supply chain management research and production and operations management: review, trends, and...

Supply Chain Management Research andProduction and Operations Management:

Review, Trends, and Opportunities

Panos Kouvelis • Chester Chambers • Haiyan WangOlin School of Business, Washington University, St. Louis, Missouri 63130, USA

Cox School of Business, Southern Methodist University, Dallas, Texas 75275, USAOlin School of Business, Washington University, St. Louis, Missouri 63130, USA

[email protected] • [email protected] • [email protected]

We reviewed the manuscripts focused on Supply Chain Management that had been published inProduction and Operations Management (POM) over roughly 15 years (1992 to 2006). The manu-

scripts covered dealt with topics including supply chain design, uncertainty and the bullwhip effect,contracts and supply chain coordination, capacity and sourcing decisions, applications and practice, andteaching supply chain management. In the process of this review, we highlight the significant contri-bution of POM to the field of supply chain management, and illustrate how this body of work has servedto further the mission of the journal. We then highlight works from this group along with the discussionof selected papers from other top journals in an effort to provide a reasonably complete overview ofimportant issues addressed in recent supply chain management research. Using our research survey andconceptual overview of the area as a foundation, we offer comments which highlight opportunities andsuggest ideas on how to usefully expand the body of work in the supply chain management area.

Key words: supply chain management research; literature review; research opportunities in supply chainmanagement

Submissions and Acceptance: Received September 30, 2006; Accepted by Kalyan Singhal after one revision.

1. IntroductionThe first 60 issues of Production and Operations Man-agement (POM) included 49 manuscripts accepted bythe Supply Chain Management department. This con-stitutes a relatively small portion of the 399 papersprinted in the pages of POM over this 15 year period.However, only one of these papers dates prior to 1997,while 18 appeared in the nine issues from Spring 2004to Spring 2006. Consideration of the pages of other topjournals in our field reveals a similar pattern. Thisstrongly suggests that supply chain management(SCM) has recently become the dominant theme inoperations management research. This paper beginswith a survey of many of the papers on supply chainmanagement published in POM, including those ac-cepted by other departments. It is useful to organizethese papers by topic in Sections 2 to 6. These sectionsfocus on supply chain dynamics and the BullwhipEffect (BWE), supply chain capacity and sourcing de-

cisions, supply chain management (SCM) applicationsand practices, supply chain planning and scheduling,and approaches to teaching SCM.

Obviously, the review of such a small number ofpapers cannot provide or even suggest a completeoverview of the literature on supply chain manage-ment that has emerged over the past 15 years. We alsobelieve it to be unwise to attempt a complete overviewof this literature in this limited space. In fact, one canargue that the collective reviews appearing in editedsupply chain management research books by deKokand Graves (2003), Simchi-Levi, Wu, and Shen (2003),and Tayur, Ganeshan, and Magazine (1998) still misssome works in the area, even though they collectivelyconsider over 1,000 manuscripts. Therefore, we havedecided to use several papers from POM as spring-boards into discussions of selected topics which haveattracted well deserved attention in other researchvenues. Specifically, we will focus on supply chain

POMSPRODUCTION AND OPERATIONS MANAGEMENTVol. 15, No. 3, Fall 2006, pp. 449–469issn 1059-1478 � 06 � 1503 � 449$1.25 © 2006 Production and Operations Management Society

449

coordination with contracts, coordination issues in-volving internet channels, postponement, and the roleand value of operational hedging in managing supplychains. To reiterate, by no means do we imply that thisis an exhaustive list. Rather, our intent is to highlighta number of major areas that are arguably at theforefront of supply chain research at the present time.Further, we hope to go beyond an outline of the SCMresearch landscape and to suggest open areas in thefield that seem ripe for significant contributions by thePOM community.

2. Supply Chain Dynamics and theBullwhip Effect

The “Bullwhip Effect” (BWE) refers to the phenome-non that order variability increases as orders moveupstream along the supply chain. This phenomenon isso well known that it is sometimes referred to as, “thefirst law of supply chain dynamics.” This effect is welldocumented by researchers both in theory and prac-tice and has several important implications. The BWEcan lead to excess inventory as well as unused oroverused capacity. It dramatically increases the oper-ating costs of the supply chain system and often leadsto serious supply and demand mismatches and dete-rioration in customer service levels. A variety ofcauses of the BWE have been identified in the litera-ture. These factors have been effectively classified inthe seminal articles of Lee et al. (1997ab) into fourmajor categories: (a) Informational inefficiencies andaccentuating factors (e.g., propagation of distorted de-mand information via erratic orders, long and/or vari-able lead times, uncontrollable product production,etc.); (b) Order batching effects (e.g., pervasive impactof fixed costs and other economies of scale consider-ations in batching orders, periodic reviews, and“hockey-stick” phenomena in sales behavior etc.); (c)Dynamic pricing and promotional campaigns (pro-vide incentives for customers to wait and place largeorders, thus creating hard to predict “peaks” and “val-leys”); and (d) System gaming behaviors (often inenvironments where capacity and/or material con-straints might lead to order rationing, thus leading toorder inflations/deflations as perception about systemconstraints change over time).

Warburton (2004) captures the systemic nature ofthe BWE by studying the fundamental differentialdelay equations that describe the evolution of the re-tailer’s inventory level in the supply chain, whichwere originally described by Forrester (1961). In par-ticular, he considers the impact of a single jump indemand. This work finds analytical solutions for theresulting equations. With these solutions in hand onecan explicitly quantify the BWE, consider various pol-icies exactly without the use of simulation, and predict

the impact of specific changes to the chain’s structureor practices.

Based on a case study of the machine tool industry,Anderson, Fine, and Parker (2000) investigate the am-plification of demand variability by presenting a sys-tem dynamics simulation model. This work highlightsa phenomenon which they label the “investment ac-celerator effect.” For machine tools, there is a fairlystable demand for replacement units, but when dura-ble goods producers invest in anticipation of even asmall surge in their demand, this often shows itself asa huge increase in the orders placed with machine toolmanufacturers. For example, a plant may use 100 ma-chine tools and replace 5% of them in a typical year. Ifthis firm anticipates a 5% increase in sales, it may planto buy 10 machine tools this period; 5 new and 5replacements. Thus, a 5% surge in expected sales atone level becomes a 50% surge for the supplier. This isparticularly problematic for machine tools becausethis pattern is correlated across the entire customerbase. The authors recommend that forecasting policieswhich tend to produce smoother orders be adopted bydownstream customers to reduce the variability seenby the upstream suppliers.

Our experience suggests that the most intuitive re-sponse to the BWE is some form of information shar-ing to help coordinate the supply chain. Chatfield,Kim, Harrison, and Hayya (2004) present a simulationstudy to evaluate causes of the BWE. Their resultsinclude the following insights: (1) without forecastupdating there is no BWE if simple ordering rules areadhered to, (2) lead time variability is a significantcause of the BWE, (3) information sharing is the mostdirect way to reduce the BWE, and (4) making use ofinformation without sharing it is typically worse thenignoring the information completely. This last point isrelated to the first one. It is the updating following thereceipt of new information that leads to the amplifi-cation of orders, and particularly so when customerorders are the only data available to a buyer whenplacing orders with his supplier. These results suggestthat policies which stimulate information sharing andcoordination are critical mitigants to the BWE.

3. Supply Chain Design, Capacity,and Sourcing Decisions

Supply Chain design involves strategic decisions andplans regarding where to locate facilities (for production,storage, distribution, retail, etc.), how to allocate capaci-ties or assign production tasks to the various facilities,how to choose and develop supplier and distributionchannels, and how to organize the interfaces among thevarious parties in the supply chain. Supply chain designis not stationary and in order to be effective has to beintegrated with the processes of product (service) and

Kouvelis, Chambers, and Wang: Supply Chain Management Research and POM: Review, Trends, and Opportunities450 Production and Operations Management 15(3), pp. 449–469, © 2006 Production and Operations Management Society

manufacturing (delivery) system design. By studyingthe supply chains of fast-clockspeed industries (such asInternet service, personal computers, and multimediaentertainment), Fine (2000) concludes that the ultimatecore competency of an organization is supply chain de-sign, which is defined as “choosing what capabilitiesalong the value chain to invest in and develop internallyand which to allocate for development by suppliers.”This work advocates the concept of “three-dimensionalconcurrent engineering”, i.e., integrating supply chaindesign, product development, and process develop-ment. Product development includes architecturalchoices (e.g., integrality vs. modularity decisions) anddetailed design choices, while process development in-cludes unit process and manufacturing systems devel-opment. Supply chain development encompasses deci-sions on whether to make or buy a component, sourcingand contracting decisions, and logistics and coordinationdecisions.

When facing either demand or supply uncertainty,it is often valuable to have multiple sources of inputsor of finished items. Consequently questions regard-ing the optimal number and types of sources are stud-ied in a number of POM papers. For example,Agrawal and Nahmias (1997) focuses on the optimalnumber of suppliers when demand is known and theyield from the supplier’s production system is nor-mally distributed. Given their formulation, increasingthe number of suppliers reduces the variability ofdelivered quantities but increases the cost of manag-ing the supplier relationships. The objective is to max-imize profits considering holding, shortage, and sup-plier management costs. Sodhi (2005) considers atactical planning model that uses linear programming.First a deterministic model is solved to generate aproduction plan. Second a stochastic problem issolved to devise an “ideal” replenishment plan in theface of demand uncertainty. The gap between theoutputs of the two models can be used to reallocatecapacity among different products as a plan to dealwith the demand/inventory risks.

Agrawal, Smith, and Tsay (2002) develops a method-ology for capacity planning and vendor managementwhen the retailer places orders with two vendors whodiffer in lead times, costs, and production flexibility. Themodel uses multiple demand scenarios and capturesseveral supplier attributes to populate an LP formulationof the planning problem. The system is developed incooperation with a retail chain that ultimately used themodel as part of a decision support system.

Yan, Liu, and Hsu (2003) studies how an updateddemand forecast affects a manufacturer’s ordering de-cisions when sourcing raw materials in a dual-sup-plier system. Their model includes the ability to placetwo raw material orders before the selling season: oneat a low cost from a slow supplier, and one at a high

cost from a fast supplier. The manufacturer puts effortinto updating his demand forecast and demand fore-cast accuracy increases with time. Consequently, thebuyer faces a tradeoff between the increased cost offast delivery and the value of the information gainedbetween the ordering points. A two-stage stochasticprogramming model is established and the manufac-turer’s expected cost function is proved to be convexunder the assumption that the optimal order level is alinear function with respect to the mean of the de-mand forecast.

Increasing the flexibility of production capacity canprovide an effective response to uncertain demandand reduce the cost induced by that uncertainty.Fisher, Hammond, Obermeyer, and Raman (1997) de-velop the idea of “accurate response” as a way tomanage demand uncertainty. One critical notion hereis to increase effective response capability by dividingproduction capacity into two distinct parts: specula-tive production capacity used before the observationof demand, and reactive production capacity, usedafter an early demand indicator. Other approachesinclude increasing total capacity, reducing lead times,throughput times, and/or transportation times, gath-ering market data earlier, improving market intelli-gence, and reducing minimum lot sizes. The authorsthen use data from a fashion/sporting goods producerto discuss the feasibility and relative impacts of thesemechanisms on supply chain responsiveness.

Hazra, Mahadevan, and Seshadri (2004) model anelectronic market where a buyer conducts an onlinereverse auction to select suppliers to meet his capacityrequirements. The reverse auction mechanism de-scribed in the paper works as follows: the buyer se-lects a subset of suppliers from the enlisted pool andthen equally allocates his capacity requirementsamong the selected suppliers at the uniform price ofthe lowest bid among the unselected suppliers. Thecontracts awarded through this electronic market arelong term. The suppliers can also sell their capacity ina traditional open market with a fixed historical price,but incur a selling expense (e.g., searching cost, ad-ministrative cost, and negotiating cost, etc.) and facethe risk of unsold capacity. The authors derive thesuppliers’ bidding price as a function of their availablecapacity and show that suppliers with larger capaci-ties would quote a lower price in the electronic mar-ket. Based on the derived supplier’s capacity-pricecurve, the authors model a scenario where the buyerannounces the number of suppliers to be selected foraward of a contract that minimizes the cost for meet-ing his capacity requirements.

As firms become increasingly global, they are ex-posed more and more to the policies of foreign gov-ernments. Many governments (or regional tradeagreements, like NAFTA) have set local content rules

Kouvelis, Chambers, and Wang: Supply Chain Management Research and POM: Review, Trends, and OpportunitiesProduction and Operations Management 15(3), pp. 449–469, © 2006 Production and Operations Management Society 451

for companies that wish to operate in their country (orwithin the region protected by the trade agreement).Local content rules require manufacturers to purchasea certain amount of components from suppliers lo-cated in that country (or the protected region). Mun-son and Rosenblatt (1997) formulates and solves amathematical programming model to select supplierswhile satisfying local content provisions. In their nu-merical studies, Munson and Rosenblatt consider asingle plant producing several components under lo-cal content constraints. As the required local contentpercentage goes up, the plant increasingly preferssourcing from local suppliers and the chances becomemuch greater that we shall see high values of the localcontent percentage. However, if the government setsthe local content percentage too high, it may causecompanies to shift production to other countries. Suchpolicies eventually become detrimental to local sup-pliers rather than helpful as they lead to reducedproduction and investment in the country.

The issue of designing a global facility network isexamined in Kouvelis and Munson (2004). The au-thors develop a structural equations model, using theglobal sensitivity approach of Wagner (1995), based ona mixed integer program that captures the facilitiesnetwork design complexities and tradeoffs. The pre-sented work can be used both as a conceptual frame-work and as a decision support tool for global supplychain design. Other work emphasizes the robustnessof global supply chain designs in the presence ofsignificant uncertainty. The designed supply chain isrobust in the sense that it hedges the firm’s perfor-mance against the worst contingencies in terms ofuncertain factors (demand, exchange rates, commod-ity prices, etc.) over a planning horizon. The work ofGutierrez and Kouvelis (1995) formally introducesmany of these concepts and presents effective algo-rithmic procedures for the solution of the relatedmixed integer programs. The wide applicability of theapproach is further illustrated in the work of Lowe etal. (2002) and illustrated in the popular case, Appli-chem (Flaherty 1996). This work addresses supplychain design in the face of exchange rate uncertaintyand makes the material easier to transfer into MBAand Executive classrooms.

4. Supply Chain ManagementPractice: Vendor ManagedInventory and ReengineeringPrograms

The literature review by Cohen and Mallik (1997)points out that the theoretical studies in supply chainmanagement often suggest policies that are impracti-cal and hard to implement. This work argues that dueto the gap between theory and practice, there is a need

for research which reflects the complexity of real sup-ply chains and highlights important practices and ap-plications. Several papers published in POM do rise tothis challenge.

Campbell Soup’s continuous replenishment pro-gram (CRP) is a notable innovation to improve theefficiency of inventory management throughout thesupply chain. It is a successful Vendor Managed In-ventory (VMI) program for “functional products.” Useof the term “functional products” stems from the orig-inal reference of Fisher (1997). Cachon and Fisher(1997) describes Campbell Soup’s program. This sys-tem involves a combination of ‘Every Day Low Pric-ing’ (EDLP), daily transmissions of store inventorylevels and sales for each SKU, and decisions by Camp-bell’s concerning shipments to the retailer’s distribu-tion center. The authors also use data from this systemto develop improved inventory management rules toimplement continuous replenishment. They show thatimplementation of these rules would greatly increasethe impact of the CRR approach.

Kreipl and Pinedo (2004) discuss how to integratemedium term production planning models and shortterm scheduling models in the design and implemen-tation of decision support systems which help managesupply chains. Production planning models describedin this paper optimize several consecutive stages in acollaborative supply chain. Particularly, the produc-tion planning model is designed to allocate produc-tion of different products to the various facilities ineach time period in order to minimize the systematicproduction costs, holding costs and transportationcosts under the given capacity (production and trans-portation) constraints. The output of the planningmodels is an input to the detailed scheduling process,which has considerably narrower scope in terms oftime and space and aims at minimizing the total setuptime on the machines at the bottleneck as well as thetotal weighted tardiness. Their ideas are operational-ized through a real application in Carlsberg A/S, aDenmark-based beer brewer with global distribution.

The pages of POM have included several surveystudies, which address the application and practice ofsupply chain management in industry. Through a sur-vey of 30 ongoing partnerships, Kopczak (1997) inves-tigates the interaction between the formation of logis-tics partnerships and supply chain restructuring in theu.s. computer industry. Logistics partnership is de-fined as a business relationship involving the out-sourcing of a set of logistics activities, which involvemore than one logistics function and are of a long-ternnature. Restructuring is defined as involving signifi-cant changes in the structure of the outsourced part ofthe supply chain via the partnership. These changesmay include a change in the warehouse structure(number of tiers, number of warehouse, etc.), reassign-


ment of tasks between tiers, redistribution of inven-tory between tiers (centralized vs. distributed stock-ing), a structural change in the transportationnetwork, designating new consolidation points, andreassignment of roles and responsibilities among sup-ply chain entities. The investigation found that bothrestructuring and non-restructuring organizationswere common in the computer industry and that bothtypes of organizations use outsourcing, but for differ-ent reasons. Restructurers are driven by the desire torestructure the supply chain while the non-restructur-ers are driven by a desire to focus on the firm’s corebusiness. The survey results support the hypothesisthat relationships which involve efforts to restructurethe supply chain are likely to provide greater benefits.

Clark and Hammond (1997) discusses the relation-ship between business process reengineering andchannel performance for firms implementing elec-tronic data interfaces (EDI) in the U.S. grocery indus-try. They combine data on inventory turns, stockouts,and sales with variables indicating the use of EDI,EDLP, and CRP to isolate the effects of each practice.They find that EDLP alone provides statistically sig-nificant benefits, while EDI alone does not. Their worksuggests that the process reengineering required toimplement CRP is the largest driver of value, whileEDI is a common enabler of the process improvement.

The emergence of the internet has had a significantimpact on how firms interact with each other and theircustomers. Johnson and Whang (2002) provides a de-tailed review of the existing literature on the impact ofe-business on supply chain management. The authorscategorize the various forms of e-business applicationsinto e-commerce, e-procurement and e-collaboration andexamine research in the three forms of e-business.Basically, e-commerce uses the internet to facilitatesupply chain partners’ interaction with customers,while e-procurement provides e-enabled firms effi-cient material procurement. E-collaboration is definedas business-to-business interactions facilitated by theinternet, which go beyond transactions and may in-clude information sharing and integration, decisionsharing, process sharing, and resource sharing.

Klassen and Vachon (2003) take an interesting per-spective on supply chains. In particular, the authorsinvestigate how the supply chain management activ-ities (specifically, the collaborative activities and up-stream evaluative activities) affect the plant-level en-vironmental investment through a survey conductedamong plant managers from selected manufacturingsites in Canada. This work finds that the collaborativeactivities such as site visits, exchange of personnel,and technical assistance, etc., significantly impact en-vironmental investment while the evaluative activitiesincluding assessing supplier performance, relying onsupplier certification to reduce the need for inspection,

recognizing supplier achievements, and offering feed-back has limited influence on the environmental in-vestment.

5. Supply Chain Planning andScheduling

In a supply chain, entities such as suppliers, manufac-turers, distributors, and retailers, can belong to a sin-gle organization or independent organizations. How-ever, the distinction between centralized and de-centralized systems is more properly related to theincentive structures within the chain. At the mostbasic level, in a centralized supply chain, there is acentral planner who makes decisions for the entiresystem, while each entity in a decentralized systemfunctions as an autonomous unit. Decentralized con-trol policies can be easily implemented and analyzedat the local level (function, department, firm, etc.),however coordinated planning of the individual enti-ties in a way that optimizes the value of the overallsupply chain (system) is a difficult undertaking. Re-search tools that are used for planning such systemsinclude network flow models and Mixed Integer Pro-gramming (MIP) models.

Chen, Zhao, and Ball (2002) present an MIP modelwhich allows a decision maker to specify suppliers ofinputs and production resources on a job by job basisin a make to order environment. Their model alsoaccommodates order rejection if the supply chain can-not meet order requirements. The model simulatessettings in which firms accept orders, make tentativepromises, then make final schedules by planningbatches of orders on an hourly, daily, or weekly basis.The authors point out that such policies are quitecommon in firms that gather orders via the internet.They also experiment with the approach using indus-try data to answer questions about how often theorders should be scheduled as a batch.

Souza, Zhao, Chen, and Ball (2004) develop a modelusing data on Toshiba’s global notebook supply chain.This paper uses an MIP model to investigate whethera global manufacturer wants to pass temporary com-ponent price discounts offered by an upstream sup-plier on to the final customers. The analysis accountsfor the fact that this component is included in multipleproducts made in various locations and deliveredthrough a range of channels. The resulting modelincludes as many as a million variables because it isnecessary to account for each path that a part may takethrough the system. The resulting model leads to in-teresting conclusions on the extent to which it is opti-mal to pass raw material discounts on to consumers.For example, short-term raw material discounts mayrequire longer term end-item discounts in order toincrease total supply chain profits. Additionally, the


merits of passing on the discount are intimately re-lated to the elasticity of the demand curve as well asthe correlations of demand across multiple items.

Even when multiple parts of a supply chain arewithin the same organization, it is still attractive tooperate the system in a decentralized way due to theease of implementing site-specific control policies. It isalso easier to evaluate an independent site rather thanan integrated system. These factors drive much re-search in the field. Zhang (1997) studies a decentral-ized assemble-to-order system with multiple compo-nents and multiple products facing correlateddemands. Inventories are kept at the component leveland are used to support service level targets for finalproducts. Each component has a deterministic replen-ishment lead time, and the assembly time for the finalproducts is assumed to be negligible. An independentorder-up-to policy is employed for each componentand a product-type-based priority rule is used to al-locate the components to final products. Given fill ratetargets for final products the author uses an optimiza-tion model to find the optimal order-up-to levels forcomponents.

Andersson, Axsater, and Marklund (1998) apply adecomposition approach to coordinate a system withone central warehouse and an arbitrary number of re-tailers. Two main techniques are employed to enable thedecentralized near-optimal order quantity decisions forthe system. One is cost approximation for each retailervia replacing the random ordering lead time from thewarehouse by its expected value. The other is to modifythe warehouse’s cost structure to include stockout pen-alty costs for late deliveries to the retailers. An iterativeapproach is used starting with a set of expected leadtimes for retailers. It is shown that in the case of normallydistributed lead time demand the existence of a NashEquilibrium for the decomposed problem is guaranteed.The proposed policy is shown to be quite close to opti-mal in several useful cases and has the advantage ofbeing easy to implement.

Urban (2000) analyzes the multi-period inventory or-dering policy of a retailer who is offered a “periodic,stationary commitment ”contract, in which the buyercommits to purchase a certain quantity every period butwill incur a cost each time a change in the order quantityis made. This type of commitment contract aims to re-duce the variability faced by a supplier with low volumeflexibility, by making it costly for the buyer to pass alongthe variability of market demand. A solution approachfor the stochastic demand model is presented and sev-eral specific demand distributions are considered. Forthe case of deterministic demand, two alternative mod-els are employed: one is a mixed-integer linear programand the other is a network model.

6. Teaching Supply ChainManagement

As the demand for supply chain expertise exploded inindustry, many teaching tools have been developed bybusiness and engineering school faculty. Undoubt-edly, the “Beer Game” has gained the most popularityamong the simulation games used as teaching tools inundergraduate as well as graduate programs. Thesetup of the game involves students playing four roleswithin a single supply chain; Retailer, Wholesaler,Distributor, and Factory. Customers’ demand for“Beer” arrives at the retailer. Material flows from up-stream to downstream is often represented by somephysical proxy such as chips or cards. Informationflows upstream through the placement of orders. Ineach period, each player in the chain decides howmuch to order from their respective supplier and thefactory must decide how much to produce. Orderprocessing, transportation, and production lead timesare built into the game. Each member experiences aholding or backlog cost in each period and the objec-tive of the team is to minimize the sum of these costswithout any communication between levels of thechain other then orders moving upstream and mate-rial moving downstream.

While many instructors have students play thegame using some physical tokens to symbolize mate-rial flows, computerized versions of the game facili-tate a wide array of variants. Chen and Samroengraja(2000) present a variant known as of the stationarybeer game. This version of the game models the ma-terial and information flows in a production-distribu-tion channel serving a stationary market where thecustomer demands in different periods are indepen-dent and identically distributed, and each player isaware of the distribution. This modification addressesstudents’ most common criticism of the game (that, inreality, we must have some a’priori sense of demandeven if we do not know it exactly) and retains thegame’s major teaching points. Playing the game usingPC’s or the web allows the instructor to deliver thegame’s lessons without sacrificing much class time.

A computer program was also developed for thepurpose of playing the stationary beer game. Jacobs(2000) describes a internet version of the beer game, inwhich students work at personal computers in a class-room using a web browser to play the game while theinstructor keeps tract of the game using software re-siding on the web server. The instructor determinesthe pace of the game and manually triggers the updateof the system. Besides the obvious advantage of sav-ing time, the internet version of the beer game caneasily incorporate some of the major causes that leadto the bullwhip effect. For example, by inputting al-


ternative demand files, the game can simulate thebuying pattern influenced by price fluctuation.

Johnson and Pyke (2000) argues that teaching sup-ply chain management is much more than repackingtopics taught in “Operations management” or “logis-tics” courses since “integration” is the essential themeof supply chain management. While the planning andmanagement of procurement, conversion and logisticsfunctions are indispensable components, supply chainmanagement emphasizes the importance of integra-tion, which comprises the functional integrationwithin the company (e.g., the integration of opera-tions, logistics, marketing, and finance) as well as thecoordination and collaboration with channel partnersacross company boundaries.

By examining the curricula used by many top engi-neering and graduate business schools for courses insupply chain management, Johnson and Pyke (2000)present a framework consisting of 12 components (ar-eas) of a typical supply chain management course.Besides a description on how to teach each compo-nent, the authors provide extensive lists of cases andbusiness news stories and categorize using the frame-work they propose. Several abbreviated class syllabiare also listed. Vollmann et al. (2000) present ap-proaches for teaching supply chain management spe-cifically to business executives. They argue that thechain focus should start with the customer and workbackwards, instead of starting with the supplier andworking forward. Therefore, they propose the term“demand chain management” to replace “supplychain management” in order to reflect their point ofview. The authors believe that the following four man-agerial concerns are particularly important to execu-tives: flawless execution, a move from supply to de-mand chain management, outsourcing and supplybase development and, implementing demand chainpartnerships. For each concern, the authors provideone or two cases that are appropriate for illustratingthe underlying ideas and concepts. Kopczak and Fran-soo (2000) share their supply chain managementteaching experience through a Master’s level offeringcalled Global Project Coordination Course (GPC) inwhich project teams composed of three students fromeach of two overseas universities execute company-sponsored projects dealing with global supply chainmanagement issues. A detailed course schedule anddiscussion of the logistics involved are reported. Aproject focused on the reverse logistics at Quantum isused to illustrate the scope and nature of the projectsin the course. The authors conclude that, in conjunc-tion with lecture and case-based approaches, the GPCcourse successfully trained students’ project manage-ment and consulting skills. The course also enhancedthe students’ knowledge and understanding of supplychain management and information system theory.

Finally, the course helped enable the students to applythe theory in a real setting and to work effectively in aglobal, cross-cultural project team.

Mehring (2000) describes another supply chain simu-lation game: the Siemens Brief Case Game (BCG), usedextensively within Siemens. Compared with the beergame, the BCG has more details and complexity, whichenables the instructors to develop learning exercises thatfocus on a wider range of supply chain managementissues. The BCG supply chain consists of 9 activities:receiving a customer order, sales order processing, plantorder processing, plant procurement, subassembly, finalassembly, managing plant traffic and warehousing,managing sales traffic and warehousing, and dealingwith a supplier. The paper describes two typical exer-cises. One targets the undergraduate audience and aimsto provide a concrete example of typical activities in asupply chain and their interactions. The other is tailoredfor a graduate level audience and leads students to dis-cover what creates a need for coordination, what activi-ties require coordination, and what types of methods aremore likely to be effective.

Anderson and Morrice (2000) propose a simulationgame designed to teach service-oriented supply chainmanagement principles. The authors argue that the BeerGame is not an accurate reflection of many service set-tings because service providers typically cannot holdinventory and can only manage backlogs through capac-ity adjustments. This game simulates the processing of amortgage application. Each application passes through 4stages: initial processing, credit checking, surveying theproperty that the applicant wants to purchase, and titlechecking. During each period, each player sees the back-log of jobs at his position. He then makes decisions tohire or fire in an effort to manage this backlog. Lags arebuilt into the game to simulate the time it takes to findnew workers and train them, or the notice that is givento workers being released. The objective of the team of 4positions is to minimize the total costs of capacity ad-justments and service delays. The behavior of the systemcan become quite complex due to the lags between ca-pacity adjustment decisions and their realization at eachposition. Campbell, Goentzel, and Savelsbergh (2000)shares their experiences with use of supply chain man-agement software in teaching, and presents many usefulideas on how to integrate popular software into theclassroom.

7. Supply Chain Coordination:Information Sharing, Incentives,and Contracts

Actions or approaches which lead supply chain part-ners to act in ways that are best for the chain as awhole are known as supply chain coordination. One ofthe most obvious means to achieve this coordination is


the use of contracts, whose terms lead to the desiredactions. Such a mechanism is valuable when theseactions are not in the best interest of the individualmembers of the chain unless some type of transferpayment is built into the scenario. Several works inPOM have attempted to show the effectiveness ofvarious supply chain coordination mechanisms andcontracts. This is particularly important in a competi-tive environment in which entire supply chains arecompeting for customers.

POM included one of the earliest contributions onincentive alignment in decentralized supply chains.Ernst and Cohen (1992) introduces a stochastic modelof a decentralized distribution system consisting of adealer (who controls the stock policy) and a manufac-turer (who is responsible for the delivery of the prod-uct through regular or expedite shipment modes).They derive optimal inventory control policies underthree different scenarios: the dealer is dominant, themanufacturer is dominant, and the system is undercentralized control. It is shown that the expected profitfor a dominant decision maker is higher under decen-tralized control than the expected profit he can get inthe centralized system.

Extensive reviews of the more recent literature onsupply chain coordination are available in Whang(1995), Cachon (1998), Lariviere (1998), Tsay, Nah-mias, and Agrawal (1998) and Cachon (2003). Supplychain coordination with internal markets is discussedin Porteus and Whang (1991) and Kouvelis andLariviere (2000). For illustrative purposes, let us con-sider a simple form of the most commonly treatedclass of problems in the supply chain coordinationarea. Consider a simple chain with one supplier andone retailer. There is one selling season with uncertaindemand, and a single opportunity for the retailer toorder before the selling season begins. If the contractconsists only of a wholesale cost per unit, it is easy toshow that the retailer does not order enough to max-imize the supply chain’s total profit because he ig-nores the impact of his action on the supplier’s profit.Hence, coordination requires that the retailer be givenan incentive to increase his order. This incentive canfollow from a wide variety of contract types andterms. For example, the supplier can offer the buyer awholesale price plus some adjustment that dependson realized demand, such as a promise to buy backleftover inventory. In effect, this presents the retailerwith a salvage value which increases his optimal orderquantity. See Pasternack (1985) for a detailed analysisof buyback contracts.

Another commonly used approach involves a quan-tity flexibility contract. Under quantity flexibility con-tracts, the retailer is allowed to change the quantityordered from the supplier after observing early de-mand or a demand signal. These contracts are similar

to buy-back contracts in that the supplier shares someof the risks of having excess inventory with the re-tailer. Quantity flexibility contracts may induce theretailer to purchase larger quantities, thereby increas-ing total supply chain profits. See Tsay (1999) for adetailed study of supply chain coordination with thistype of contract. Another common approach existswhen the supplier charges some amount per unit pur-chased and the retailer gives the supplier a percentageof his revenue. This effectively splits both the benefitsof coordination and the risks related to market uncer-tainty. Cachon and Lariviere (2005) provide an analy-sis of these contracts in a general setting. A relatedapproach is known as the sales-rebate contract. Underthis approach the supplier charges an amount per unitpurchased, but then gives the retailer a rebate per unitsold above a threshold level.

A more complex setting exists when retail prices arenot taken as given. If demand is a function of retailprices, then the contract types just discussed may notcoordinate the chain, especially if salvage values orgoodwill penalties are present. For these settings aprice-discount contract is often used. It is essentially abuyback contract with parameters that are set onlyafter the retailer chooses his price.

The general topic of supply chain coordination wasthe focus of a special issue of POM in 2004. As part ofthis issue, Boyaci and Gallego (2004) present a game-theoretic approach to modeling competition between2 supply chains consisting of one wholesaler and oneretailer. In the depicted scenario, market share isbased strictly on customer service levels. The resultingmodel shows that supply chain coordination is anequilibrium strategy for both firms, but not necessarilyone that increases supply chain profits. This is truebecause, in some instances where service is the onlyorder winner, the consumers extract all of the benefitsresulting from coordination. This suggests that coor-dination may be paramount to survival even thoughindustry profits may not rise as a result of the coordi-nating activity.

Gerchak and Wang (2004) study two distinct con-tract schemes involving an assembler and its suppliersin a decentralized assembly system with uncertaindemand. One scheme combines vendor-managed in-ventory (VMI) with revenue sharing. The otherscheme is a simpler wholesale-price driven contract.In the VMI system the downstream assembler offersshares to the suppliers who then select the compo-nents production quantities. It is shown that with theassembler as the leader, the revenue shares are chosensuch that all suppliers produce the same quantity andthis amount is independent of the other supplier’sproduction cost. To achieve channel coordination, theauthors also propose a revenue-plus-surplus-subsidyscheme where, in addition to a share of revenue, a


supplier is partially paid by the assembler for hisunsold delivered components. In the more conven-tional wholesale price contract, the suppliers first si-multaneously choose their individual componentwholesale prices, then the assembler chooses the orderquantity from each suppliers. It is shown that theproduction quantity and the system profit are decreas-ing in the number of suppliers. Similarly as in thesingle-supplier-single-buyer system, by adding a buy-back policy to the wholesale price contract, the multi-supplier, decentralized assembly system can be coor-dinated. The authors also provide one example toshow that the VMI with revenue sharing system dom-inates the wholesale-price-based system in terms ofthe channel profit.

An asymmetric incentive structure can follow fromthe way in which the damage resulting from a lost saleis split among the channel partners. Kraiselburd,Narayanan, and Raman (2004) consider such a case viaanalysis of a scenario in which the retailer carriessubstitutable products from different manufacturersand customers make substitution decisions. In thiscase, an individual wholesaler feels the impact of lostsales even though no impact is felt by the retailer. Theauthors propose VMI as a solution to such a misalign-ment of incentives problem, but the performance ofthis approach depends on the manufacturer’s effortwhich is ‘non-contractible’ meaning contracts regard-ing this level of effort are not fully enforceable becausethe retailer lacks perfect information. This leads toscenarios in which VMI can actually increase channelinefficiency when there is a high likelihood of substi-tution and the manufacturer’s effort has little impacton product demand.

The coordination of supply chains with risk-averseagents is studied by Gan, Sethi, and Yan (2004). Basedon the concepts used in group decision theory, thiswork defines a coordinating contract as one that leadsto a Pareto-optimal solution acceptable to each agentin the supply chain. The decisions faced by the groupinclude external choices and the internal split of thepayoff among the group members. In supply chainswith risk-averse agents, each Pareto-optimal solutionis characterized by a pair: the optimal channel’s exter-nal action and the corresponding Pareto-optimal shar-ing rule. A coordinating contract is then defined asone with a specific set of parameters that achieves theselected Pareto-optimal solution. The authors demon-strate how to find the set of Pareto-optimal solutionsand how to design a contract to achieve the solutionsfor three specific cases: (1) the supplier is risk neutraland the retailer maximizes his expected profit subjectto a downside risk; (2) the supplier and the retailereach maximizes his own mean-variance tradeoff; (3)the suppler and the retailer each maximizes his ownexpected utility. For general concave utility functions,

it is not always possible to get the Pareto-sharing rulein closed form. Therefore, the authors consider thespecial case of exponential utility functions, for whichthe Pareto-optimal sharing rule and the optimal exter-nal action are derived and explicit coordinating con-tracts (revenue sharing contract and buy-back con-tract) are designed.

Poundarikapuram and Veeramani (2004) present adistributed decision-making framework for the play-ers in a supply chain or a private e-marketplace inwhich players have limited information about eachother but they do collaborative planning. The playershave a global goal that is dependent on some commonvariables and their unique local private objectives thatare dependent on local variables. Based on the IntegerL-shaped method, the authors propose an iterativeprocedure to address the collaborative decision mak-ing problem in the decentralized system. In this ap-proach a master problem (modeled as a mixed-integerprogram) is solved to propose global solutions andeach player solves his local problems (modeled asLinear programs) to construct cuts on the feasiblespace of the master problem. This model enables thedistributed decision-making process to achieve near-optimal global performance if the players are willingto disclose limited information.

Zhang (2002) considers a single manufacturer andtwo competing retailers. The retailers may compete byselecting either prices or output quantities. They alsomust decide whether to share signals regarding de-mand prior to receiving the signal. The major resultsare that the manufacturer is always better off withmore information and that each retailer is damaged ifhe shares information, while his competitor does not.Consequently, the unique equilibrium is to avoid in-formation sharing in the supply chain unless somemechanism exists to induce sharing such as a sidepayment or a guarantee that your rival will sharetruthful information along with you.

A quantity discount can serve as an inventory coor-dination mechanism between buyers and suppliers.Shin and Benton (2004) study the effectiveness ofquantity discounts under different conditions. Theyshow that the link between quantity discounts and theperformance of the chain is influenced by severalother factors including the variability of demand, therelative inventory cost structures, and the buyer’s eco-nomic reorder intervals.

In the context of high-tech industries, Erkoc and Wu(2005) study capacity reservation contracts between amanufacturer, who faces convex capacity expansioncost, and her OEM customer. They share the stochasticdemand information but the OEM places orders whenthe demand is realized. In the absence of a capacityreservation, the manufacturer bears all demand risk.To mitigate the manufacturer’s capacity expansion


risk and to encourage the manufacturer to expandcapacity aggressively, the OEM may engage in a ca-pacity reservation contract, in which, the buyer re-serves capacities upfront by paying a deductible fee.However, such a contract is not always desirable forboth parties. There is a threshold for the reservationfee that the buyer will accept, and such a reservation isonly beneficial for the supplier when the product mar-gin is sufficiently high. Consequently, the fully de-ductible reservation fee contract rarely coordinates thechain. The authors propose two coordination mecha-nisms: a partially deductible reservation fee contract,and a cost-sharing contract. Supply chain coordinationis also discussed for cases in which the manufacturerunder-expands capacity, and cases where demand in-formation can be partially updated.

8. Multi-Channel CoordinationChallenges: Coordinating Offlineand Online Procurement andDistribution

The emergence and development of the internet, andits role in information management, marketing, andsupply chain coordination has opened the door to ahost of research and business opportunities. The in-ternet offers manufacturers and retailers new avenuesto conduct their business and most, if not all estab-lished firms are at least considering expanding ontheir historic business practices and relationships toinclude some internet enabled component. In this sec-tion, we will focus on the unique issues that arisewhen one needs to coordinate the activities of anonline and offline approach to procurement, sales, anddelivery. We will also highlight a few research effortsin these areas as examples of the styles of analysisused so far and types of questions present which havenot been fully addressed.

In our exposition, we follow the classificationscheme of the review paper by Cattani, Gilland, andSwaminathan (2003). Focusing mainly on the manu-facturer, they identify three areas of research focus;procurement, pricing, and distribution and fulfill-ment. Specifically, firms find themselves evaluatingoptions to use the internet in procurement to eitherenhance or replace traditional practices. POM pub-lished one of the earliest works on the topic by Peleg,Lee, and Hausman (2002). This work develops a two-period model with uncertain demand in which thedecision maker has options to use long-term contracts,simple auctions, or some combination of the two. Theattraction of the long term contract is that the supplieris willing to promise reduced prices in the secondperiod of a two-period game, if the buyer commits atthe start of period 1 to some minimum order quantityin period two. This is rationalized as a learning effect

in that the promised price in the second period isp(1 � �), where p is the first period price, and � isessentially a discount rate which reflects supplierlearning. The attraction of using auctions is the oppor-tunity to get lower prices. However, the manufacturerdoes not know a’priori what price will be realized inthe auction. The model assumes that he does havesome prior distribution describing the realized auctionprice. The buyer can also apply a combined strategy inwhich he has a contract to cover most of his expectedneeds but has the option to use the auction to coverany shortfall or buy additional units from the supplierat the agreed upon price.

Peleg et al. (2004) shows that there exists a value of� beyond which it is optimal for the buyer to prefer astrategic long term partnership as opposed to utilizingan auction in the second period. They also show thatthe combined strategy of using a contract and anauction in the second period to meet higher than ex-pected demand may be either superior or inferior to apure auction strategy, depending upon the distribu-tion of prices obtained from the auction. This modelalso allows the authors to consider the impact of thenumber of players in the auction in that more playersshould result in lower prices, but including additionalplayers involves some coordination cost. They con-clude that the optimal number of suppliers dependson the variance of demand in the first period, and itincreases with the mean demand level in the secondperiod.

The growing literature on supply contracts whichintegrates markets in various contexts (storable vs.non-storable goods, various types of long term supplycontracts, negotiation etc.) is surveyed in Kleindorferand Wu (2003). Among the seminal works in this areais Lee and Whang (2002). This paper studies howinternet based surplus (or secondary) markets serve asmechanisms through which inventory can be pooledand balanced among participants in reaction to de-mand information. The market price in their two-period model is determined as the equilibrium pricethat clears the market under the assumption of a largenumber of market participants. They show that anopen secondary market increases the allocation effi-ciency though it has an unclear effect on the monop-olistic supplier’s sales. Recent work of Milner andKouvelis (2006) elucidates some aspects of how spec-ulative online exchanges with a small number of par-ticipants might behave and the impact they have onthe use of long term contracts for supply. The authorsshow that participating buyers accrue network bene-fits as the number of participating firms increasesthrough the inventory pooling effects, resulting in re-duced cost for them. However, a strategically actingsupplier will counteract such benefits by restrictingavailability of goods to the spot market, sacrificing


short term spot market revenue for long term contractvolume.

The work of Granot and Sosic (2005) considers thebenefits of joining a consortia of industry players,using a model with three firms. In this work, each firmmay form alliances with none, one, or two other firms.A variety of issues are involved which increase thecomplexity of the decision. This work focuses largelyon the substitutability of the products of the firmsconsidering the formation of an alliance. If the prod-ucts are not substitutes, then the optimal strategy is tojoin and the authors argue that this outcome is rela-tively easy to sustain. On the other hand, if two firmshave substitute products then the third firm is likely tovalue the alliance differently then either of them andsome 2 firm alliances in which a seller of a substitut-able product connects with the party not making sucha product may be more stable. In settings in which allthree players have substitutable products, a three firmalliance may be stable, but it is not likely. This followsfrom the fact that the benefit of such an alliance is veryunlikely to be uniformly distributed across the threefirms. The problem is also complicated by the fact thatfirm A may wish to form an alliance with firm Bsimply to keep firm C from forming such a 2-firmalliance. The authors point out that their work sug-gests much future research that considers differentdemand functions and information states for the play-ers involved.

The consideration of a manufacturer exploring orbeing presented with a new “e-tail” channel presentsat least four possibilities. The manufacturer may sim-ply find himself as the supplier to two competingchannels (an E-tailer, and a retailer.) He may decide toforward integrate by creating an e-tail channel whichcompetes with a pre-existing retail channel. He mayfind himself serving a combined online and offlinere-seller. In some cases, he may even decide to becomeboth a retailer and an e-tailer himself. These four casespresent a plethora of new research topics. In an effortto better frame the broad research issues involved, wediscuss some representative work outside POM in thearea, and then proceed to position POM publicationcontributions to this field.

Considering a manufacturer supplying a retailercompeting with an e-tailer, the issue of product pric-ing is paramount. One obvious concern of manufac-turers is that the end users’ added ability to compareprices on line will result in reduced profits for theentire chain as buyers select the low cost provider formore and more goods, especially those that are notobviously differentiated between online and offlineofferings. While much work remains to be done ana-lyzing the ultimate impact of internet sales on manu-facturers, some research has suggested that the fearthat all buyers will simply select on price has proven

to be overblown. Brynjolfsson and Smith (2000) con-sider the prices paid for items easily advertised online.This work considers the pricing and sale of CD’s andbooks. They found that the internet offered a pricediscount of 15.5% over traditional channels on books,and 16.1% on CD’s, and that these discounts are 9%and 13%, respectively, after considering sales tax,shipping and handling, and transportation charges.This work also found that price changes are muchmore common with online retailers as opposed totraditional brick and mortar stores.

The work of Smith and Brynjolfsson (2001) extendedthis analysis and found that customers do not alwaysopt for the lowest price. In fact, when consideringonline book sales they found that the majority of cus-tomers do not chose the lowest price offered evenwhen they are fully informed about the range of pricesavailable. Among the customers that do not chose thelowest priced version of the product, the average priceof the selected offer is 20.4% higher then the lowestprice available. Surprisingly, they found that buyerswere willing to pay $1.72 more for an item from one ofthe “big three” online booksellers (Amazon, Barnesand Noble, and Borders) when compared with othervendors. The authors suggest that this result showsthat the online vendor is perceived as a brand by thepotential buyers, and they are willing to pay a pre-mium for the brand if this is associated with somenotion of quality, reliability, or performance.

The issues which determine the optimal price be-come even more complex when the same item is beingoffered online and offline by the same firm. This“bricks and clicks” approach raises several difficultquestions whose complete answers are still forthcom-ing from academic researchers. For example, Lal andSavary (1999) model scenarios in which competingproducts are offered in both retail stores and online.They assume that some attributes of the products areeasily conveyed digitally while others are not. Theyalso account for the cost of searching for items withwhich the customers have insufficient prior experi-ence. The customers can get digital information onlinebut must travel to stores to get non-digital informationsuch as feel and fit. When no internet outlet exists,customers must get all information by store searches.After the internet channel is introduced, customerscan get some info with much less effort. This workshows that, depending on the customer search cost,the addition of an internet option may result in lower,or higher average prices paid. Thus the intuition thatan internet outlet must drive prices down does notnecessarily hold. It depends on how much relevantinformation can be conveyed online, the customersearch cost, and the probability that the search willresult in an unfavorable evaluation.

The work of Chiang, Chhajed and Hess (2003) con-


siders a manufacturer selling through traditional retailchannels while contemplating a move to integrate for-ward by creating an online option for end users topurchase their products. They focus on the issue ofdouble marginalization that occurs when both themanufacturer and the retailer behave in a manner thatis locally optimal. They consider a customer specificparameter 0 � � � 1, such that the customer’s utilityfor the product purchased online is � times the utilityreceived when purchased in a retail store. This workshows that differing equilibria exist depending on thevalue of �. When this value is low, the internet offeringhas no impact on the supply chain. There exists arange of moderate values of � such that the presence ofthe internet channel serves to drive prices down atboth the manufacturer and retail levels even thoughno sales actually take place online. In this case, themargins of both the manufacturer and retailer areaffected but both parties may actually be more profit-able as the reduced price increases sales. Finally, forrelatively high values of �, the manufacturer’s profitsincrease but the retailer’s profits decline as the onlinechannel becomes a very effective competitor with theretailer.

An influential dual channel conflict and coordina-tion work in the POM journal is by Tsay and Agrawal(2004). This paper develops a model where a tradi-tional and an internet retailer compete against eachother and the customer purchase decisions are drivenby both price and the effort that each channel putsforth. Their findings include the surprising result thatthe apparent conflict created is not necessarily detri-mental to the retailer because there is asymmetry inthe cost structures of the different channels. Thus,both a reseller and its supplier can be better off whenthe supplier implements a direct sales channel. Thisfollows from the fact that the price charged by themanufacturer to the retailer is driven down by thecompetition, but the determination of the optimalprice at the retail level is more complex because henow has more options in offering a response.

A consideration of channel coordination involvinginternet outlets would be incomplete without somediscussion of distribution and fulfillment issues thatarise in this more complex environment. One excellentexample of the work in this area was recently pub-lished in POM. Cattani and Souza (2002) consider ascenario in which the firm takes advantage of the factthat internet customers display a greater willingnessto wait for the delivery of a product when comparedwith retail customers. This is evidenced by the factthat the internet customer must wait for product de-livery in any case, and he does not know the inventorystatus of the item at the time of purchase. This sug-gests that a seller offering both retail and internetchannels should dynamically allocate inventory be-

tween them. This work suggests a continuous reviewinventory management policy for a single product. Itmodifies the standard approach to such inventorymanagement systems by determining three criticalvalues. Specifically, they determine a reorder point, atarget inventory position, and a third value which isthe cutoff point for internet sales. In this system, onceinventory drops below a determined level, all internetorders are backlogged even if inventory still exists inthe system and is available to retail shoppers. Theresults show that the increase in the seller’s profitsdepends on the fraction of total demand that is in theform of retail sales. Obviously, if this fraction is closeto 0% or 100%, then the policy has no application.However, if this fraction is significantly different fromthese end points, then the seller’s profits may increaseby as much as 7%. Recently research work on ration-ing of inventories with dynamic discounts appeared inDing, Kouvelis, and Milner (2006). The presentedmodel considers a single period stochastic demandproblem with variable ordering costs and multipledemand classes. The period is divided into multiplestages, allowing updating of demand information andmaking inventory allocation decisions. The authorsargue on the effectiveness of rationing policies in low-ering inventory levels and improving profitability forsuch settings, as well the use of dynamic discounts to“keep” lower class customers around and meet servicelevel constraints.

9. Design for Supply ChainManagement: Postponement andProduct Variety

One can use the design of a product, as well as thedesign of the manufacturing and supply chain pro-cess, to delay the point of product differentiation suchthat it is closer, both in terms of physical location andtime, to the final customer demand. This increases thefirm’s ability to handle the uncertain and continuouslychanging demand for one or multiple products. Thisapproach was first termed postponement by Alderson(1950). The value of postponement lies in the reduc-tion of inventory related costs due to a better matchingof supply to demand. For complete reviews of impor-tant works in the design for supply chain managementand postponement literature we refer readers to Lee(1993), Lee (1996), Garg and Tang (1997), Garg and Lee(1998), and Swaminathan and Lee (2003).

Swaninathan and Lee (2003) cite three specific en-ablers of postponement—process standardization,process resequencing, and component standardiza-tion. One classic process standardization example isfound in the description of Hewlett Packard’s DeskJetPrinter business given in Lee, Billington, and Carter(1993). This printer line had distribution centers in the


U.S., Europe, and the Far East. The earlier strategy hadbeen to make region-specific models all in the u.s. andto ship these completed items to each area as finishedgoods. The variability of demand within each region,and the one month lead time in shipping led to sig-nificant costs for the system. HP decided to redesignthe product so that standardized steps in the produc-tion process could all be done in the U.S., but stepswhich make a model specific to one area such aspower supply installation, and inclusion of manualswas deferred until after customer orders arrive in eachregion. The documented success of this approach atHP became the prime instigator of subsequent designfor supply chain management practices at the com-pany as discussed in Feitzinger and Lee (1997), Leeand Sasser (1995), and Lee et al. (1997).

A general model of this type of arrangement ispresented in Garg and Tang (1997). In this work, theauthors consider a system with two points of differ-entiation. The first is the family differentiation point,and the second is the product differentiation point. Inother words, the process consists of three stages. Instage 1, steps common to the production process of theentire group of products are undertaken. In stage 2,steps which differentiate one family from another aretaken. The steps which define a specific product aredeferred until stage three. This work shows that wheninventory is stored at the end of each stage, significantsavings in inventory costs can be achieved when com-pared to a system that only stores finished products.They also discuss problem characteristics that increasethe value of later postponement such as when productdemand across a family is more negatively correlated.

Related work appeared in POM in Graman andMagazine (2002). This model considers postponementuntil the second of a 2-stage process, with capacityconstraints for each stage where the differentiatingsteps take place in the second stage and inventory canbe stored between the two stages. This is sometimesreferred to as a ‘vanilla box problem’ because genericmodules are produced in stage one and demand isrealized for specific models after the additional stepsof stage 2. This model derives analytical expressionsfor service measures and inventory levels. The authorsalso use a numerical study to show that very littlepostponement capacity can provide virtually all of thebenefits related to inventory reduction. Some of theoriginal work on delayed product differentiationthrough vanilla boxes appeared in Swaminathan andTayur (1998).

A story frequently used in MBA classrooms whichdisplays the value of process resequencing can befound in Dapiran’s (1992) description of Benetton. Inthis scenario, sweaters had traditionally been pro-duced by dyeing yarn to the appropriate colors, andthen weaving them into the selected styles and sizes.

The managers at Benetton observed that the majorsource of uncertainty was not over sizes and styles asmuch as it was over color. In other words, which colorwould be “hot” this season was very difficult to pre-dict. As a result, Benetton found it profitable to changeits traditional production sequence to allow for theweaving of uncolored sweaters, thus postponing thepoint in the process at which a unit became differen-tiated by color.

Lee and Tang (1998) formally considers a simplifiedversion of the problem by modeling a two-stage sys-tem where a distinct feature is introduced into theproduct at each stage. Specifically, they consider twostages and two possible features at each stage. Thus,there are four possible products available to a cus-tomer. In this formulation, changing the sequencedoes not alter either the raw material or finishedgoods inventory levels. However, it may alter thevariance of the buffer levels between the stages. Thishighlights the fact that the notions of holding andshortage costs for finished goods may not reflect all ofthe costs associated with managing inventory in thesupply chain. This paper argues that the cost of man-ufacturing, such as overtime charges or the disruptioncaused by expediting, is often linked to the variancesof inventory levels and production requirementswithin the system.

An excellent example of the use of a componentstandardization strategy involves Lucent Technolo-gies and is relayed in Hoyt and Lopez-Tello (2001). In1998, Lucent recognized a huge sales opportunity fortelecommunications equipment in Saudi Arabia.However, to take advantage of this opportunity, Lu-cent needed to be able to offer a short lead time thatwould make it impractical to perform the usual de-tailed site engineering work that had always precededproduct configuration. In addition, the Spanish plantwhich was closest to the customer did not have thecapacity to fill the order, even if all specifications wereknown at once. Lucent ultimately decided to redesignthe product so that it had common building blocks.This enabled them to pre-build these common blockswithout full knowledge of the final configuration, andto use capacity in other plants to help generate therequired output.

These approaches are closely related to earlier workfocused on the benefits of component commonality.The bulk of the formal modeling which deals withcomponent commonality focuses on inventory cost atthe component level. One notable exception is in VanMieghem (2004). This model considers two productswhere each product is assembled from two compo-nents and both common and product specific compo-nents are stocked. This work derives conditions underwhich commonality should be adopted. This determi-nation is driven by the level and nature of demand


correlation for the two end items, and the cost re-quired to achieve commonality. When this cost is high,postponement and the use of common componentsmay not be attractive. A strategy in which each prod-uct is assembled using a common component is onlyoptimal if the cost to manage components is increasingin the number of different components in play. Fi-nally, this work shows that while the value of thecommonality strategy decreases in the correlation be-tween product demands, commonality is optimal evenwhen the product demands move in lockstep if thereis a sufficient profit differential between the two prod-ucts.

10. Operational Hedging and RiskManagement in Supply Chains

Due to the inherent uncertainties/risks in SCM, Sup-ply Chain Risk Management (SCRM) was virtuallyborn alongside the concept of supply chain manage-ment itself. Even though many articles in the literaturedo not articulate the term “risk management” they aremotivated by the need to deal with the management ofuncertain demand, uncertain supply, and uncertaincosts. Supply chain risks can be categorized into twobasic levels: operational risks, and disruption risks(Tang 2006). Kleindorfer and Saad (2005) refer to themas supply-demand coordination risks and disruptionrisks. Operational risks are directly associated with theday to day management of the supply chain, while theterm “disruption risks” is generally reserved for nat-ural or purposeful man-made disasters (earthquakes,hurricanes, terrorism, etc.).

Tang (2006) offers an extensive and excellent reviewof various quantitative models in the literature dealingwith the risks associated with supply chains and pre-sents a unified framework for classifying SCRM arti-cles. Tang (2006) first defines SCRM as “the manage-ment of supply chain risks through coordination orcollaboration among the supply chain partners so as toensure profitability and continuity.” By “profitabil-ity,” the above definition implicitly assumes that thesupply chain members are risk-neutral and conse-quently, this work basically reviews those articles thatprovide cost effective solutions for managing supplychain risks. Based on the above definition of SCRM,Tang (2006) classifies the risk management ap-proaches appearing in the supply chain managementliterature into four groups: supply management, de-mand management, product management, and infor-mation management. As implied by the definition,these four approaches are all intended to improvesupply chain operations via coordination or collabo-ration among supply chain members. Accepting thedefinition, we would like to highlight the relevant

issues in several approaches that are used to mitigatesupply chain risk.

There are five intertwined issues in supply manage-ment: supply network design, supplier relationships,supplier selection, supplier order allocation, and sup-ply contracting. Tang (2006) discusses three ap-proaches for demand management: shifting demandacross time, shifting demand across markets, andshifting demand across products. That work also high-lights three product management strategies: post-ponement, process sequencing, and product substitu-tion. The author argues that postponement is aneffective way to reduce variability in a supply chain.Researchers note that this variability can also be re-duced by reversing the sequence of manufacturingprocesses in a supply chain. An example of using theprocess sequencing strategy is the reengineering effortat Benetton, which pioneered the knit-first-dye-laterprocess instead of the traditional dye-first-knit-latersequence in the woolen garment industry. For infor-mation management, Tang (2006) classifies the workinto strategies for fashion products (shorter life cycleand higher demand uncertainty) and strategies forfunctional products. The author reviews informationsharing, vendor managed inventory, and collaborativeforecasting. It is also pointed out that there is a lack ofquantitative models which consider disruption risksin an explicit manner and therefore the author pro-poses several ideas for future research to close the gapbetween the literature and actual practices.

Note that few works reviewed by Tang (2006) ex-plicitly address risk identification, risk measurement,or risk hedging. It seems fair to say that Tang (2006)provides an operational perspective of SCRM. How-ever, many firms are deeply concerned about the vari-ability of their profits and need insights regarding themanagement of this variability. Substantial academicresearch in the corporate finance literature has ex-plained why firms should hedge risks. Violations ofthe assumptions underlying the Modigliani-Miller(Modigliani and Miller (1958) irrelevance theoremsmotivate corporate risk management. From a practicalperspective risk management is indeed relevant dueto a number of factors including tax laws, financialdistress costs, managerial risk aversion (Smith andStulz 1985), capital market imperfection and costs in-curred when accessing external capital markets (Froot,Scharfstein, and Stein 1993). Consequently, we haveseen the rigorous development of risk managementconcepts and techniques in the finance literature in-cluding both the quantification of risks and the devel-opment of risk hedging tools in the past two decades.Recently, there has been rapidly growing recognitionthat risk management concepts and techniques devel-oped in the finance field can be applied to supply


chain risk management. This has spawned severalnew research streams in operations management.

One such research stream is the incorporation of asupply chain agent’s risk aversion into operationaldecision rules. A fundamental theoretical frameworkfor modeling risk-sensitive decision makers is ex-pected utility theory (see Mas-Colell, Whinston, andGreen 1995, Chapter 6). The literature in financialeconomics offers a quite general and analytically trac-table approach based on the mean-variance modeldeveloped by Markowitz (see Markowitz 1952;Markowitz 1987; and others). Both the general utilityfunction and mean-variance approach have been em-ployed in risk-aversion inventory management mod-els (see Bouakiz and Sobel 1992; Chen and Federgruen(2000). However, there are fewer papers in supplychain management that incorporate risk aversion. Sev-eral papers appearing in POM contribute to this bur-geoning research direction. Gan, Sethi, and Yan (2004)first consider the supply chain coordination problemwith risk-averse agents. In a companion paper (Gan,Sethi, and Yan 2005), they specialize the definition ofcoordination for a supply chain with a risk-neutralsupplier and a downside-risk constrained newsven-dor. The authors illustrate that the buy-back contractsand revenue sharing contracts that coordinate the sup-ply chain in the risk-neutral case will not always co-ordinate the channel in the presence of risk aversion ifthe downside risk of the retailer is higher than hisacceptable level. In this case, the retailer prefers toorder less than the coordinating quantity. A risk-shar-ing contract then is proposed, based on the buy-back/revenue-sharing contract. The basic idea of the risk-sharing contract is to provide the downside protectionto the retailer by refunding a certain amount of unsoldunits. Some other notable research in the use of risksharing contracts between retailers and suppliers inthe presence of price uncertainty appeared in Li andKouvelis (1999).

Another rising research stream explores the value ofoperational hedging, as well as the benefit of integrat-ing operational hedging and financial hedging in thecontext of a global supply chain. As noted in Boyabatliand Toktay (2004), there are two essentially similardefinitions of operational hedging in the operationsmanagement literature. One stems from a real optionsview (Huchzermeier and Cohen 1996), while anotherbuilds upon a counterbalancing-action view (VanMieghem 2003). The real options view considers theoperational hedging strategies as real options that areexercised in response to demand, price, and exchangerate contingencies. A variety of types of operationalhedging strategies have been explored in the litera-ture. Kogut and Kulatilaka (1994) consider the oper-ating flexibility to shift production between two man-ufacturing plants located in different countries under

exchange rate uncertainty. Huchzermeier and Cohen(1996) explores the value of global manufacturingstrategy options including international supply flexi-bility and manufacturing and distribution flexibilitiesvia maintaining multiple plants globally and switch-ing production and distribution among the global sup-ply chain network. Ding, Dong, and Kouvelis (2006)study the strategy of postponing the allocation of ca-pacity to specific markets until currency and demanduncertainties are resolved.

Though it is clearly recognized that financial andoperational options should be integrated in order toeffectively manage exposure to exchange rate fluctu-ations (Hodder 1982; Flood and Lessard (1986), fewquantitative models exist to help accomplish this. Onesuch work is Ding, Dong, and Kouvelis (2006). Thispaper develops a structural model to integrate thecapacity allocation option with financial hedging strat-egies for a risk-averse multinational firm. It is as-sumed that the firm uses a production facility in itshome country and sells to markets in both its homecountry as well as a foreign market. Their investiga-tions show that both the operational flexibility andfinancial hedging have an impact on value-creationand the control of the variance of the firm’s profits.The firm with an allocation option tends to increaseoverall capacity after adopting financial hedging.Firms using financial hedging are less affected byvolatilities and their own risk attitude when comparedto firms not using financial hedging. Dong, Kouvelis,and Su (2006) expand this burgeoning research line byincorporating the competitive exposure faced by aglobal firm which sells to both a domestic market anda foreign market while competing with a local incum-bent in the foreign market. The authors compare threeoperational strategies: one with involves no opera-tional flexibility but assumes matching currency foot-prints (“natural hedge.”) The second includes post-ponement flexibility. The third strategy includes bothpostponement and allocation flexibility. It is shownthat the natural hedge strategy is outperformed bystrategies which take advantage of operational flexi-bility in terms of both profit and downside risk. Thiswork also finds that operational flexibility will some-times benefit the competitor in the sense that theglobal firm may decide not to participate in the foreignmarket when the exchange rate is not favorable.

Several papers in POM address the issue of thesupply chain disruption risk. Hendricks and Singhal(2005) conduct an empirical analysis of the effect ofsupply chain disruptions on long-term stock price per-formance and equity risk of the firm based on a sam-ple of 827 disruption announcements made during1989 to 2000. They found that the average abnormalstock return of firms that experienced disruptions isnearly –40% and furthermore, firms do not quickly


recover from the negative effects of disruptions. Klein-dorfer and Saad (2005) offer a conceptual frameworkfor managing disruption risks in supply chains. Thiswork discusses the need for continuous and concur-rent efforts along three dimensions. They argue thatfirms must focus on “specifying” the sources of dis-ruption risks and vulnerabilities. Firms also need toquantify the risks through a disciplined approach torisk “Assessment.” Finally, the firms need to designpolicies and plan actions which serve to “Mitigate”these risks. The author’s elaboration on their frame-work is supplemented with data from an empiricalstudy on the U.S. Chemical industry.

Martinez-de-Albeniz and Simchi-Levi (2005) con-sider the sourcing problem for commodity productsthat are supplied by a variety of suppliers as well as aspot market. A multi-period inventory managementmodel with uncertain demand, a supply contract inwhich the purchasing cost is convex in the purchasequantity, and a spot market is analyzed. The authorsspecialize the model to consider the case of a portfolioconsisting of option contracts and the optimal replen-ishment policy is characterized for a selected portfolioof option contracts.

11. Future Research DirectionsWhile we have made every effort to consider a rich setof topics thus far in this review, we must concede thatwe have omitted a great many research streams sim-ply due to the lack of space. However, we would besorely remiss if we did not at least mention three moreemerging areas in the field. Recent months have seena surge in the efforts of many researchers in the area ofrisk management related specifically to supply chaindisruptions stemming from man-made or natural di-sasters. (Consider the many presentations on the topicat the Annual meeting of POMS in May 2006.) Forsome influential work in this area, we refer our read-ers to Tomlin (2006) and references therein. More at-tention has been paid to this topic following experi-ences with hurricanes Katrina and Rita, which hit thegulf coast of the U.S. in 2005. There is also a growingeffort to cover topics involving closed loop supplychains (see Flapper et al. 2005 for a recent review).These supply chains typically involve previously soldor produced items which are “looped” back to a man-ufacturer and used, either partially or completely, toproduce new units or models. This work is also relatedto works in the SCM area which focus on environmen-tal or “green” issues (Rao and Holt 2005). Upcomingissues of POM will direct specific focus to each of thesetopics.

In the limited space which we have remaining, wewould like to call attention to a number of areasthrough a broader style of argumentation. While the

breadth of SCM research in POM and similar journalsis quite impressive and far beyond what we havediscussed here, we would like to bring attention todimensions along which we perceive an opportunityfor significant improvement in the near and interme-diate term. First and foremost, judging from the pleth-ora of anecdotes about poor service, large inventorylevels, and friction between suppliers and manufac-turers, it seems quite safe to say that SCM researchersstill need to do a better job of packaging the insightsderived from SCM research for SCM professionals.This issue itself is multi-dimensional. First, many ofthe complex decisions rules and contract structuresdeveloped in SCM research are simply not compre-hensible to most managers. For example, Narayananand Raman (2005) report that their straw polls ofSenior Executives reveal that almost all of them admitthat they had not even thought that incentive align-ment might be a problem in their supply chains. Ap-parently, researchers have developed a large literatureon coordinating contracts and other mechanismswhile many practitioners are yet to even recognize theproblem that these contracts seek to address. Nara-yanan and Raman (2005) also point out that supplychain incentives are often mis-aligned due to issuesnot easily addressed in contracts, such as hidden ef-forts which effect supply chain performance, and hid-den information which can make contract formulationor enforcement impossible. Unfortunately, even whencontract formation is possible it is not clear that re-searchers have conveyed the significance of contractterms in clear ways that managers understand. We canargue that this is natural in the sense that this researchstream is still relatively young and that it takes timefor these ideas to filter to the scattered army of SCMprofessionals. But the glaring inefficiencies which re-main in fields such as Health Care, TransportationLogistics, and Disaster Relief demonstrate the greatand immediate need for SCM researchers to find bet-ter and faster ways to help translate research findingsinto firm behavior. Applied projects that address im-plementation difficulties of new supply chain initia-tives, and offer creative and well tested ways to over-come them, might deserve publication when welldocumented and clearly written to demonstrate bothunderstanding of state-of-the art theory and the spe-cifics of their application context.

While the stream of empirical work focused on SCMhas grown steadily over the past decade, it still re-mains insufficiently developed. For example, Changand Grimm (2006) argue that many of the 40 strategypapers which focus on SCM issues and use secondarydata are reporting results from the analysis of datagathered 10 or more years ago. Given the relativelyrecent emergence of SCM as a discipline and the dy-namic nature of the field, findings based on such data


must be interpreted with great caution. Again, this isnot a criticism of the researchers involved. Rather, it isa recognition that vast opportunities exist in this areabecause the efforts to gather relevant data which re-flects the application of SC improvement efforts arestill quite young. Some influential empirical researchin the SCM field has recently appeared in Hendricksand Singhal (2005ab) connecting supply chain execu-tion and firm operating and financial performance.Empirical testing of the “bullwhip effect” is currentlyunder study by Cachon, Randall, and Schmidt (2006).Pioneering work on explaining the wide variations ininventory turnover performance, and the link betweenoperational and financial performance in retail opera-tions has recently appeared in Gaur, Fisher, and Ra-man (2005).

Much research in the academic literature focusesupon single agent problems even though the nature ofSCM almost always involves multiple parties. Suchfocus is often necessary to formulate tractable prob-lems but it may raise concerns about credibility in theminds of some managers who instinctively argue thatthere is no such thing as a single-agent SCM problem.Chang and Grimm (2006) reports that there are still arelatively small number of empirical studies usingdyadic or triadic methodologies. SCM scholars haveurged researchers to broaden the context of SCM re-search from a focal firm to the dyads or the network offirms. However, most strategy research with an inter-organizational context has used data only from a sin-gle firm perspective. Of course, on the modeling sidethere has been an explosion of work addressing thecompetitive interactions at all levels of the supplychain (suppliers, retailers, etc.). We will avoid repli-cating what can be found in the superb survey paperby Cachon and Netessine (2003).

Dischinger et al. (2006) argues that the ideal SCMprofessional must have skills and capabilities in atleast 5 major areas including: (1) functional skills inareas such as procurement, demand/supply planning,manufacturing, global logistics, and customer fulfill-ment; (2) technical skills, particularly in informationtechnology selection, implementation, and applicationas it relates to SCM issues; (3) leadership skills includ-ing communication, negotiation, problem solving,team leadership and project management; (4) globalmanagement experience including work outside thehome country, and (5) high levels of experience andcredibility because virtually all SCM initiatives in-volve multiple parties and often parties that are notintimately familiar with the qualifications and historyof the manager leading the efforts. On one hand, itseems safe to say that we have made great strides inproviding tools, insights, and materials which help toconvey technical knowledge. On the other hand, italso seems that for these professionals, research find-

ings that downplay the significance of inter-firm rela-tionships and management across functional, national,and corporate boundaries miss the truly “hard part”of SCM.

Arguing that managers are increasingly facing anew supply chain competitive landscape, Christopher(1998) stated that successful SCM must model fu-ture—not past—relationships. With this in mind, itseems that managers have a distinct need for researchthat gives them clues about where SCM is going,rather then lessons about fine-tuning an existing sys-tem. Managers want to know how the economic, tech-nological, social, and political forces will serve toshape business and SC practices in the future. Man-agers also seem to be searching for insights which tellthem what types of SC related capabilities can beleveraged for a sustainable advantage, rather thenpassed along to customers through lower margins.SCM researchers may need to collaborate with re-searchers in the fields of economics, and strategy toexplore the relationships between SCM practices andscale economies. For example, as the approaches toSCM become more sophisticated do they become abarrier to entry, and which practices protect the firmfrom competition even if they add inefficiencies to thechain. In a related vein, it is important to consider howgovernment regulation or market interactions can leadto the adoption of supply chain behavior that is in thepublic good, even if it does not benefit any particularfirm.

Additionally, there is a clear need to package andinterpret lessons from the SCM research to benefitcustomers and providers in service industries, includ-ing health care, and education. It appears that oppor-tunities to apply lessons from SCM research abound.For example the linkages between the huge marketingforces and product delivery efforts in the Pharmaceu-tical industry involve billions of dollars and relatedexpenditures which seem to grow faster then inflation.Recent years have seen a sharp growth in efforts tomodel and consider disruptions in supply chains forphysical goods, but this problem may be much moreacute in service chains because of the inability to storethe product at any level of the chain.

Several researchers have argued that practitionerswill benefit from more research at the interface ofother traditional OM topics. We may be missing glar-ing research opportunities which are right under ournoses if we are not willing to view “old” topics in“new” ways related to supply chain dynamics. Forexample, Robinson and Malhotra (2005) finds thateven though the philosophies of quality managementand SCM have been researched extensively in theliterature, few studies examine these agendas jointly,and that research about quality management in thesupply chain is highly disjointed and lacks treatment


as a significant dimension of SCM. They offer a defi-nition of a rarely used term “Supply Chain QualityManagement” as “the formal coordination and inte-gration of business processes involving all partnerorganizations in the supply channel to measure ana-lyze, and continuously improve products, services,and processes in order to create value and achievesatisfaction of intermediate and final customers.” Theyconclude that the establishment of programs and tac-tics to manage and monitor supply chain quality is afundamental step towards maximizing the competi-tiveness and market leadership of supply chains, yetthe extant literature does not provide a comprehensiveexamination of this topic. This work goes on to pointout that managers focused on quality need to moveaway from a product focus to a process focus, and inparticular, need help in understanding how to im-prove processes that cross firm boundaries to improvethe performance of the supply chain. They also notethat few firms even have measurements and standardswhich relate to quality from a chain-wide perspective.Basic questions remain unanswered in this area. Forexample, what performance measures can a supplychain utilize to monitor supply chain processes andtheir alignment with customer desires?

Finally, a recent stream of literature examines thebenefits of RFID technologies in managing supplychains (and it will be featured in a soon to be pub-lished Special Issue in POM, which will include: Heese2007, Ngai et al. 2007, Delen et al. 2007, Barratt andChoi 2007, Karaer and Lee 2007, Amini et al. 2007, andWhitaker et al. 2007). Some research focuses on theapplication of RFID to retail environments (Gaukler etal. 2004; Karaer and Lee 2007; Lee and Ozer 2005),while other research focuses on the application ofRFID to logistics processes (Gaukler et al. 2005) and tomanufacturing and assembly operations (Gaukler etal. 2005). Some recent research work in Kouvelis andLi (2006) is pioneering in its study of the value of RFIDtechnologies in obtaining more accurate lead-time in-formation and then using it for smart response strat-egies in uncertain supply systems. While most of thecurrent research on RFID implications for supplychain management practices has focused on the use ofupdated demand and inventory information, it is im-portant that research also focuses on understandinghow updated lead-time information can be exploitedvia dynamic response strategies in an effort to bettermatch supply and demand. We believe that a greaterfocus on lead-time information will have a profoundimpact on the global supply chain literature and willlead to a better understanding of the sources andvalue of new information and tracking technologieswhich allow the dynamic and accurate updating oflead-time information.

ReferencesAgrawal, N., S. A. Smith, A. A. Tsay. 2002. Multi-vendor sourcing in

a retail supply chain. Production and Operations Management11(2) 157–182.

Agrawal, N., S. Nahmias. 1997. Rationalization of the supplier basein the presence of yield uncertainty. Production and OperationsManagement 6(3) 291–308.

Alderson, W. Marketing efficiency and the principle of postpone-ment. Cost and Profit Outlook, September 1950.

Amini, M., R. F. Otondo, B. D. Janz, M. G. Pitts. 2007. Simulationmodeling and analysis: A collateral application and expositionof RFID technology. Production and Operations Management(Forthcoming).

Anderson, E. G., C. H. Fine, G. G. Parker. 2000. Upstream volatilityin the supply chain: The machine tool industry as a case study.Production and Operations Management 9(3) 239–261.

Anderson, E. G., D. J. Morrice. 2000. A simulation game for teachingservice-oriented supply chain management: Does informationsharing help managers with service capacity decisions? Produc-tion and Operations Management 9(1) 40–55.

Andersson, J., S. Axsater, J. Marklund. 1998. Decentralized multi-echelon inventory control. Production and Operations Manage-ment 7(4) 370–386.

Barratt, M., T. Choi. 2007. Mandated RFID and Institutional Re-sponses: Cases of decentralized business units. Production andOperations Management (Forthcoming).

Bouakiz, M., M. J. Sobel. 1992. Inventory control with an exponentialutility criterion. Operations Research 40(3) 603–608.

Boyabatli, O., B. Toktay. 2004. Operational hedging: A review withdiscussion. Working Paper, INSEAD.

Boyaci, T., G. Gallego. 2004. Supply chain coordination in a marketwith customer service competition. Production and OperationsManagement 13(1) 3–22.

Brynjolfsson, E., M. D. Smith. 2000. Frictionless commerce? A com-parison of internet and conventional retailers. Management Sci-ence 46(4) 563–585.

Cachon, G. P. 1998. Competitive supply chain inventory manage-ment in Quantitative Models for Supply Chain Management, S.Tayur, R. Ganeshan, M. Magazine (eds.), Kluwer AcademicPublishers, Boston, Massachusetts, pp. 111–146.

Cachon, G. P. 2003. Supply chain coordination with contracts inSupply Chain Management: Design, Coordination and Operation,A. G. Kok, S. C. Graves (eds.), Elsevier, Amsterdam, pp. 229–340.

Cachon, G. P., M. Fisher. 1997. Campbell Soup’s continuous replen-ishment program: Evaluation and enhanced inventory decisionrules. Production and Operations Management 6(3) 266–276.

Cachon, G. P., M. Lariviere. 2005. Supply chain coordination withrevenue sharing: strength and limitations. Management Science51(1) 30–44.

Cachon, G. P., S. Netessine. 2006. Game theory in supply chainanalysis in Handbook of Quantitative Supply Chain Analysis: Mod-eling in the E-Business Era, D. Simchi-Levi, S. D. Wu, Z. M. Shen(eds.), Kluwer Academic Publishers, Boston, Massachusetts,pp. 13–66.

Cachon, G., T. Randall, G. Schmidt. 2006. In search of the bullwhipeffect. Working paper, Wharton School, University of Pennsyl-vania, Pennsylvania.

Campbell, A., J. Goentzel, M. Savelsbergh. 2000. Experiences withthe use of supply chain management software in education.Production and Operations Management 9(1) 66–80.

Cattani, K. D., W. G. Gilland, J. M. Swaminathan. 2003. Coordinat-ing traditional and internet supply chains in Handbook of Quan-titative Supply Chain Analysis, D. Simchi-Levi, S. D. Wu, Zou-JunShen (eds.), Springer, Boston, Massachusetts, pp. 643–680.


Cattani, K. D., G. C. Souza. 2002. Inventory rationing and shipmentflexibility alternatives for direct market firms. Production andOperations Management 11(4) 441–457.

Chang, L., C. M. Grimm. 2006. The application of empirical strategicmanagement research to supply chain management. Journal ofBusiness Logistics 27(1) 1–56.

Chatfield, D. C., J. G. Kim, T. P. Harrison, J. C. Hayya. 2004. Thebullwhip effect—impact of stochastic lead time, informationquality, and information sharing: A simulation study. Produc-tion and Operations Management 13(4) 340–353.

Chen, C.-Y., Z. Zhao, M. O. Ball. 2002. A model for batch advancedavailable-to-promise. Production and Operations Management11(4) 424–440.

Chen, F., A. Federgruen. 2000. Mean-Variance Analysis of BasicInventory Models. Working Paper, Columbia University.

Chen, F., R. Samroengraja. 2000. The stationary beer game. Produc-tion and Operations Management 9(1) 19–30.

Chiang, W. K., D. Chhajed, J. D. Hess. 2003. Direct marketing,indirect profits: A strategic analysis of dual-channel supplychain design. Management Science 49(1) 1–20.

Christopher, M. 1998. Logistics and Supply Chain Management, 2ndedition. Financial Times, London.

Clark, T. H., J. H. Hammond. 1997. Reengineering channel reorder-ing processes to improve total supply-chain performance. Pro-duction and Operations Management 6(3) 248–265.

Cohen, M. A., S. Mallik. 1997. Global supply chains: Research andapplications. Production and Operations Management 6(3) 193–210.

Dapiran, P. 1992. Benetton—Global logistics in action. InternationalJournal of Physical Distribution and Logistics Management 22(6)7–11.

De Kok, S.C. Graves. 2003. Supply chain management: Design, coordi-nation, and operation. Elsevier, The Netherlands.

Delen, D., B. C. Hardgrave, R. Sharda. 2007. RFID for better supply-chain management through enhanced information visibility.Production and Operations Management (Forthcoming).

Ding, Q., L. Dong, P. Kouvelis. 2006. On the integration of produc-tion and financial hedging decisions in global markets. Opera-tions Research (Forthcoming).

Ding, Q., P. Kouvelis, J. M. Milner. 2006. Dynamic pricing throughdiscounts for optimizing multi-class demand fulfillment. Oper-ations Research 54(1) 169–183.

Dischinger, J., D. J. Closs, E. McCulloch, C. Speier, W. Grenoble, D.Marshall. 2006. The emerging supply chain management pro-fession. Supply Chain Management Review 10(1) 62–78.

Dong, L., P. Kouvelis, P. Su. 2006. Operational hedging strategiesand competitive exposure to exchange rates. Working Paper,Olin School of Business, Washington University in St. Louis,Missouri.

Erkoc, M., S. D. Wu. 2005. Managing high-tech capacity expansionvia reservation contracts. Production and Operations Management14(2) 232–251.

Ernst, R., M. A. Cohen. 1992. Coordination alternatives in a manu-facturer/dealer inventory system under stochastic demand.Production and Operations Management 1(3) 254–268.

Feitzinger, E., H. L. Lee. 1997. Mass customization at Hewlett-Packard: The power of postponement. Harvard Business Review75(1) 113–121.

Ferdows, K. 1997. Made in the world: The global spread of produc-tion. Production and Operations Management 6(2) 102–109.

Fine, C. H. 2000. Clockspeed-based strategies for supply chain de-sign. Production and Operations Management 9(3) 213–221.

Fisher, M. L. 1997. What is the right supply chain for your produc-tion? Harvard Business Review, March–April 1997 105–116.

Fisher, M., J. Hammond, W. Obermeyer, A. Raman. 1997. Config-

uring a supply chain to reduce the cost of demand uncertainty.Production and Operations Management 6(3) 211–225.

Flaherty, T. 1996. Global Operations Management. McGraw Hill, NewYork, New York.

Flapper, S. D. P., J. Van Nunen, L. Van Wassenhove (eds.). 2005.Managing Closed-Loop Supply Chains, Springer Science and Busi-ness Media, New York, New York.

Flood, E. Jr., D. R. Lessard. 1986. On the measurement of operationalexposure to exchange rates: A conceptual approach. FinancialManagement, Spring 25–36

Forrester, J. 1961. Industrial Dynamics. John Wiley and Sons, NewYork, New York.

Froot, K., D. Scharfstein, J. Stein. 1993. Risk management: Coordi-nating corporate investment and financing policies. Journal ofFinance 48(5) 1629–1658.

Gan, X., S. P. Sethi, H. Yan. 2004. Channel coordination of supplychains with risk-averse agents. Production and Operations Man-agement 13(2) 135–149.

Gan, X., S. P. Sethi, H. Yan. 2005. Channel coordination with arisk-neutral supplier and a downside-risk-averse retailer. Pro-duction and Operations Management 14(1) 80–89.

Garg, A., H. L. Lee. 1998. Managing product variety: An operationalperspective in Quantitative Models for Supply Chain Management,S. Tayu, R. Ganeshan, M. Magazine (eds.), Kluwer AcademicPublishers, Boston, Massachusetts, pp. 467–490.

Garg, A., C.S. Tang. 1997. On postponement strategies for productfamilies with multiple points of differentiation. IIE Transactions29 641–650.

Gaukler, G. M., W. H. Hausman. 2005. RFID in assembly operations:Process and quality savings. Working Paper, Stanford Univer-sity, Stanford, California.

Gaukler, G. M., O. Ozer, W. H. Hausman. 2004. RFID and productprogress information: Improved dynamic emergency orderingpolicies. Working Paper, Stanford University, Stanford, Califor-nia.

Gaukler, G. M., R. W. Seifert, W. H. Hausman. 2004. Item-level RFIDin the retail supply chain. Production and Operations Management(Forthcoming).

Gaur, V., M. L. Fisher, A. Raman. 2005. An econometric analysis ofinventory turnover performance in retail services. ManagementScience 51(2) 181–194.

Gerchak, Y., Y. Wang. 2004. Revenue-sharing vs. wholesale pricecontracts in assembly systems with random demand. Productionand Operations Management 13(1) 23–33.

Graman, G. A., M. J. Magazine. 2002. A numerical analysis ofcapacitated postponement. Production and Operations Manage-ment 11(3) 340–357.

Granot, D., Greys Sosic. 2005. Formation of alliances in internet-based supply exchanges, Management Science 51(1) 92–105.

Gutierrez, G. J., P. Kouvelis. 1995. A robustness approach to inter-national sourcing. Annals of Operations Research 59 165–193.

Hazra, J., B. Mahadevan, S. Seshadri. 2004. Capacity allocationamong multiple suppliers in an electronic market. Productionand Operations Management 13(2) 161–170.

Heese, H. S. 2007. Inventory record inaccuracy, double marginal-ization and RFID adoption. Production and Operations Manage-ment (Forthcoming).

Hendricks, K. B., V. R. Singhal. 2005a. An empirical analysis of theeffect of supply chain disruption on long-run stock price per-formance and equity risk of the firm. Production and OperationsManagement 14(1) 35–52.

Hendricks, K. B., V. R. Singhal. 2005b. Association between supplychain glitches and operating performance. Management Science51(5) 695–711.

Hodder, J. E. 1982. Exposure to exchange rate movements. Journal ofInternational Economics 13 375–386.


Hoyt, D., E. Lopez-Tello. 2001. Lucent Technologies—Provisioningand Postponement, Stanford University Case No. GS-02.

Huchzermeier, A., M. A. Cohen. 1996. Valuing operational flexibil-ity under exchange rate risk. Operations Research 44(1) 100–113.

Jacobs, F. R. 2000. Playing the beer game distribution game over theinternet. Production and Operations Management 9(1) 31–39.

Johnson, M. E., S. Whang. 2002. E-business and supply chain man-agement: An overview and framework. Production and Opera-tions Management 11(4) 413–423.

Johnson, M. E., D. F. Pyke. 2000. A framework for teaching supplychain management. Production and Operations Management 9(1)2–18.

Karaer, O., H. Lee. 2007. Managing the reverse channel with RFID-enabled negative demand information. Production and Opera-tions Management (Forthcoming).

Klassen, R. D., S. Vachon. 2003. Collaboration and evaluation in thesupply chain: The impact on plant-level environmental invest-ment. Production and Operations Management 12(3) 336–352.

Kleindorfer, P. R., G. H. Saad. 2005. Managing disruption risks insupply chains. Production and Operations Management 14(1) 53–68.

Kleindorfer, P. R., D. J. Wu. 2003. Integrating long-and short-termcontracting via business-to-business exchanges for capital in-tensive industries. Management Science 49(1) 1597–1615.

Kogut, B., N. Kulatilaka. 1994. Operating flexibility, global manu-facturing, and the option value of a multinational network.Management Science 40(1) 123–139.

Kopczak, L. R. 1997. Logistics partnerships and supply chain re-structuring: Survey results from the U.S. computer industry.Production and Operations Management 6(3) 226–247.

Kopczak, L. R., J. C. Fransoo. 2000. Teaching supply chain manage-ment through global projects with global project teams. Produc-tion and Operations Management 9(1) 91–104.

Kouvelis, P., M. A. Lariviere. 2000. Decentralizing cross-functionaldecisions: Coordination through internal markets. ManagementScience 46(8) 1049–1058.

Kouvelis, P., J. Li. 2006. Flexible backup supply and the manage-ment of lead-time uncertainty. Working Paper, Olin School ofBusiness, Washington University in St. Louis, Missouri.

Kouvelis, P., C. L. Munson. 2004. Using a structural equationsmodeling approach to design and monitor strategic interna-tional facility networks in Handbook of Quantitative Supply ChainAnalysis: Modeling in the E-Business Era, D. Simchi-Levi, S. D.Wu, Z. M. Shen (eds.), Kluwer Academic Publishers, Boston,Massachusetts, pp. 681–769.

Kraiselburd, S., V. G. Narayanan, A. Raman. 2004. Contracting in asupply chain with stochastic demand and substitute products.Production and Operations Management 13(1) 46–62.

Kreipl, S., M. Pinedo. 2004. Planning and scheduling in supplychains: An overview of issues in practice. Production and Oper-ations Management 13(1) 77–92.

Lal, R. and M. Savary. 1999. When and how is the internet likely todecrease price competition. Marketing Science 18(4) 485–503.

Lariviere, M. 1998. Supply chain contracting and co-ordination withstochastic demand in Quantitative Models for Supply Chain Man-agement, S. Tayur, R. Ganeshan, M. Magazine (eds.), KluwerAcademic Publishers, Boston, Massachusetts, pp. 233–268.

Lee, H. L. 1993. Design for supply chain management: Concepts andexamples in Perspectives in Operations Management, R. Sarin(ed.), Kluwer Academic Publishers, Boston, Massachusetts.

Lee, H. L. 1996. Effective management of inventory and servicethrough production and process design. Operations Research44(1) 151–159.

Lee, H. L., C. A. Billington, B. Carter. 1993. Hewlett Packard gainscontrol of inventory and service through design for localiza-tion. Interfaces 23(3) 1–11.

Lee, H. L., E. Feitzinger, C. A. Billington. 1997. Getting ahead ofyour competition through design for mass customization. Tar-get 13(2) 8–17.

Lee, H. L., O. Ozer. 2005. Unlocking the value of RFID. Productionand Operations Management (Forthcoming).

Lee, H. L., P. Padmanabhan, S. Whang. 1997a. The bullwhip effect insupply chains. Sloan Management Review 38 93–102.

Lee, H. L., P. Padmanabhan, S. Whang. 1997b. Information distor-tion in a supply chain: The bullwhip effect. Management Science43(4) 546–558.

Lee, H. L., M. Sasser. 1995. Product universality and design forsupply chain management. Production Planning and Control:Special Issue on Supply Chain Management 6(3) 270–277.

Lee, H. L., C. S. Tang. 1998. Variability reduction through operationsreversal. Management Science 44(2) 162–172.

Li, C. L., P. Kouvelis. 1999. Flexible and risk sharing supply con-tracts under price uncertainty. Management Science 45(10) 1378–1398.

Lowe, T. J., R. E. Wendell, G. Hu. 2002. Screening location strategiesto reduce exchange rate risk. European Journal of OperationalResearch 136(3) 573–590.

Mas-Colell, A., M. D. Whinston, J. R. Green. 1995. MicroeconomicTheory. Oxford University Press, Inc.

Markowitz, H. M. 1952. Portfolio selection. Journal of Finance 7(1)77–91.

Markowitz, H. M. 1987. Mean-Variance Analysis in Portfolio Choiceand Capital Markets, Basil Blackwell, Oxford, UK.

Martinez-de-Albeniz, V., D. Simchi-Levi. 2005. A portfolio approachto procurement contracts. Production and Operations Manage-ment 14(1) 90–114.

Mehring, J. S. 2000. A practical setting for experiential learningabout supply chains: Siemens Brief Case supply chain simula-tor. Production and Operations Management 9(1) 56–65.

Milner, J. M., P. Kouvelis. 2006. Inventory, speculation and sourcingstrategies in the presence of online exchanges. Manufacturing &Service Operations Management. Forthcoming.

Modigliani, F., M. Miller. 1958. The cost of capital, corporate finance,and the theory of investment. American Economic Review 30(3)261–297.

Munson, C. L., M. J. Rosenblatt. 1997. The impact of local contentrules on global sourcing decisions. Production and OperationsManagement 6(3) 277–290.

Narayanan, V. G., A. Raman. 2005. Aligning incentives in supplychains. Harvard Business Review 82(11) 94–102.

Ngai, E. W. T., T. C. E. Cheng, K. Lai, P. Y. F. Chai, Y. S. Choi, R. K.Sin. 2007. Development of an RFID-based traceability system:Experiences and lessons learned from an aircraft engineeringcompany. Production and Operations Management (Forthcoming).

Paternack, B. (1985). Optimal pricing and returns policies for per-ishable commodities. Marketing Science 4(2) 166–176.

Peleg, B., H. L. Lee, W. H. Hausman. 2002. Short-term e-procure-ment strategies versus long-term contracts. Production and Op-erations Management 11(4) 458–479.

Porteus, E., S. Whang. 1991. On manufacturing/marketing incen-tives. Management Science 37(9) 1166–1181

Poundarikapuram, S., D. Veeramani. 2004. Distributed decisionmaking in supply chains and e-Marketplaces. Production andOperations Management 13(1) 77–92.

Rao, P., D. Holt. 2005. Do green supply chains lead to competitive-ness and economic performances? International Journal of Oper-ations and Production Management 25(9) 898–916.

Robinson, C. J., M. K. Malhotra. 2005. Defining the concept ofsupply chain quality management and its relevance to aca-demic and industrial practice. International Journal of ProductionEconomics 16(3) 315–916.


Shin, H., W. C. Benton. 2004. Quantity discount-based inventorycoordination: Effectiveness and critical environmental factors.Production and Operations Management 13(1) 63–76.

Simchi-Levi, D., S. D. Wu, Z. M. Shen. 2003. Handbook of QuantitativeSupply Chain Analysis. Springer, Boston, Massachusetts.

Smith, M. D., E. Brynjolfsson. 2001, Consumer decision-making atan internet shopbot: Brand still matters. The Journal of IndustrialEconomics 49(4) 541–558.

Smith, C., R. Stulz. 1985. The determinants of firms’ hedging poli-cies. Journal of Financial and Quantitative Analysis 20(4) 391–402.

Sodhi, M. S. 2005. Managing demand risk in tactical supply chainplanning for a global consumer electronics company. Productionand Operations Management 14(1) 69–79.

Souza, G. C., Z. Zhao, M. Chen, M. O. Ball. 2004. Coordinating salesand raw material discounts in a global supply chain. Productionand Operations Management 13(1) 34–45.

Swaminathan, J. M., H. L. Lee. 2003. Design for postponement inSupply Chain Management: Design, Coordination and Operation,A. G. de Kok, S. C. Graves (eds.), Elsevier, Amsterdam.

Swaminathan, J. M., S. R. Tayur. 1998. Managing broader productlines through delayed differentiation using vanilla boxes. Man-agement Science 44(12.2) S161–S172.

Tang, C. S. 2006. Perspectives in supply chain risk management.International Journal of Production Economics 103 451–488.

Tayur, R., R. Ganeshan, M. Magazine. 1998. Quantitative Models forSupply Chain Management. Kluwer, Boston, Massachusetts.

Tomlin, B. 2006. On the value of mitigation and contingency strat-egies for managing supply chain disruption risks. ManagementScience 52(5) 639–657.

Tsay, A. 1999. Quantity-flexibility contract and supplier-customerincentives. Management Science 45(10) 1339–1358.

Tsay, A. A., N. Agrawal. 2004. Channel conflict and coordination inthe e-commerce age. Production and Operations Management13(1) 93–110.

Tsay, A., S. Nahmias, N. Agrawal. 1998. Modeling supply chaincontracts: a review in Quantitative Models for Supply Chain Man-agement, S. Tayur, R. Ganeshan, M.Magazine (eds.), Kluwer,Boston, Massachusetts.

Urban, T. L. 2000. Supply contracts with periodic, stationery com-mitment. Production and Operations Management 9(4) 400–413.

Van Mieghem, J. 2003. Capacity management, investment, andhedging: Review and recent developments. Manufacturing &Service Operations Management 5(4) 269–302.

Van Mieghem, J. 2004. Component commonality strategies: Valuedrivers and equivalence with flexible capacity strategies. Man-agement Science 50(3) 419–424.

Vollmann, T. E., C. Cordon, J. Heikkila. 2000. Teaching supply chainmanagement to business executives. Production and OperationsManagement 9(1) 81–90.

Wagner, H. M. 1995. Global sensitivity analysis. Operations Research43(6) 948–969.

Warburton, R. D. H. 2004. An analytical investigation of the bull-whip effect. Production and Operations Management 13(2) 150–160.

Whang, S. 1995. Coordination in operations: A taxonomy. Journal ofOperations Management 12 413–422.

Whitaker, J., S. Mithas, M. S. Krishnan. 2007. A field study of RFIDdevelopment and return expectations. Production and OperationsManagement (Forthcoming).

Yan, H., K. Liu, A. Hsu. 2003. Optimal ordering in a dual-suppliersystem with demand forecast updates. Production and Opera-tions Management 12(1) 30–45.

Zhang, A. X. 1997. Demand fulfillment rates in an assemble-to-ordersystem with multiple products and dependent demands. Pro-duction and Operations Management 6(3) 309–324.

Zhang, H. 2002. Vertical information exchange in a supply chainwith duopoly retailers. Production and Operations Management11(4) 531–546.


supply chain management research and production and operations management: review, trends, and...

Documents