Knowledge that Transforms

We study a dynamic pricing problem, where a firm offers a product to be sold over a fixed time horizon. The firm has a given initial inventory level, but there is uncertainty about the demand for the product in each time period. The objective of the firm is to determine a dynamic pricing strategy that maximizes revenue throughout the entire selling season. We develop a tractable optimization model that directly uses demand data, therefore creating a practical decision tool. We show computationally that regret‐based objectives can perform well when compared to average revenue maximization and to a Bayesian approach. The modeling approach proposed in this study could be particularly useful for risk‐averse managers with limited access to historical data or information about the true demand distribution. Finally, we provide theoretical performance guarantees for this sampling‐based solution.

To reduce preventable readmissions, many healthcare systems are transitioning from fee‐for‐service (FFS) to other reimbursement schemes such as pay‐for‐performance (P4P) or bundled payment (BP) so that the funder of a healthcare system can transfer to the hospital some of the financial risks associated with patient rehospitalizations. To examine the effectiveness of different schemes (FFS, P4P, and BP), we develop a “health co‐production” model in which the patient's readmissions can be “jointly controlled” by the efforts exerted by both the hospital and the patient. Our analysis of the equilibrium outcomes reveals that FFS cannot entice the hospital and the patient to exert readmission‐reduction efforts. Relative to BP, we find that P4P is more effective in reducing readmissions over a wider range of scenarios. However, BP tends to be more effective in keeping lower combined patient costs and funder payments to the hospital. Finally, we find that some patient cost‐sharing can be optimal for the funder under both P4P and BP.

Vacation Timeshare is a form of ownership or “right to use” of a resort property for a specific time period (typically a week) each year. Timeshare exchange refers to the non‐monetary trading of timeshare weeks among owners, so that they can interchange their vacation homes to experience new destinations. The need for member participation during the exchange process has been well‐recognized for a variety of practical reasons, including the reluctance of members to accept an authoritarian solution that does not provide any information about the exchange process and their desire to experience some control over the process. Another important need is to ensure that, given the members’ preferences, an exchange solution offers collectively the best‐possible improvement over their currently‐owned weeks, while being “fair” to all participants. We suggest two objectives to capture the efficiency and fairness of an exchange solution. For the resulting bi‐criteria problem, we show that a solution that is simultaneously near‐optimal on both objectives may not exist. Our main contribution is an efficient algorithm in which (i) each member uses her private preference list to communicate with other members, and the members, through such communications, collectively achieve an individually rational allocation, and (ii) for any desired approximation bounds α and β on, respectively, efficiency and fairness, the following property holds: if an ( α, β)‐approximate solution exists, then the solution provided by the algorithm satisfies this approximation guarantee; otherwise, the solution is an α‐approximation on the efficiency measure and, among all such allocations, has the best fairness measure.

Contemporary products are usually designed and produced in large inter‐firm supply networks rather than by single firms. However, our understanding of such networks is still limited due to the lack of network‐wide empirical data as well as the complexity and nonlinearity of supply networks. Herein, we introduce a network formation model to extend and generalize the prior empirical studies that have revealed variable hierarchy topologies and firm‐level transaction specificities across the supply networks for automobiles and electronics. We call it the “ hierarchy‐niche model.” With tuning the parameters for transaction specificity and transaction breadth, the model can generate a wide spectrum of stochastic networks that comply with the production hierarchy to varied degrees. Our simulation analyses show that the model‐generated stochastic networks capture hierarchical and cyclic topologies of real‐world automobile and electronics supply networks. The model, which relates firm‐level transaction patterns to network‐wide emergent topologies, can be further utilized to inform and guide firms’ transaction strategies concerning the overall supply network.

We consider a decentralized supply chain (DSC) under resale price maintenance (RPM) selling a limited‐lifetime product to forward‐looking customers with heterogeneous valuations. When customers do not know the inventory level, double marginalization under RPM leads to a higher profit and aggregate welfare than without RPM under a two‐part tariff contract (TT). Both RPM and TT profits are higher and aggregate welfare is lower than in a centralized supply chain (CSC). When customers know the inventory, RPM coincides with CSC. Thus, overestimation of customer awareness may lead to overcentralization of supply chains with profit loss comparable with the loss from strategic customers. The case of RPM with unknown inventory is extended to an arbitrary number of retailers with inventory‐independent and inventory‐dependent demand. In both cases, the manufacturer, by setting a higher wholesale price, mitigates the inventory‐increasing effect of competition and reaches the same profit as with a single retailer. The high viability and efficiency of RPM in using double marginalization as a strategic‐behavior‐mitigating tool may serve as another explanation of why manufacturers may prefer DSC with RPM to a vertically integrated firm.

We experimentally investigate horizontal coordination between suppliers where group output is limited by the lowest‐performing member and groups are formed endogenously. All participants first choose between one of two groups, where one group has an entry fee. Participants then simultaneously make capacity choices, and the minimum choice within each group dictates profits for group members. Allowing participants to select their group, thereby indirectly determining the group size, has strong implications for equilibrium outcomes. We find both theoretically and experimentally that the group with an entry fee always achieves higher output, while members of both groups earn equal profits in equilibrium. From a managerial perspective, costly membership fees for exclusive groups can separate high‐performing and low‐performing subjects when group selection is endogenous, even when the costly fee provides no other benefits. Interestingly, the group with an entry fee always has fewer subjects, suggesting that a group membership fee acts as a deterrent to poor‐performing subjects.

Firms in advanced economies are increasingly outsourcing software and technology development as well as other knowledge work to a worldwide supply base. Standard economic and learning models predict that focal firms should outsource either all or none of a particular activity unless extra resources are required during cyclical demand peaks or access is needed to some tightly appropriable intellectual property. However, recent evidence shows that, even when these exceptions do not apply, many firms pursue a partial outsourcing strategy. We develop a dynamic optimization model to provide a rational explanation for this observation. In our model, learning from prior projects occurs at both the subsystem level and the overall systems level (e.g., systems integration and architecture). Learning at the two levels interacts such that integration capabilities can dynamically build and decay. The model generates conditions where partial outsourcing is rational and dominates the extreme conditions of complete insourcing or complete outsourcing. Our model also specifies the conditions for regime change between insourcing and outsourcing and cycling between insourcing and outsourcing, and overshooting or undershooting the long‐run outsourcing target. Furthermore, we show that these results are highly path‐dependent under short horizons. The model also provides explanation for interesting questions such as why the rate of outsourcing might be U‐shaped in the rate of technological change and why startups so often insource in contrast to more established counterparts in similar industries.

We model a real‐world service center with cross‐trained agents serving customer requests that are heterogeneous with respect to complexity and priority levels: High priority requests preempt low priority requests and low‐skilled agents can only serve less complex requests, while high skilled agents can serve all requests. Our main aim is to dynamically assign requests to agents considering the priority and complexity levels of requests. We model this system as a Markov chain that is infinite in multiple dimensions and thus is not amenable to exact analysis. We therefore apply approximation and bounding techniques to develop a tractable, novel algorithm using the Matrix Analytic Method. Our algorithm closely approximates the operations of the real‐world service system under a simple but effective threshold‐based request‐assignment policy. Extensive computational results demonstrate the usefulness of our algorithm to minimize costs given an existing staffing configuration, as well as in helping to make long‐term staffing decisions. In addition, our algorithm also has at least two orders of magnitude shorter computation times than each replication of simulation. Hence, it is both fast and accurate.

We experimentally study first‐price split‐award auction formats as they can be found in procurement markets where suppliers have economies of scale. Our analysis includes sequential and combinatorial auctions, which allow for bids on the package of two shares and single shares. We derive equilibrium predictions as hypotheses for bidder behavior in our laboratory experiments. These equilibrium predictions help explain important patterns in our experimental results. The combinatorial first‐price sealed‐bid auction yields lower prices than the other mechanisms and is highly efficient independent of the extent of scale economies. With strong economies of scale both combinatorial auction formats let the auctioneer incur significantly lower procurement costs and generate high efficiency compared to the sequential auction. We also find high efficiency of the combinatorial first‐price sealed‐bid auction in experiments with diseconomies of scale, making these auctions attractive if the buyer has uncertainty about the economies of scale in a market. Our analysis shows that combinatorial split‐award auctions can be an attractive alternative for the buyer compared to their sequential counterparts.