load periods. These contracts place limits on the number of calls and total number of hours of power reduction per customer per year as well as the duration of each call. The stochastic dynamic program that determines how many customers to call and the timing and duration of each call for each day is an extremely difficult (NP-hard) optimization problem. We design a scenario-based approximation method to generate probabilistic allocation polices in a reasonable amount of time. Our approach consists of three approximations: deterministic approximation of demand, discretization of the expected demand, and aggregation/disaggregation of the resources. We show the relative information error resulting from the deterministic approximation is [Formula: see text], the discretization error is [Formula: see text], and the aggregation/disaggregation error is [Formula: see text], where n represents the length of the horizon. Finally, we show the total relative error is [Formula: see text]. Our error analysis establishes that our approximation method is near optimal. In addition, our extensive numerical experiments verify the high quality of our approximation approach. The error, conservatively measured, is quite small and has an average and standard deviation of 8.6% and 1.4%, respectively. We apply our solution approach to the data provided by three major utility companies in California. Overall, our study shows our procedure improves the savings in energy-generation cost by 37.7% relative to current practices. This paper was accepted by Chung Piaw Teo, optimization. Funding: The authors thank the UCLA Ziman Center’s Howard and Irene Levine Program in Housing and Social Responsibility and the Morrison Family Center for Marketing Studies and Data Analytics for generous funding. Supplemental Material: The online appendices are available at https://doi.org/10.1287/mnsc.2022.4493 .

We study the effect of proximity to corporate headquarters on the productivity of inventors and research and development (R&D) facilities. Distant inventors and R&D facilities are less productive, and plausibly exogenous reductions in the travel time from these inventors or facilities to headquarters increase their productivity and creativity. We hypothesize that these improvements in the management of innovation production occur because proximity improves monitoring, managerial ability to provide direction, relationship building that supports creativity, and information exchange, including advice from headquarters. Consequently, our results suggest that proximity can help managers balance both exploration and exploitation when overseeing innovation. Naturally, these results then beg the question of why firms do not locate all inventors and R&D facilities in close proximity to headquarters. We find that distant inventors and R&D facilities are located in areas with better general economic indicators, and especially more favorable tax rates, than headquarters, suggesting firms balance these benefits against the benefits of proximity when locating innovation production. This paper was accepted by Brian Bushee, accounting. Supplemental Material: Data and the internet appendix are available at https://doi.org/10.1287/mnsc.2022.4469 .

We consider a principal who periodically offers a fixed and costly nonmonetary reward to agents to incentivize them to invest effort over the long run. An agent’s output, as a function of his effort, is a priori uncertain and is worth a fixed per-unit value to the principal. The principal’s goal is to design an attractive reward policy that specifies how the rewards are to be given to an agent over time based on that agent’s past performance. This problem, which we denote by [Formula: see text], is motivated by practical examples from both academia (e.g., a reduced teaching load) and industry (e.g., “Supplier of the Year” awards). The following “limited-term” (LT) reward policy structure has been quite popular in practice. The principal evaluates each agent periodically; if an agent’s performance over a certain (limited) number of periods in the immediate past exceeds a predefined threshold, then the principal rewards him for a certain (limited) number of periods in the immediate future. When agents’ outputs are deterministic in their efforts, we show that there always exists an optimal policy that is an LT policy and also, obtain such a policy. When agents’ outputs are stochastic, we show that the class of LT policies may not contain any optimal policy of problem [Formula: see text] but is guaranteed to contain policies that are arbitrarily near optimal. Given any [Formula: see text], we show how to obtain an LT policy whose performance is within ϵ of that of an optimal policy. This guarantee depends crucially on the use of sufficiently long histories of the agents’ outputs. We also analyze LT policies with short histories and derive structural insights on the role played by (i) the length of the available history and (ii) the variability in the random variable governing an agent’s output. We show that the average performance of these policies is within 5% of the optimum, justifying their popularity in practice. We then introduce and analyze the class of “score-based” reward policies; we show that this class is guaranteed to contain an optimal policy and also, obtain such a policy. Finally, we analyze a generalization in which the principal has a limited number for rewards in any given period and show that the class of score-based policies, with modifications to accommodate the limited availability of the rewards, continues to contain an optimal solution for the principal. This paper was accepted by Jeannette Song, operations management. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4482 .

In this paper, we consider the yet-uncharted assortment optimization problem under the exponomial choice model, in which the objective is to determine the revenue-maximizing set of products that should be offered to customers. Our main algorithmic contribution comes in the form of a fully polynomial-time approximation scheme, showing that the optimal expected revenue can be efficiently approached within any degree of accuracy. We synthesize several ideas related to approximate dynamic programming, intended to construct a compact discretization of the continuous state space by keeping track of “key statistics” in rounded form and by operating with a suitable bit precision complexity. We complement this result by a number of NP-hardness reductions to natural extensions of this problem. Moreover, we conduct empirical and computational evaluations of the exponomial choice model and our solution method. Focusing on choice models with a simple parametric structure, we provide new empirical evidence that the exponomial choice model can achieve higher predictive accuracy than the multinomial logit (MNL) choice model on several real-world data sets. We uncover that this predictive performance correlates with certain characteristics of the choice instance—namely, the entropy and magnitude of choice probabilities. Finally, we leverage fully ranked preference data to simulate the expected revenue of optimal assortments prescribed using the fitted exponomial and MNL models. On semisynthetic data, the exponomial-based approach can lift revenues by 3%–4% on average against the corresponding MNL benchmark. This paper was accepted by Chung Piaw Teo, optimization. Funding: The research of Danny Segev on this project was supported by the Israel Science Foundation [Grants 148/10 and 1407/20]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4492 .

We examine the relationship between pre-grant patent disclosure and analyst forecast accuracy. We take advantage of the passage of the American Inventor’s Protection Act (1999), which mandates the pre-grant public disclosure of all information in patent application documents within 18 months of the initial filings. We find that, on average, the pre-grant patent disclosure of corporate inventions significantly improves the accuracy of analyst forecasts about the patenting firm and this improvement is greater for firms with higher research and development intensity. Nevertheless, improvements in the accuracy of analysts’ forecasts are smaller when firms issue more original and scientifically broader patents. Also, this effect is weaker for firms in states without legal protection for trade secrets. This paper was accepted by Brian Bushee, accounting. Funding: Ali Mohammadi gratefully acknowledges support from the Danish Finance Institute. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4420 .

We examine the incentive effects of subjectivity in allocating tournament-based rewards and punishments. We use data from a company where reward and punishment decisions are based on a combination of objective metrics and subjective performance assessments. Rankings based on the objective metrics and the ultimate payoff allocations are disclosed to all members of the organization. This information allows employees to observe whether and how managers subjectively override the objective rankings. Consistent with expectancy theory, we predict and find that subjective rewards and punishments manifesting as favorable (unfavorable) deviations from formula-based payoff expectations are associated with subsequent performance improvements (declines). These performance responses are incremental to the effects of receiving a reward or punishment per se. Our results suggest that managers can benefit from using subjective rewards, but using subjective punishments can be very costly in the absence of sufficiently strong ex ante incentive effects associated with the prospect of subjective penalties. Our findings contribute to the literature on subjectivity in performance evaluations and have important practical implications for designing incentive systems. This paper was accepted by Brian Bushee, accounting. Funding: The authors appreciate the Harvard Business School for financial support during the development of this study. Supplemental Material: Data are available at https://doi.org/10.1287/mnsc.2022.4501 .

Decision makers often ask experts to forecast a future state. Experts, however, can be biased. In the economics and psychology literature, one extensively studied behavioral bias is called herding. Under strong levels of herding, disclosure of public information may lower forecasting accuracy. This result, however, is derived only for point forecasts. In this paper, we consider experts’ probabilistic forecasts under herding, find a closed-form expression for the first two moments of a unique equilibrium forecast, and show that the experts report too similar locations and inflate the variance of their forecasts because of herding. Furthermore, we show that the negative externality of public information no longer holds. In addition to reacting to new information as expected, probabilistic forecasts contain more information about the experts’ full beliefs and interpersonal structure. This facilitates model estimation. To this end, we consider a one-shot setting with one forecast per expert and show that our model is identifiable up to an infinite number of solutions based on point forecasts but up to two solutions based on probabilistic forecasts. We then provide a Bayesian estimation procedure for these two solutions and apply it to economic forecasting data collected by the European Central Bank and the Federal Reserve Bank of Philadelphia. We find that, on average, the experts invest around 19% of their efforts into making similar forecasts. The level of herding shows an increasing trend from 1999 to 2007 but drops sharply during the financial crisis of 2007–2009 and then rises again until 2019. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Supplemental Material: The electronic companion and data are available at https://doi.org/10.1287/mnsc.2022.4487 .

This paper studies the investor’s investment decision in a standard financial contracting model. The investor chooses among entrepreneurs with heterogeneous production technologies who generate different probability distributions of cash flow under effort. I provide a complete characterization of optimal contracts, agency rents, and the investor’s investment decision. Differences in entrepreneurs’ production technologies imply differences in optimal contracts and agency rents across equally productive entrepreneurs, which biases and potentially distorts the investor’s investment decision. The results provide a complete characterization of entrepreneurs’ financial constraints in the presence of heterogeneous production technologies and uncover a fundamental link between production technologies, financial contracts, and financial constraints. This paper was accepted by Gustavo Manso, finance. Funding: This work was supported by the Swedish House of Finance.

Using cross-state and intertemporal variation in whether a state’s minimum wage is bound by the federal minimum wage, we provide evidence that minimum wage increases lead U.S. public firms in minimum-wage-sensitive industries (i.e., retail, restaurant, and entertainment) to cut capital expenditures. These effects are concentrated one to two years after the law goes into effect. Prior to the minimum wage increase, investment trends are similar across minimum-wage-sensitive firms in bound versus unbound states, and we find little evidence that minimum wage changes affect U.S. public firm investment outside of these industries. This paper was accepted by Victoria Ivashina, finance. Supplemental Material: The internet appendix is available at https://doi.org/10.1287/mnsc.2022.4430 .

Exaptation refers to the emergence of novel functionalities in existing products. Exaptations frequently arise in the context of users who creatively modify (or hack) existing products to accommodate new needs. Here, we examine how product-first (compared with problem-first) search affects the occurrence of exaptations. In a product-first search, the user identifies the product to be hacked before seeking a viable need. In a problem-first search, the user has a defined problem before seeking a viable solution. We argue that users are more likely to achieve exaptations following a problem-first (compared with a product-first) search. Indeed, with problem-first search, they are less likely to face functional fixedness, and they can leverage their greater awareness of problems that may not have readily adaptable solutions, which leads them to generate exaptations. Using a novel data set comprising user hacks of IKEA products, we present evidence that hacks originating from a product-first search are less likely to generate exaptations than hacks originating from a problem-first search. We also show that this difference is mitigated when the user has hacking experience or when the IKEA product being hacked is more modular. We also explore how the mitigation happens. Increased hacking experience appears to reduce functional fixedness; meanwhile, increased product modularity increases the likelihood that users will make serendipitous discoveries leading to exaptations. We contribute to the growing literature on exaptation as a source of novelty and discuss the implications of this phenomenon for managing user innovation. This paper was accepted by Sridhar Tayur, entrepreneurship and innovation. Funding: Financial support from the Singapore Ministry of Education Academic Research Fund [Tier 1 Grant 251RES1907] is gratefully acknowledged. Supplemental Material: The data and online appendix are available at https://doi.org/10.1287/mnsc.2022.4486 .

We provide evidence that investor behavior is sensitive to two dimensions of subjective uncertainty concerning future asset values. Investors vary in the extent to which they attribute market uncertainty to (1) missing knowledge, skill, or information (epistemic uncertainty) and (2) chance or stochastic processes (aleatory uncertainty). Investors who view stock market uncertainty as higher in epistemicness (knowability) are more likely to reduce uncertainty by seeking guidance from experts and are more responsive more responsive to available information when choosing whether to invest. In contrast, investors who view stock market uncertainty as higher in aleatoriness (randomness) are more likely to reduce uncertainty through diversification, and their risk preferences better predict whether they choose to invest. We show, further, that attributions of uncertainty can be perturbed by the format in which historical information is presented: charts displaying absolute stock prices promote perceptions of epistemicness and greater willingness to pay for financial advice, whereas charts displaying the change in stock prices from one period to the next promote perceptions of aleatoriness and a greater tendency to diversify. This paper was accepted by Yuval Rottenstreich, behavioral economics and decision analysis. Funding: This material is based upon work supported by the National Science Foundation [Grant SES-1427469] to C. R. Fox and Gülden Ülkümen. Additional support for this work was provided by INSEAD and UCLA Anderson School of Management. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2022.4489 .

This paper shows that investments based on deep learning signals extract profitability from difficult-to-arbitrage stocks and during high limits-to-arbitrage market states. In particular, excluding microcaps, distressed stocks, or episodes of high market volatility considerably attenuates profitability. Machine learning-based performance further deteriorates in the presence of reasonable trading costs because of high turnover and extreme positions in the tangency portfolio implied by the pricing kernel. Despite their opaque nature, machine learning methods successfully identify mispriced stocks consistent with most anomalies. Beyond economic restrictions, deep learning signals are profitable in long positions and recent years and command low downside risk. This paper was accepted by Kay Giesecke, finance. Funding: D. Avramov acknowledges the Israel Science Foundation (Grant 288/18) for financial support. S. Cheng acknowledges the General Research Fund of the Research Grants Council of Hong Kong [Project 14502318] for financial support. Supplemental Material: The data files and online appendix are available at https://doi.org/10.1287/mnsc.2022.4449 .

This paper seeks to demonstrate the impact of the gig economy on product quality in seemingly unrelated local industries through the labor market. Our empirical context is the quality of service for restaurants in the city of Austin, and we examine how they were impacted by the exogenous exit and reentry of rideshare platforms, Uber and Lyft, because of regulatory changes. We leverage these exogenous shocks and combine them with sentiment-analyzed data from Yelp reviews that capture how customers assess the quality of service at each restaurant. We show that, compared with control cities, customers in Austin become more negative about service quality when Uber and Lyft are present in the city. Additionally, we use rich data on employee turnover and wages to demonstrate that service staff turnover increases in Austin when Uber and Lyft are present compared with the control cities. We also conduct several additional studies and robustness checks that are all congruent with our hypothesis that Uber and Lyft lower the quality of service in Austin restaurants by raising their staff turnover. Together, these results suggest significant ramifications of the gig economy on the broader industries through the labor market. This paper was This paper was accepted by Matt Shum, marketing. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2022.4481 .

Subjects observe a private signal and make an initial guess; they then observe their neighbors’ guesses, update their own guess, and so forth. We study learning dynamics in three large-scale networks capturing features of real-world social networks: Erdös–Rényi, Stochastic Block (reflecting network homophily), and Royal Family (that accommodates both highly connected celebrities and local interactions). We find that the Royal Family network is more likely to sustain incorrect consensus and that the Stochastic Block network is more likely to persist with diverse beliefs. These patterns are consistent with the predictions of DeGroot updating. It lends support to the notion that the use of simple heuristics in information aggregation is prevalent in large and complex networks. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Funding: The authors thank the Keynes Fund (University of Cambridge), the Creative-Pioneering Researchers Program (Seoul National University), and C-BID (NYUAD) for financial support. Supplemental Material: The data files and e-companion are available at https://doi.org/10.1287/mnsc.2023.4680 .

In this research we reexamine and update prior research on the association between decision conflict and choice deferral. Decision conflict has featured prominently in psychology literature, going back to the seminal work of Lewin (1933) , Miller (1944) , and Janis and Mann (1977) . Building on this earlier work and the assumption of conflict as an inhibitor of choice, Tversky and Shafir (1992 ; see also Dhar 1997 ) argued that the tendency to defer a decision can increase when decision makers experience conflict because of the need to make trade-offs between options in a set. Tversky and Shafir (1992) showed that decision makers may defer choice more when faced with two options presenting trade-offs compared with when they are offered either a single option (see also Dhar 1997 ) or two options where one alternative dominates the other. Relatedly, Luce (1998) showed that decision makers may be more likely to defer choice when faced with multiple options presenting relatively difficult (versus easy) trade-offs. We reexamine published data and report the results of 40 well-powered direct and conceptual replications of these effects (total n = 26,703) that cast doubt on the replicability of prior research. Our data suggest that additional research is needed in order to reveal a robust association between experimental manipulations of decision conflict and choice deferral. We conclude our investigation by delineating a set of directions for future research on this topic. This paper was accepted by Yuval Rottenstreich, behavioral economics and decision analysis. Supplemental Material: The data and e-companion are available at https://doi.org/10.1287/mnsc.2022.4484 .

This paper exploits a major food safety crisis to estimate a full demand model for the unsafe product and its substitutes, recovering consumers’ preference parameters for different product characteristics. Counterfactual exercises quantify the relevance of different mechanisms driving consumers’ responses, such as changes in safety perceptions, idiosyncratic taste, product characteristics, and price. We find that consumers’ reactions are limited by their preferences for the product’s observable and unobservable characteristics. Because of the costs associated with switching from the affected product, the decline in demand following a product-harm crisis tends to understate the true weight of such events in consumers’ utility. We find that unobservable taste is a crucial driver of consumers’ responses. Our counterfactual exercises illustrate that the demand would have declined further if consumers had had access to a closer substitute. For an accurate assessment of product-harm crises, managerial strategies should therefore account for how different demand drivers bind consumers’ substitution patterns. This paper was accepted by Matthew Shum, marketing. Funding: This work was supported by the Ministerio de Ciencia e Innovación [Grants ECO2008-01116 and ECO2010-15052], the Ministerio de Economía y Competitividad [Grant ECO2013-43011-P], and the Deutsche Forschungsgemeinschaft [Grant CRC TR 224]. R. Ferrer received support from [Grant ECO2016-76998-P], and the Severo Ochoa Programme for Centers of Excellence in R&D [SEV-2015-0563].

Could a salary history ban (SHB) reduce the gender wage gap? Proponents of this intervention believe the gap is sustained by the practice of eliciting salary histories from job applicants. Although observational studies suggest that SHB operates as envisioned, two features complicate the interpretation of its effects. These are, respectively, the passage of relevant legislation alongside SHB, and the presence of public campaigns that propel SHB into law. We assessed SHB in the United Kingdom, where neither potential confound was present, and did not find evidence that the intervention operated as intended. An intention-to-treat analysis of a 16-month field experiment, conducted with 230 staff hires at a private educational institution, indicates that SHB was about as likely to harm new hires as it was to help them. Additional analyses did not reveal significant differences between women and men. We supplement these results with an interrupted time series analysis of 3,687 placements made by a recruitment firm that voluntarily adopted SHB for its job candidates. Salaries were significantly lower under SHB, but were not significantly different for women versus men. Taken together, our results suggest that SHB was ineffective in isolation from contemporaneous legislative changes, proequality messaging, or some combination thereof. This paper was accepted by Olav Sorenson, organizations. Supplemental Material: Data are available at https://doi.org/10.1287/mnsc.2022.4658 .

This paper examines an innovation in capital formation that has spurred contentious debate: the Umbrella Partnership Corporation (“Up-C”) IPO. Advisors and underwriters argue that the Up-C deal structure is a driver of post-IPO value and, thus, is a value-enhancing means of raising capital that may be one solution to concerns regarding the drop in the number of publicly traded companies. Consistent with these claims, recent research suggests that organizing soon-to-be public businesses as pass-through entities (as is the case for Up-Cs) leads to superior future performance. Yet, broadening the analysis to consider abnormal stock performance and post-IPO litigation of a larger and more recent sample of exclusively Up-C IPOs, we conclude just the opposite. While the Up-C deal structure increases IPO valuations and predicts positive post-IPO operating performance, the return performance of Up-C IPOs indicates that Up-C deals harm public shareholders. Further, despite their superior earnings performance, Up-C IPOs face a significantly higher rate of post-IPO litigation as compared with non-Up-C IPOs. Because IPO investors seemingly do not anticipate the myriad ways in which the Up-C deal structure might facilitate opportunism by pre-IPO owners, they frequently turn to litigation as an ex-post settling-up mechanism. Consequently, our paper offers the first empirical evidence of downsides associated with the Up-C deal structure for public shareholders and, in so doing, affords the rarity of having academic evidence lead (as opposed to respond to) a controversial debate. This paper was This paper was accepted by Victoria Ivashina, finance. Funding: The authors gratefully acknowledge the financial support of the NYU Stern School of Business, the J. Reuben Clark Law School, and the Marriott School of Business. Supplemental Material: Data are available at https://doi.org/10.1287/mnsc.2022.4454 .

Negative advertising provides information about the weaknesses of a competitor’s product. We study negative advertising with a focus on how its regulation impacts product positioning for profit-maximizing firms. We build a model of informative advertising competition, where product positioning is endogenous, and consumers have rational expectations. We show that despite the informational benefits of negative advertising, permitting it (as the Federal Trade Commission in the United States does) may lead to reduced product differentiation and lower consumer welfare, even in markets where firms do not use negative advertising in equilibrium. We then extend our model to political competition, where a candidate’s objective is to obtain a larger share of votes than the competitor. We show that political competition supports higher positional differentiation, along with more negative advertising than product competition, in line with observed high use of negative advertising in political races and their rarer use in product competition. This paper was accepted by Dmitri Kuksov, marketing. Funding: P. Yildirim thanks Mack Institute and the Dean’s Research Funds of the Wharton School for their generous funding of this study. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4439 .

Assortment optimization concerns the problem of selling items with fixed prices to a buyer who will purchase at most one. Typically, retailers select a subset of items, corresponding to an "assortment" of brands to carry, and make each selected item available for purchase at its brand-recommended price. Despite the tremendous importance in practice, the best method for selling these fixed-price items is not well understood, as retailers have begun experimenting with making certain items available only through a lottery. In this paper we analyze the maximum possible revenue that can be earned in this setting, given that the buyer's preference is private but drawn from a known distribution. In particular, we introduce a Bayesian mechanism design problem where the buyer has a random ranking over fixed-price items and an outside option, and the seller optimizes a (randomized) allocation of up to one item. We show that allocations corresponding to assortments are suboptimal in general, but under many commonly-studied Bayesian priors for buyer rankings such as the MNL and Markov Chain choice models, assortments are in fact optimal. Therefore, this large literature on assortment optimization has much greater significance than appreciated before -- it is not only computing optimal assortments; it is computing the economic limit of the seller's revenue for these fixed-price substitute items. We derive several further results -- a more general sufficient condition for assortments being optimal that captures choice models beyond Markov Chain, a proof that Nested Logit choice models cannot be captured by Markov Chain but can be partially captured by our condition, and suboptimality gaps for assortments when our condition does not hold. Finally, we show that our mechanism design problem provides the tightest-known LP relaxation for assortment optimization under the ranking distribution model., Full version of EC2022 paper. Forthcoming in Management Science

This paper analyzes how the degree of task-related team diversity influences the team outcome. The focus is on a novel benefit of team diversity that does not depend on complementary skills or mutual learning. We consider a multiagent setting where the employer makes use of career competition to motivate the team. Our analysis reveals incentive effects that do not exist in the well-studied case of teams of size 2. In teams with more than two members, diversity may have a positive effect on the efforts of some members although diversity leads to unbalanced career competition. As a consequence, expected profits are higher under a moderate degree of diversity compared with a situation with completely homogeneous agents. This result shows that the employer prefers a positive degree of diversity for pure incentive reasons. This paper was accepted by Suraj Srinivasan, accounting. Funding: This work was supported by the Deutsche Forschungsgemeinschaft [Grants EXC 2126/1–390838866 and EXC-2047/1–390685813].

It is generally acknowledged that sunk cost bias leads to suboptimal decisions, such as escalation of commitment. Some researchers, however, suggest that sunk cost bias can be beneficial when consumers have self-control problems. In this paper we explore the case when consumers with sunk cost bias have time-inconsistent preferences and, therefore, suffer from self-control problems. We experimentally demonstrate that sunk costs can make subjects better off by inducing higher effort. We then develop an analytical model to explore the implications of sunk cost bias for firm’s pricing strategy. We find that, in the presence of sunk cost bias, higher prices can lead to higher experienced quality. We show that sunk cost bias can sometimes improve firm’s profits, lead to lower prices, and increase welfare. Our results suggest that, when consumers use a product for multiple periods, pricing policies such as 0% financing, which are often viewed as exploitative, can instead lead to lower total prices, higher profits, and higher welfare. This paper was accepted by Matthew Shum, marketing. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2022.4479 .

Facial information is essential in daily life, but relatively little is known about whether seeing a face improves people’s decision quality. This experimental paper studies the loan-approval decisions based on the historical cash-loan data with real repayment outcomes and exogenously varies whether and how a borrower’s facial information is provided. We find that facial information does not improve subjects’ decisions, despite the fact that it can predict repayment behavior in a machine-learning algorithm. This is because subjects have various biases in evaluating facial photos, and they rely excessively on facial information in making the loan-approval decisions. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Funding: J. Meng received financial support from the National Natural Science Foundation of China [Grants 71103003, 71471004, and 71822301]. Supplemental Material: Data files and the online appendix are available at https://doi.org/10.1287/mnsc.2022.4436 .

Massive online text reviews can be a powerful market research tool for understanding consumer experiences and helping firms improve and innovate. This research exploits the rich semantic properties of text reviews and proposes a novel machine learning modeling framework that can reliably and efficiently extract consumer opinions and uncover potential interaction effects across these opinions, thereby identifying hidden and nuanced areas for product and service improvement beyond existing modeling approaches in this domain. In particular, we develop an opinion extraction and effect estimation framework that allows for uncovering customer opinions’ average effects and their interaction effects. Interactions among opinions can be synergistic when the co-occurrence of two opinions yields an effect greater than the sum of two parts, or as what we call dysergistic, when the co-occurrence of two opinions results in dampened effect. We apply the model in the context of large-scale customer ratings and text reviews for hotels and demonstrate our framework’s ability to screen synergy and dysergy effects among opinions. Our model also flexibly and efficiently accommodates a large number of opinions, which provides insights into rare yet potentially important opinions. The model can guide managers to prioritize joint areas of product and service improvement and innovation by uncovering the most prominent synergistic pairs. Model comparison with extant machine learning approaches demonstrates our improved predictive ability and managerial insights. This paper was accepted by Gui Liberali, marketing. Funding: The authors acknowledge the support of research funding from the National Natural Science Foundation of China [Grant 72072173]. Supplemental Material: The data files and online appendix are available at https://doi.org/10.1287/mnsc.2022.4443 .

Although leveraged exchange-traded funds (ETFs) are popular products for retail investors, how to hedge them poses a great challenge to financial institutions. We develop an optimal rebalancing (hedging) model for leveraged ETFs in a comprehensive setting, including overnight market closure and market frictions. The model allows for an analytical optimal rebalancing strategy. The result extends the principle of “aiming in front of target” introduced by Gârleanu and Pedersen (2013) from a constant weight between current and future positions to a time-varying weight because the rebalancing performance is monitored only at discrete time points, but the rebalancing takes place continuously. Empirical findings and implications for the weekend effect and the intraday trading volume are also presented. This paper was accepted by Agostino Capponi, finance. Funding: M. Dai acknowledges support from the National Natural Science Foundation of China [Grant 12071333], the Hong Kong Polytechnic University [Grant P0039114], and the Singapore Ministry of Education [Grants R-146-000-243/306/311-114 and R-703-000-032-112]. H. M. Soner acknowledges partial support from the National Science Foundation [Grant DMS 2106462]. C. Yang acknowledges support from the Chinese University of Hong Kong [Grant 4055132 and a University Startup Grant]. Supplemental Material: Data and the online supplement are available at https://doi.org/10.1287/mnsc.2022.4407 .

In-store promotions are a highly effective marketing tool that can have a significant impact on revenue. In this research, we study the question of dynamic promotion planning in the face of Bounded-Memory Peak-End demand models. In order to determine promotion strategies, we establish that a High-Low pricing policy is optimal under diagonal dominance conditions (so that the current period price dominates both past period price effects and competitive product price effects on the demand), as well as conditions on the price dispersion. We show that finding the optimal High-Low dynamic promotion policy is NP-hard in the strong sense. Nevertheless, for the special case of promotion planning for a single item, we propose a compact Dynamic Programming (DP) approach that can find the optimal promotion plan that follows a High-Low policy in polynomial time. When the diagonal dominance conditions do not hold, and, hence, a High-Low policy is not necessarily optimal, we show that the optimal High-Low policy that is found by our proposed DP can find a provably near-optimal solution. Using the proposed DP as a subroutine, for the case of multiple items, we propose a Polynomial-Time-Approximation Scheme (PTAS) that can find a solution that can capture at least [Formula: see text] of the optimal revenue and runs in time that is exponential only in [Formula: see text]. Finally, we test our approach on data from large retailers and demonstrate an average of [Formula: see text] increase in revenue relative to the retailer’s current practices. This paper was accepted by Chung Piaw Teo, optimization. Funding: Financial support from the Oracle Corporation [External Research Office grant] and the National Science Foundation [Grant CMMI-1162034] is gratefully acknowledged.

Candidate self-funding, in particular self-loans, is a significant source of funding of political campaigns. Self-funding clusters among nonincumbent campaigns, republican campaigns, and more expensive campaigns. Self-funded campaigns raise less money from individuals and special interests and also spend less. Self-funders are wealthier on average and run in more competitive elections. The analysis of self-funders’ legislative decisions shows that self-funders’ votes, especially those of republicans, are significantly more sensitive to contributions from special interests that are affected by the votes. The results highlight the importance of considering politicians’ self-funding choices in analyzing voting behavior and the value of political activism. This paper was accepted by Gustavo Manso, finance. Supplemental Material: Data and the online appendices are available at https://doi.org/10.1287/mnsc.2022.4404 .

To shed light on the factors that affect who speaks up in teams in the workplace, we study willingness to speak up after someone has raised an opinion. We call voicing disagreement overriding and study this behavior in a laboratory experiment where participants answer multiple choice questions in pairs. In a control treatment, participants interact anonymously. In a photo treatment, both participants see the photo of the person they are matched with at the beginning of the group task. Using a series of incentivized tasks, we elicit beliefs about the likelihood that each possible answer option to a question is correct. This allows us to measure disagreement and to tease apart the role of disagreement versus preferences in the decision to override ideas in teams. Results show that anonymity increases overriding. This treatment effect is driven by social image costs. Analysis of heterogeneity in behavior by gender reveals no differences between the likelihood that men and women override. However, we find some evidence that men and women are treated differently; when participants disagree with their partner, they are more likely to override a woman than a man. Preferences seem to in part explain the differential treatment of men and women. Studying group performance, we find that overriding helps groups on average, while the gender composition of teams does not affect team performance. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Funding: This work was supported by the University of Melbourne and the Australian Research Council [Grant DE190100585]. J. Guo received a Kinsman Studentship from the University of Melbourne to conduct this research. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.4434 .

This paper studies a pricing problem for a single-server queue where customers arrive according to a Poisson process. For each arriving customer, the service provider announces a price rate and system wait time. In response, the customer decides whether to join the queue, and, if so, the duration of the service time. The objective is to maximize either the long-run average revenue or social welfare. We formulate this problem as a continuous-time control model whose optimality conditions involve solving a set of delay differential equations. We develop an innovative method to obtain the optimal control policy, whose structure reveals interesting insights. The optimal dynamic price rate policy is not monotone in the wait time. That is, in addition to the congestion effect (the optimal price rate increases in the wait time), we find a compensation effect, meaning that the service provider should lower the price rate when the wait time is longer than a threshold. Compared with the prevalent static pricing policy, our optimal dynamic pricing policy improves the objective value through admission control, which, in turn, increases the utilization of the server. In a numerical study, we find that our revenue-maximizing pricing policy outperforms the best static pricing policy, especially when the arrival rate is low, and customers are impatient. Interestingly, the revenue-maximizing policy also improves social welfare over the static pricing policy in most of the tested cases. We extend our model to consider nonlinear pricing and heterogeneous customers. Nonlinear pricing may improve the revenue significantly, although linear pricing is easier to implement. For the hetergeneous customer case, we obtain similar policy insights as our base model. This paper was accepted by Baris Ata, stochastic models and simulation. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2022.4474 .

One of the most commanding powers of the U.S. Patent and Trademark Office (USPTO) is to compel inventions into secrecy, withholding patent rights and prohibiting disclosure, to prevent technology from leaking to foreign competitors. This paper studies the impacts of compulsory secrecy on firm invention and the wider innovation system. In World War II, USPTO issued secrecy orders to more than 11,000 patent applications, which it rescinded en masse at the end of the war. Compulsory secrecy caused implicated firms to shift their patenting away from treated classes, with effects persisting through at least 1960. It also restricted commercialization and impeded follow-on innovation. Yet it appears it was effective at keeping sensitive technology out of public view. The results provide insight into the effectiveness of compulsory secrecy as a regulatory strategy and into the roles, and impacts, of formal intellectual property in the innovation system. This paper was accepted by Toby Stuart, entrepreneurship and innovation. Funding: This work was supported by Harvard Business School (Division of Research and Faculty Development) and the National Bureau of Economic Research (Innovation Policy Grant). This material is based upon work supported by the National Science Foundation [Grant 1951470]. Supplemental Material: Data and the online appendices are available at https://doi.org/10.1287/mnsc.2022.4457 .

We investigate the causal effect of performance pay and conversations about performance in 224 stores of a retail chain implementing a field experiment with a 2x2 factorial design. In the performance pay treatments, managers receive a bonus, which is a simple linear function of the profits achieved above a threshold value. In the performance review treatments, managers have to report their activities undertaken to increase profits in regular meetings. We find that whereas performance pay did not yield significant profit increases, performance review conversations increased profits by about 7%. However, when additionally receiving performance pay, the positive effect of performance reviews vanished. We provide evidence from surveys and meeting protocols that performance pay changes the nature of conversations, leading to a stronger self-reliance of store managers, which undermines the value of the performance reviews. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Funding: Funding from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany’s Excellence Strategy [Grant EXC 2126/1 390838866] is gratefully acknowledged. Supplemental Material: Data and the online appendices are available at https://doi.org/10.1287/mnsc.2022.4431 .

Combinatorial auctions have found widespread application for allocating multiple items in the presence of complex bidder preferences. The enumerative exclusive OR (XOR) bid language is the de facto standard bid language for spectrum auctions and other applications, despite the difficulties, in larger auctions, of enumerating all the relevant packages or solving the resulting NP-hard winner determination problem. We introduce the flexible use and efficient licensing (FUEL) bid language, which was proposed for radio spectrum auctions to ease both communications and computations compared with XOR-based auctions. We model the resulting allocation problem as an integer program, discuss computational complexity, and conduct an extensive set of computational experiments, showing that the winner determination problem of the FUEL bid language can be solved reliably for large realistic-sized problem instances in less than half an hour on average. In contrast, auctions with an XOR bid language quickly become intractable even for much smaller problem sizes. We compare a sealed-bid FUEL auction to a sealed-bid auction with an XOR bid language and to a simultaneous clock auction. The sealed-bid auction with an XOR bid language incurs significant welfare losses because of the missing bids problem and computational hardness, the simultaneous clock auction leads to a substantially lower efficiency than FUEL because of the exposure problem. This paper was accepted by Axel Ockenfels, behavioral economics and decision analysis. Funding: This work was supported by Deutsche Forschungsgemeinschaft [Grant BI 1057-1/8]. P. Milgrom gratefully acknowledges support from the U.S. National Science Foundation [Grant SES-1947514]. M. Bichler and G. Schwarz was supported by the German Research Foundation [Grants BI 1057 I-9 and 277991500]. Supplemental Material: The data files are available at https://doi.org/10.1287/mnsc.2022.4465 .

We develop an approach to solve the Barberis casino gambling model [Barberis N (2012) A model of casino gambling. Management Sci. 58(1):35–51] in which a gambler whose preferences are specified by the cumulative prospect theory (CPT) must decide when to stop gambling by a prescribed deadline. We assume that the gambler can assist their decision using independent randomization. The problem is inherently time inconsistent because of the probability weighting in CPT, and we study both precommitted and naïve stopping strategies. We turn the original problem into a computationally tractable mathematical program from which we devise an algorithm to compute optimal precommitted rules that are randomized and Markovian. The analytical treatment enables us to confirm the economic insights of Barberis for much longer time horizons and to make additional predictions regarding a gambler’s behavior, including that, with randomization, a gambler may enter the casino even when allowed to play only once and that it is prevalent that a naïf never stops loss. This paper was accepted by Kay Giesecke, finance. Funding: The first author acknowledges support from National Natural Science Foundation of China [Grant 11901494] and Natural Science Foundation of Guangdong Province [Grant 2019A1515011396]. The second author gratefully acknowledges support from the European Union’s Seventh Framework Programme/European Research Council [ERC Starting Grant RobustFinMath 335421]. The third author gratefully acknowledges financial support through start-up grants at both the University of Oxford and Columbia University, as well as through the Oxford–Nie Laboratory for Financial Big Data and the Nie Center for Intelligent Asset Management. Supplemental Material: Data are available at https://doi.org/10.1287/mnsc.2022.4414 .

In this paper, we study the learning problem in contextual search, which is motivated by applications such as crowdsourcing and personalized medicine experiments. In particular, for a sequence of arriving context vectors, with each context associated with an underlying value, the decision maker either makes a query at a certain point or skips the context. The decision maker will only observe the binary feedback on the relationship between the query point and the value associated with the context. We study a probably approximately correct learning setting, where the goal is to learn the underlying mean value function in context with a minimum number of queries. To address this challenge, we propose a trisection search approach combined with a margin-based active learning method. We show that the algorithm only needs to make [Formula: see text] queries to achieve an ε-estimation accuracy. This sample complexity significantly reduces the required sample complexity in the passive setting where neither sample skipping nor query selection is allowed, which is at least [Formula: see text]. This paper was accepted by J. George Shanthikumar, data science. Funding: X. Chen and Q. Liu were supported by the National Science Foundation [Grant IIS-1845444].

To maximize productivity, manufacturers must organize and equip their workforces to efficiently handle variable workloads. Their success depends on their ability to assign experienced and skilled workers to specialized tasks and coordinate work on production lines. Worker turnover may disrupt such efforts. We use staffing, productivity, and pay data from within a major consumer electronics manufacturer’s supply chain to study how firms should manage worker turnover and its effects using production decisions, wages, and inventory. We find that worker turnover impedes coordination between assembly line coworkers by weakening knowledge sharing and relationships. Publicly available unit-cost estimates imply that worker turnover accounts for $206–274 million in added direct expenses alone from defectively assembled units failing the firm’s stringent quality control. To evaluate managerial alternatives, we structurally estimate a dynamic equilibrium model (an Experience-Based Equilibrium) encompassing (1) workers’ endogenous turnover decisions and (2) the firm’s weekly planning of its production scheduling and staffing in response. In counterfactual analyses, a less turnover-prone, hence more productive, workforce significantly benefits the firm, reducing its variable production costs by 4.5%, or an estimated $928 million for the studied product. Such benefits justify paying higher efficiency wages even to less skilled workforces; furthermore, interestingly, rational inventory management policies incentivize self-interested firms to reduce rather than tolerate turnover. This paper was accepted by Vishal Gaur, operations management. Funding: The authors gratefully acknowledge support from the data sponsor, and K. Moon gratefully acknowledges support from the Wharton Dean’s Research Fund and the Claude Marion Endowed Faculty Scholar Award of the Wharton School. Supplemental Material: The online appendices are available at https://doi.org/10.1287/mnsc.2022.4476 .

The timely flow of financial information is critical for efficient capital market functioning, yet we have little understanding of firms’ and auditors’ collective abilities to maintain timely financial reporting when under duress. We use COVID as a stress test case to examine whether reporting systems can withstand systemic increases in complex economic events and coordination challenges. Despite COVID-related challenges persisting through 2020 and beyond, we document surprisingly modest average delays in financial reports during COVID and only in Q1-2020. Reporting timeliness reverts to pre-COVID levels no later than Q2-2020. We find no evidence of meaningful declines in actual reporting quality during COVID, but we do find some evidence consistent with declines in perceived reporting quality. Overall, our findings indicate that current financial reporting processes are remarkably robust and provide insights about financial reporting more broadly. In particular, given that nearly all firms were able to weather the unprecedented disruptions caused by COVID, our findings imply that most material reporting delays observed outside of COVID are likely a result of either a firm’s strategic choices or exceptionally fragile reporting processes. This paper was accepted by Ranjani Krishnan, accounting. Supplemental Material: The data and online appendix are available at https://doi.org/10.1287/mnsc.2023.4670 .

The design of performance-based incentives—commonly used in online labor platforms—can be naturally posed as a moral hazard principal-agent problem. In this setting, a key input to the principal’s optimal contracting problem is the agent’s production function: the dependence of agent output on effort. Although agent production is classically assumed to be known to the principal, this is unlikely to be the case in practice. Motivated by the design of performance-based incentives, we present a method for estimating a principal-agent model from data on incentive contracts and associated outcomes, with a focus on estimating agent production. The proposed estimator is statistically consistent and can be expressed as a mathematical program. To circumvent computational challenges with solving the estimation problem exactly, we approximate it as an integer program, which we solve through a column generation algorithm that uses hypothesis tests to select variables. We show that our approximation scheme and solution technique both preserve the estimator’s consistency and combine to dramatically reduce the computational time required to obtain sound estimates. To demonstrate our method, we conducted an experiment on a crowdwork platform (Amazon Mechanical Turk) by randomly assigning incentive contracts with varying pay rates among a pool of workers completing the same task. We present numerical results illustrating how our estimator combined with experimentation can shed light on the efficacy of performance-based incentives. This paper was accepted by Chung Piaw Teo, optimization. Supplemental Material: The data files and e-companion are available at https://doi.org/10.1287/mnsc.2022.4450 .

We examine the effect of plant managers on productivity using unique matched manager-plant panel data on U.S. auto-assembly plants during 1993–2007. Our econometric approach is two-pronged. Our first approach relies on using the panel nature of our data to measure variation in productivity due to managerial influence. We estimate the interquartile range of the effect of individual plant managers on average hours-per-vehicle to be about 30%. Further, we find that plant managers’ experience with the models that are in production ameliorates the negative impact of new model introductions on productivity. We also observe evidence that managers’ plant-specific tenure has a positive impact on productivity. In our second approach, we use high-frequency time-series data, along with structural-break tests and machine-learning methodologies, to predict variation in production using plant-manager switches. We find that a plant manager’s identity is predictive of changes in both the mean and variance of production, further highlighting their channels of managerial influence. These findings are robust to narrowing the sample to focus on retirements as an exogenous source of managerial switches. This paper was accepted by Serguei Netessine, operations management. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.4427 .

In order to thrive, organizations need to build and maintain an ability to meet unexpected external challenges. Yet many organizations are sluggish: their capabilities can only undergo incremental changes over time. What are the stochastic processes governing “routinely occurring” challenges that best prepare a sluggish organization for unexpected challenges? We address this question with a stylized principal-agent model. The “agent” represents a sluggish organization that can only change its capability by one unit at a time, and the “principal” represents the organization’s head or its competitive environment. The principal commits ex ante to a Markov process over challenge levels. We characterize the process that maximizes long-run capability for both myopic and arbitrarily patient agents. We show how stochastic, time-varying challenges dramatically improve a sluggish organization’s preparedness for sudden challenges. This paper was accepted by Joshua Gans, business strategy. Funding: This work was supported by the Foerder Institute for Economic Research.

We consider a mixed duopoly service system with two service providers (SPs): one private and the other public. The public SP’s objective is welfare maximization, whereas the private SP is profit driven. Customers are heterogeneous in terms of their quality tastes and choose from three options, namely, joining the private queue, joining the public queue, and balking (or taking an outside option). We first consider the scenario where the private SP provides a premium service and the public SP provides a regular service. Paradoxically, we find that although the public SP’s objective is welfare maximization, welfare in the system can be less than that obtained with only profit-seeking private SPs. We further demonstrate that the maximum social welfare is achieved by partially privatizing the public SP, that is, by including both welfare and profit maximization as arguments in its objective function. In extreme cases in which the public SP’s capacity is very small, fully privatizing the public SP can be socially desirable. We then consider the alternate setting where the public SP provides a premium service and the private SP provides a regular service and obtain similar conclusions. These findings, however, rely on the assumption that an outside option exists for balking customers, and one should be cautious in generalizing to no-balking situations. This paper was accepted by Jayashankar Swaminathan, operations management. Funding: W. Zhou was supported by the National Natural Science Foundation of China [Grants 71925002 and 71731006]. W. Huang acknowledges the financial support from the National Natural Science Foundation of China [Grants 71801096 and 72101090] and the Guangdong Basic and Applied Basic Research Foundation [Grant 2022A1515011983]. V. N. Hsu was supported by the General Research Fund of the Hong Kong Research Grants Council [Grant CUHK14526216]. P. Guo received financial support from the Hong Kong Research Grants Council [Grant 15502820]. Supplemental Material: The e-companion and data are available at https://doi.org/10.1287/mnsc.2022.4424 .

We extend the existing literature on gender differences in competitive behavior by investigating tournament entry choices when a principal decides for an agent. In a laboratory experiment, we randomly assign subjects the role of either principal or agent. The principal decides whether the agent performs a real-effort task under piece-rate or tournament incentives. When deciding, the principal is informed about the agent’s previous performance, age, and residency. Between treatments, we vary whether the principal knows the agent’s gender. In a baseline treatment, we replicate the standard setting in which subjects decide for themselves whether to compete. Our main findings are, first, that there is no gender gap in tournament entry when principals decide for agents as opposed to the baseline treatment. Second, the gender gap closes because more women are made to compete by principals. Third, whereas there is no gender gap in either of the principal treatments, revealing the agent’s gender is associated with higher overall tournament entry rates. Exploratory analyses of principals’ choice determinants reveal a positive effect of preferences to take risks, competitiveness, and confidence in agents’ performances on making agents compete. In addition, we find no difference in how principals evaluate male and female agents’ performances. Finally, we test the efficiency of principals’ competition choices and show that they lead to fewer payoff-maximizing outcomes than when subjects decide for themselves. Additionally, overall tournament performances and winners’ performances are lower when agents are made to compete, but this effect is not robust to controlling for agents’ previous performances. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Funding: This work was supported by L. Meltzers Høyskolefond; Austrian Science Fund [Grant SFB F63 “Credence Goods, Incentives and Behavior”], and the Department of Economics at the University of Bergen. Also, we are grateful for funding from the Equal Opportunity Coordination Office of the University of Regensburg and its faculty of Business and Economics. Supplemental Material: The supplementary online material and data are available at https://doi.org/10.1287/mnsc.2022.4413 .

Improving traceability in food supply chains has been a big part of ongoing effort to reduce contamination risks and food waste. Recent technology advancement such as blockchain technology has propelled innovative retailers to promote food traceability system adoption in their supply chains. We develop a three-tier supply chain model with multiple upstream (tier 2) suppliers to investigate how traceability technology adoption affects incentives of supply chain members and whether and how its anticipated benefits can be realized. We find that full traceability brings direct revenue benefit to every supply chain member by saving uncontaminated food from disposal (pure traceability effect) but also leaves each tier of the supply chain vulnerable to its immediate downstream buyer’s exploitation through strategically lowering the purchasing price (strategic pricing effect). The interplay of the two effects may result in some of the supply chain members (even the retailer) being worse off with traceability adoption and the system being exposed to higher contamination risk; the latter is because of the weakened upstream supplier’s incentive to exert contamination risk-reduction effort. Moreover, we find that the supply chain network structure also influences the benefit distribution of traceability adoption: The retailer always benefits from traceability adoption in network structures where the tier 1 supplier’s strategic pricing power is eliminated or weakened; all supply chain members benefit from traceability adoption in a network with a large number of tier 2 suppliers. Finally, we show that alternative risk-mitigation schemes such as tier 2 coordination can diminish the value of traceability adoption, and partial traceability enabled by tier 1 product inspection can be more beneficial to the retailer than full traceability. This paper was accepted by Victor Martínez-de-Albéniz, operations management. Supplemental Material: The online appendices are available at https://doi.org/10.1287/mnsc.2022.4440 .

In many markets, buyers sign advance contracts before actual decisions on transactions or consumptions are made. Therefore, a buyer may have private information on expected payoff at the contracting stage, and as time moves on, new information on other components of payoff may arrive. However, prior information can be losable, forgettable, or unattended. In this paper, we investigate how limited memory may influence the optimal design of contracts for sequential screening. Despite memory loss, the buyer can make ex post inference about her initially informed type from the chosen contract. As ex ante screening facilitates subsequent retrospection, the chosen contract can serve as a self-reminding instrument. This would yield an endogenous demand for separation in ex ante contract choice. In response, distortions in the optimal contract design can be either mitigated or intensified, leading to improved or undermined social welfare, respectively. As a result, the equilibrium buyer surplus can be higher than that under perfect memory. We also show that the buyer can exhibit the so-called flat-rate bias, even though her preference is time consistent and perfectly predicted. In addition, as memory can be perfectly recovered from the equilibrium contract choice, investing on any other memory-improving instrument is redundant. Moreover, the buyer’s demand for screening can induce her to choose dominated refund contract. Nevertheless, when dominance must be obeyed, the seller may offer a menu of refund contracts with two-way distortions. This paper was accepted by Matthew Shum, marketing. Funding: The financial support offered by the Hong Kong RGC General Research Fund [Grant 14500519] is acknowledged.

This paper provides novel evidence suggesting that securities class action lawsuits, a central pillar of the U.S. litigation and corporate governance system, can constitute an obstacle to valuable corporate innovation. We first establish that valuable innovation output makes firms particularly vulnerable to costly low-quality class action litigation. Exploiting judge turnover in federal courts, we then show that changes in class action litigation risk affect the value and number of patents filed, suggesting firms take into account that risk in their innovation decisions. A new perspective we provide is that innovation success, not only innovation failure, can increase firms’ securities class action litigation risk. This paper was accepted by Victoria Ivashina, finance. Funding: E. Kempf gratefully acknowledges financial support from the James S. Kemper Foundation at the University of Chicago Booth School of Business. Supplemental Material: Data and the internet appendix are available at https://doi.org/10.1287/mnsc.2022.4388 .

Hiring in online labor markets involves considerable uncertainty: which hiring choices are more likely to yield successful outcomes and how do employers adjust their hiring behaviors to make such choices? We argue that employers will initially explore the value of available information. When employers observe successful outcomes, they will keep reinforcing their hiring strategies; but when the outcomes are unsuccessful, employers will adjust their hiring behaviors. To investigate these dynamics, we propose a two-component framework that links hiring choices with task outcomes. The framework’s first component, a Hidden Markov Model, captures how employers transition from unsuccessful to successful hiring decisions. The framework’s second component, a conditional logit model, estimates employer hiring choices. Analysis of 238,364 hiring decisions from a large online labor market shows that, often, employers initially explore cheaper contractors with a lower reputation. When these options result in unsuccessful outcomes, employers learn and adjust their hiring behaviors to rely more on reputable contractors who are not as cheap. Such hiring tends to be successful, guiding employers to reinforce their hiring processes. As a result, the market observes employers transition from cheaper, lower-reputation options with poorer performance to more expensive reputable options with better performance. We attribute part of this behavior to employer confidence and risk attitude, which can change over time. This work is the first to investigate how employers learn to make successful hiring choices in online labor markets. As a result, it provides platform managers with new knowledge and analytics tools to target employer interventions. This paper was accepted by Anandhi Bharadwaj, information systems. Supplemental Material: The e-companion is available at https://doi.org/10.1287/mnsc.2022.4426 .

Operating under both supply-side and demand-side uncertainties, a mobile-promotion platform conducts advertising campaigns for individual advertisers. Campaigns arrive dynamically over time, which is divided into seasons; each campaign requires the platform to deliver a target number of mobile impressions from a desired set of locations over a desired time interval. The platform fulfills these campaigns by procuring impressions from publishers, who supply advertising space on apps via real-time bidding on ad exchanges. Each location is characterized by its win curve, that is, the relationship between the bid price and the probability of winning an impression at that bid. The win curves at the various locations of interest are initially unknown to the platform, and it learns them on the fly based on the bids it places to win impressions and the realized outcomes. Each acquired impression is allocated to one of the ongoing campaigns. The platform’s objective is to minimize its total cost (the amount spent in procuring impressions and the penalty incurred due to unmet targets of the campaigns) over the time horizon of interest. Our main result is a bidding and allocation policy for this problem. We show that our policy is the best possible (asymptotically tight) for the problem using the notion of regret under a policy, namely the difference between the expected total cost under that policy and the optimal cost for the clairvoyant problem (i.e., one in which the platform has full information about the win curves at all the locations in advance): The lower bound on the regret under any policy is of the order of the square root of the number of seasons, and the regret under our policy matches this lower bound. We demonstrate the performance of our policy through numerical experiments on a test bed of instances whose input parameters are based on our observations at a real-world mobile-promotion platform. This paper was accepted by Baris Ata, stochastic models and simulation. Supplemental Material: The online appendices are available at https://doi.org/10.1287/mnsc.2022.4441 .

Many important managerial outcomes hinge on the co-occurrence of multiple uncertain events, a situation termed conjunctive risk. Whereas past literature has addressed the psychology of choosing to enter situations with conjunctive risk, this article elucidates a novel way in which the psychology of managing conjunctive risk is importantly distinct. We examine a case in which there are two independent events, one is currently less likely than the other, both are required for overall success, and the decision maker must evaluate opportunities to increase the chance of the less-likely or more-likely requirement. We introduce the hypothesis of a worst-first heuristic. Decision makers intuitively evaluate improvements in conjunctive risk according to their impact on the biggest barrier to success, the least likely of the required events. We find evidence for such a worst-first heuristic across nine experiments (n = 3,653, including samples from the United States and United Kingdom in Studies 1–5 and Studies S1–S3 in the online supplement, as well as a sample of managers in Study 6). Participants invest more to improve chances of less-likely requirements than more-likely requirements, even when the latter improvements have at least as much impact on the aggregate chance of success. Moreover, we find that decision makers exhibit this behavior particularly when managing conjunctive risk, as doing so makes them attend to which threat is the worst. Conversely, they do not appear to exhibit the behavior when making formally equivalent decisions about choosing between conjunctive risks. This bias toward underinvesting in stronger-links holds important implications for decision making in contexts subject to conjunctive risk—both managerial and societal. This paper was accepted by Yuval Rottenstreich, behavioral economics and decision analysis. Supplemental Material: Data and the e-companion are available at https://doi.org/10.1287/mnsc.2022.4411 .

We develop a theory wherein a priori identical banks may trade loans in a search market with reusable information. The equilibrium is unique, but its nature depends on the probability of a future market state. When the probability of a boom is high, all banks hold no equity and do no screening. When this probability is low, all banks choose a high level of equity and screen loans. For intermediate probability values, the equilibrium is heterogeneous, with some banks posting equity and screening and others avoiding equity and screening. This endogenously arising heterogeneity generates interbank trading. The credit market is partially frozen in a recession: only high-capital banks have continued funding access. Low-capital banks obtain funding by selling legacy loans to banks with “financial muscle,” so market funding is reallocated from low-capital to high-capital banks. This paper was accepted by Bruno Biais, finance.

We study a large-scale (n = 50,000) natural field experiment implemented by the U.S. Social Security Administration aimed at increasing the timely and accurate self-reporting of wages by Supplemental Security Income (SSI) recipients. A letter reminding SSI recipients of their wage reporting responsibilities significantly increased both the likelihood of reporting any earnings and the total earnings reported. However, the specific letter content—providing social information or highlighting the salience of penalties—had no systematic effect. We develop a conservative estimate that the letters generated roughly $5.91 in savings per dollar spent, highlighting the value of such a nudge in this important context. This paper was accepted by Yan Chen, behavioral economics and decision analysis. Supplemental Material: The data files and online appendix are available at https://doi.org/10.1287/mnsc.2022.4645 .