The W. Edwards Deming Center for Quality, Productivity and Competitiveness supports applied research activities through doctoral research fellowships with the objective of producing new knowledge related to the field of operations management, and providing students with the opportunity to use company data to validate academic hypotheses, increasing their exposure to industry problems and enhancing the relevance of their research.
Sponsored Projects 2012
“The Effects of Access to Mainstream Financial Services on the American Poor: Evidence from Data on Recipients of Financial Education”
A significant fraction of low-income households in the US operates outside of the financial mainstream, relying on expensive “alternative financial services” for carrying out basic transactions and for borrowing, and lacking adequate mechanisms for saving. In this paper, Assaf Shtauber (2014), a PhD candidate in the Finance and Economics Division, advised by Gur Huberman, Professor, Columbia Business School, examines the causal effects of bank account ownership on a representative sample of the American urban poor. Using a unique dataset and exploiting quasi-random variation in the likelihood of account opening, he finds a significant reduction in debt delinquency and higher likelihood of credit score improvement among previously unbanked individuals, and provides evidence regarding the mechanisms behind these results.
“Pricing Managed Lanes”
A managed lane is loosely defined as a separated highway lane operated with active flow management. The entity responsible from operating the lane manages the flow to the lane by levying tolls or setting regulations, such as requiring a minimum number of passengers in the vehicle. Managed lanes with tolls are becoming an important way to finance new public works projects such as highway expansions or other transportation related programs. Caner Gocmen (2013), a PhD candidate in the Decision, Risk, and Operations division, advised by Robert Phillips, Professor of Professional Practice at Columbia Business School, investigates the optimal operating policies for managed lanes with tolls. For different operating environments, such as constant or time-varying demand, they identify some properties of the revenue maximizing policy.
“Organizational Efficiency and Hierarchical Functioning Following the Implementation of a Mobility Initiative”
In this project, Eric Anicich (2016), a PhD student in the Management division, advised by Adam Galinsky, Professor at Columbia Business School, highlight the importance of making a theoretical distinction between power and status and move beyond person-based, compositional explanations for relationship conflict in the workplace. Using archival data and direct surveys of senior level managers in a federal agency, they found that supervisors experienced higher levels of relationship conflict when their roles provided power without status. This intriguing finding occurred in the context of a mobility initiative within the organization and spurred further empirical investigation. Findings from several follow-up studies enhance current understanding of social structure, workplace conflict, and social relations by integrating social psychological and organizational theories of power, status, and conflict.
“Consequences of New Channel Adoption: Addressing the Self-Selection Problem”
According to the Direct Marketing Association’s report, 40 percent of retailers sell through three or more channels, for example, catalogs, Internet stores, apps for smart phones, or brick-and-mortar stores. In this paper, Hasan Tolga Bilgicer (2013), a PhD student in the Marketing division in collaboration with Don Lehmann, Professor at Columbia Business School, examine what happens after companies add new channels. They find that if the new channel has similar attributes to company’s existing channel, the new channel steals sales from the incumbent channel (cannibalization effect). If the new channel offers new services that complement the incumbent channel, such as, easing the product return processes, or providing after sales service, the new channel augments the sales at the existing channel (synergy effect).
“Recommending Content in a Changing World: What is still news?”
A new class of online services allows publishers to direct readers to other web-based content they may be interested in. Yoni Gur (2014), PhD candidate in the Decision, Risk, and Operations division, advised by Omar Besbes, Assistant Professor at Columbia Business School, and Assaf Zeevi, Professor and Vice Dean for Research at Columbia Business School, study this dynamic recommendation problem, focusing on challenges introduced by the massive stream of new content and the heterogeneous shelf-life of articles. With the objective of quantifying the aging of content in the presence of rapidly evolving features, they attempt to identify from different data sources the key underlying factors to track, and how one should do so. This work is based on a collaboration with a leading provider of online content recommendations.
“The Effect of Dynamics Incentives on Borrowers Rapayment Behavior: Evidence from a Randomized Experiment at a Department Store”
Do borrowers respond to dynamic incentives? In this work, Andres Liberman (2013), PhD candidate in the Finance and Economics Division, advised by Daniel Paravisini, Professor at London School of Economics, design and apply a field experiment in a large department store in Chile to study the effect of dynamic incentives on the take-up and performance of unsecured consumer loans. The store’s baseline credit contract features a fixed credit line. They offer a randomly selected group of borrowers (treated group) a contract that specifies an increasing path for credit lines. They further subdivide and assign the store’s baseline contract to a randomly selected subgroup within the treated group. Finally, a control group is offered and assigned the baseline contract. This allows them to separately identify the selection and incentive effects of the contract that features explicit dynamic incentives on debt repayment. They conclude by comparing the importance of dynamic incentives relative to pricing, reputation, and other incentives for unsecured borrowers to repay their debt.
“Dynamic Matching in Residential Real Estate Markets”
Hua Zheng (2014), PhD candidate in the Decision, Risk, and Operations division, in collaboration with Costis Maglaras, Professor at Columbia Business School, and Ciamac Moallemi, Associate Professor at Columbia Business School, study the residential real estate markets as dynamic matching systems with an emphasis on their microstructure. The residential units are heterogeneous with respect to their attributes such as location and size, while the agents are also characterized by their idiosyncratic preference, delay tolerance, and other financial considerations. They study the dynamics of this market and characterize its steady-state behavior, such as market depth and price dispersion.
Sponsored Projects 2011
“Analyzing Agent Productivity in Service Delivery Systems”
The goal of this research is to analyze factors that drive agent productivity in service delivery systems and utilize this information to improve efficiency in the allocation of work. More specifically, Yina Lu (2013), a PhD candidate in the Decision, Risk, and Operations division, advised by Marcelo Olivares, Associate Professor at Columbia Business School, wants to analyze how factors such as workload, multi-tasking, agent skill level, and skill-and-job matching would affect agents’ productivity using primary objective data, and based on this design a more efficient job routing scheme.
“Competition and Market Structure in Online Display Advertising with Ad Exchange”
“Endogeneity and Price Sensitivity in Customized Pricing”
Endogeneity occurs in customized pricing when a seller uses unrecorded customer characteristics that are correlated with price-sensitivity in setting the price. Endogeneity can lead traditional regression approaches to under-estimate price elasticity. In this study, Ahmet Serdar Simsek (2013), a PhD candidate in the Decision, Risk, and Operations division, working with Robert L. Phillips, Professor of Professional Practice, and Garrett van Ryzin, Paul M. Montrone Professor of Private Enterprise, and Chair of Decision, Risk, and Operations division, use two sources of data from auto loan pricing: one from an on-line lender and one from an indirect lender to test for endogeneity. They present their results as well as recommendations for how to detect and control for endogeneity in the estimation process.
“Enhancing Healthcare OM Research via Collaboration with Kaiser Permanente”
The objective of this project is to learn how to use available quantitative tools in the most effective way to improve capacity planning and resource allocation in hospitals. Hailey Song-Hee Kim (2013), a PhD candidate in the Industrial Engineering & Operations Research division, working with Carri W. Chan, Assistant Professor at Columbia Business School, and Marcelo Olivares, are studying hospital settings and operations in order to correctly structure the models, estimate the parameters, and develop performance objectives and to find ideas for potential research questions. Additionally, there are two specific questions they are trying to answer: (1) examining whether the occupancy levels of the available recovery units with different levels of care play a role in deciding what type of recovery unit a patient is routed to after a surgery or a visit to ER and if so, how this routing pattern further affects patient outcomes and (2) developing an optimal policy for giving a preventative therapy which is done in the Intensive Care Unit (ICU) to patients with a particular disease given the prevalent shortages of ICU beds in hospitals.
“Portfolio Execution with Short-Term Predictability”
In this project, Mehmet Saglam (2013), a PhD candidate in the Decision, Risk, and Operations division, advised by Ciamac C. Moallemi, Associate Professor at Columbia Business School, would like to test whether linear decision rules can be tractable in complex and realistic portfolio execution problems. If there exists any performance gain, they would like to quantify the improvement over current algorithmic practices used in the industry. In a recent working paper, they lay out the theoretical background of using linear decision rules in large class of dynamic portfolio choice problems and in this project, they would like to apply this methodology to optimal portfolio execution problems with realistic parameters. Model parameters will be estimated from real data and using the calibrated model they will compute the best linear execution policy and compare its performance to those of current algorithms used in the industry.
“Structural Estimation of a Large-Scale Procurement Combinatorial Auction: The Chilean School-meal Project”
Combinatorial auctions are particularly useful in procurement when items exhibit cost synergies. However, allowing bidders to bundle items that do not exhibit synergies may hurt the efficiency of the allocation. In this study, Sang Won Kim (2012), a PhD candidate in the Decision, Risk, and Operations division, working with Marcelo Olivares and Gabriel Y. Weintraub, develop a structural estimation method to uncover the bidders’ cost structure of a large-scale combinatorial auction; the Chilean auction for school meals. Based on these estimates they analyze and suggest important improvements to the auction design.