qThis research was made possible by a grant from Hewlett-Packardand Arizona State University, College of Business, Council of 100Summer Research Grant.

*Corresponding author. Tel.: #1-480-965-2998; fax: #1-480-965-8629.

E-mail address: lisa.ellram@asu.edu (L. Ellram).

European Journal of Purchasing & Supply Management 7 (2001) 15}27

Purchasing leverage considerations in the outsourcing decisionq

Lisa Ellram!,*, Corey Billington"

!Department of Business Administration, College of Business, Arizona State University, P.O. Box 874706, Tempe, AZ 85287-4706, USA"Strategic Planning and Modeling, Hewlett-Packard, Palo, Alto, CA, USA

Received 2 February 1999; received in revised form 20 January 2000; accepted 24 January 2000

Abstract

As outsourcing has increased, an unanticipated problem has come to light for many manufacturers: as they reduce their purchasevolume with their materials and parts suppliers for inputs to items that are now outsourced, these manufacturers lose their volumeand price leverage for the inputs. As a result, they may have to pay more for parts and materials that they continue to buy fromsuppliers. In addition, the contract manufactures to whom they outsource may also be paying a higher material/part price or may geta lower price by consolidating volume from a number of customers, but may retain the volume discount as part of their pro"t. Thispaper uses the transaction cost literature to explore how purchasing can best manage the potential loss of volume/purchase leveragefor the inputs to items that are outsourced. In exploring this issue, case studies are used to demonstrate both unsuccessful andsuccessful attempts at managing this loss of purchasing leverage. A prescriptive framework is developed that suggests how to bestmanage the risk of losing purchasing leverage when outsourcing. This framework considers the relative volume of the buyer and thecontractor, the level of trust, and the concerns of the supplier regarding its price visibility. ( 2000 Elsevier Science Ltd. All rightsreserved.

Keywords: Outsourcing; Purchasing leverage; Transaction cost analysis

1. Introduction

The issue to be examined in this study is:

Is there a way that purchasing can retain its buyingleverage for intermediate materials and componentswhen it outsources the manufacture of the item thatconsumes those inputs?

The question will be addressed primarily from an eco-nomic perspective, considering the amount of potentialsavings in purchase price that the organization mightgive up by outsourcing its purchasing of material/compo-nents used by subcontractors for outsourced items. Thisquestion was precipitated by Hewlett-Packard (H-P),a major manufacturer of personal computers. H-P wasoutsourcing its subassemblies/boards to an increasingdegree. Some members of the supply function became

concerned that H-P might lose its purchasing leverage onsome of the parts and components that made up theboard if it allowed the assemblers to purchase all thecomponents. This issue took on supply chain designissues, broader than price leverage. It also concerns as-surance of supply as key components were often onallocation, and inventory management, as prices are of-ten volatile, and H-P does not want to pay for partdevaluation due to excess inventory. Yet H-P did notwant to take possession of the parts/components andreship them to subassemblers, as that was very ine$cient,increasing inventory levels, costs and cycle times. Alter-natives were needed.

The purpose of this study is to examine the questionpresented above. It will do so by "rst brie#y exploring therelated literature on outsourcing, and examining whereoutsourcing seems to apply from a theoretical stand-point. Second, based on interviews with organizations,the paper will present an examination of the practices ofcompanies in three categories: those who have madeunsatisfactory attempts to leverage purchased inputs usedin outsourcing, those that believe they have successfullyleveraged purchased inputs used in outsourced items,and those who have had mixed results in either leveraging

0969-7012/00/$ - see front matter ( 2000 Elsevier Science Ltd. All rights reserved.PII: S 0 9 6 9 - 7 0 1 2 ( 0 0 ) 0 0 0 0 4 - 6

or not leveraging the price of purchased inputs used inthe manufacture of outsourced items. Based on an integ-ration of theory and the interviews with organizations,recommendations are made as to the applicability ofleveraging purchased input used in outsourced items invarious situations.

2. Relation of outsourcing to core competency

For purposes of this paper, outsourcing is de"ned asthe transfer of the production of goods or services thathad been performed internally to an external party. Thisis consistent with the viewpoint of H-P in pursuing out-sourcing implications on purchasing leverage. It seems tobe well understood and agreed upon that organizationsshould build on their core competencies, and leveragethese competencies. Core competencies should never beoutsourced. It is those activities where the "rm has nounique capabilities that should be considered as candi-dates for outsourcing.

Core competencies are (Quinn and Hilmer, 1994):

1. Skills or knowledge sets, not products or functions.2. Flexible, long-term platforms that are capable of

adaption or evolution.3. Limited in number: generally two to three.4. Unique source of leverage in the value chain.5. Areas where the company can dominate/perform ac-

tivities important to the customers better than others.6. Elements important to customers in the long run.7. Embedded in the organization's systems.

According to Quinn and Hilmer, core competenciesare the activities that provide long-term competitive ad-vantage. These must be closely protected. All other activ-ities are candidates for outsourcing. Thus, if suppliermarkets were e$cient, companies would outsource allactivities except core competencies. However, marketshave many ine$ciencies, transaction costs, and otheruncertainties. Thus, the market is not always the answer.

Clearly, suppliers are in the market to make money. Themanner in which they make money while providing thecustomer with lower costs than internal production are:

1. Specialized knowledge that creates competitive ad-vantage.

2. Lower cost of operations than its customers (wages,overhead, e$ciency).

3. Leverage/better raw material prices with supplier thanits customers (due to volume consolidation, owner-ship, or a host of other reasons).

4. Economies of scale and/or scope.

Outsourcing allows an organization to take advantageof the strengths within the supply market. The problem isthat US companies generally outsource to save on short-term costs such as overhead. This creates ine$ciencies,

expensive management issues and huge numbers of sub-contractors (Purchasing, 1998a,b). The creation of theseissues signi"cantly undermines the purpose of outsourc-ing in these organizations.

3. Transaction cost analysis

Transaction cost analysis (TCA, also referred to astransaction cost economics) is an approach that explicitlyconsiders the implications of an organization's choice toperform a transaction or activity internally (verticallyintegrate), or in the market (horizontally integrate oroutsource). According to Williamson the decision ofwhether or not to outsource, and the extent of outsourc-ing, depends on the transaction costs associated withoutsourcing versus internalization (Williamson, 1981).Vertical integration represents the failure of the free mar-ket to handle exchange relationships e$ciently. Thus, thetransaction costs that result in the market are essentiallythe `costs of running the systema. The level of transactioncosts depends upon a number of factors:

f the frequency of the transaction;f the level of transaction speci"c investments;f the external and internal uncertainty.

Williamson describes uncertainty as the inability topredict contingencies that may occur (Williamson, 1979).This creates a problem in developing contractual rela-tionships because contracts are somehow `incompletea.Such contingencies may create `opportunisma, the abil-ity to take advantage of the situation in order to fa-vorably interpret the contractual terms. There are twotypes of uncertainty which may exist: external/environ-mental and internal/behavioral.

The "rst type of uncertainty, external/environmentaluncertainty, deals with the level of uncertainty in themarket in which the organizations operate. For example,external uncertainty is high where technology is rapidlychanging, or where demand #uctuates signi"cantly andunpredictably. The premise of Williamson's work is thatas environmental uncertainty increases, so will the likeli-hood of vertical integration (Williamson, 1979).

The second issue, internal/behavioral uncertainty, re-#ects the idea that the organization really does not knowwhat it wants, or the situation involving the transaction issuch that the contracting parties have no assurance thatthe other party has actually ful"lled its obligation/per-formed. An example of this would be when a supplier saysthat the item it provides is `on allocationa, and it cannotship the full order. How can this truly be veri"ed? Ora buyer may request a price decrease/improved servicebased on an o!er it received from a competitor of thecurrent supplier; again very di$cult to verify. As a result,the organization(s) involved have di$culty evaluatingwhether the relationship is meeting expectations.

16 L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27

Table 1Potential advantages of obligational contracts

Cost reductionReduced production costsReduced transactions costs

Better information/reduced uncertaintyRoutinize transactions

Risk reductionSharing of assets

Lower break-even pointSharing of information both formally and informallyVolume commitmentsIncreased #exibility versus vertical integrationFuture orientation with joint planningTrustInterdependenceSharing of risks and rewards of relationship

Table 2Potential disadvantages of obligational contracts

Control/DependenceLoss of incentive to be competitive due to dependenceTrust may prolong relationship of unequal bene"tsIncreased risk if opportunistic due to dependenceDi$cult to manage multiple relationships/complex governanceDi$cult to build all contingencies into relationshipDi$cult to evaluate due to multi-dimensionalityReal or perceived loss of control over operationsInformal relationships may be di$cult to controlHeavily dependent on individuals involved; limited transferability

CostOutside "rms receive pro"t on value-added that "rm would retain ifactivities are performed internallyDuplication of e!orts may increase total cost

Transaction cost theory assumes that each of theabove factors creates potential costs. If the costs are veryhigh, complete vertical integration may be the answer.However, there is usually a contractual solution some-where between vertical integration and complete freemarket forces that will better satisfy the organization'sneeds (John and Weitz, 1988; Walker and Weber, 1984;Heide, 1994; Venkatraman, 1997)

Power is also an issue in determining organizationalboundaries. However, Williamson points out that integ-ration decisions based on power rather than e$ciencywill eventually shift based on e$ciency.

3.1. Recommendations for governance structure-based onTCA

Obligational contracting represents an ongoing, co-operative relationship with another channel memberwhich is bound by a legal agreement. This could be anongoing relationship with a supplier or a subcontractor.This section "rst presents the advantages of obligationalcontracting, then the disadvantages, followed by a reviewof situations where obligational contracts have the bestcompetitive "t.

3.2. Advantages of obligational contracts

The advantage of contractual relationships may beclassi"ed as cost reduction or risk reduction. Each isde"ned below, and summarized in Table 1.

Cost reduction refers to the ability of the "rm to lowerthe overall cost of conducting business. This could occurif the contracted supplier has more leverage in the supplybase than the buying organization. Further, risks can bereduced through commonality of both information andassets. Lack of exclusive asset ownership increases the"rm's #exibility. Information sharing can reduce costsand uncertainty, by allowing partners to anticipate/planfor the actions of other partners.

3.3. Disadvantages of obligational contracts

Contractual relationships are not without disadvan-tages. The disadvantages, as discussed below and shownin Table 2, may be classi"ed as associated with eithercontrol/dependence or cost.

Clark notes that contractual relationships may createdependence without the control to balance that depend-ence; there is a dependence risk in terms of relative power(Clark, 1961). In comparing contractual relationships tovertical integration, a disadvantage of contractual rela-tionships is a real or perceived loss of control over theactivity which is performed externally. The supply chainmay be more di$cult to manage and control with addi-tional parties involved in performing functions. Further,costs such as the price of purchased inputs, may rise whenactivities are dispersed among a number of parties, ratherthan performed internally. The outside "rms must receivepro"t on the value they add.

3.4. When to use obligational contracts

Williamson discusses the "t of obligational contractsin terms of his three critical dimensions of transactions,as discussed earlier (Williamson, 1981). He postulatesthat obligational contracts "t best when transactions arerecurrent, and assets needed for production are highlyspecialized, yet not limited to one use. However, "rmsmust recognize the hazards of opportunism. When build-ing #exibility into contracts, "rms should not allow #ex-ibility in areas where the potential for opportunism ishigh. This is congruent to Quinn and Hilmer's assess-ment that "rms must reduce their vulnerability, whileallowing contract #exibility (Quinn and Hilmer, 1994).

One factor that contributes to the success of obliga-tional contracting is that the expectation of futurebusiness discourages opportunism. In obligational con-tracting, the pressure to maintain the relationship createsa credible commitment among parties, reducing oppor-tunism. The reference point for interaction is not a single

L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27 17

transaction; it is the entire relationship that has beenestablished. Further, as uncertainty increases, or if thepotential risk is high, a "rm may want its contract toinclude more explicit terms and contingencies. Whilespeci"city has high initial costs, there may be long-termsavings.

Obligational contracts may also be important wherethere is a concentration at one node of the supply chain.Supply chain members may need to co-operate and seekout long-term relationships in order to gain access to thegoods or services in the concentrated channel node. Ob-ligational contracts may also help "rms adapt in verycompetitive, recently deregulated markets. Obligationalcontractual relationships can further allow "rms to re-spond better to shifting customer demand patterns(Ellram, 1991). The next section presents a summary ofcurrent practices and issues of a number of "rms involvedin outsourcing purchasing across a number of industries.This is followed by an assessment and recommendations.

4. Case analysis

The following provides an overview of the approachesthat a number of organizations are taking to address theissue of `the potential loss of purchasing leveragea whenoutsourcing. The data were gathered based on per-sonal/telephone interviews conducted by the researchers.Some organizations were willing to share more thanothers. Results achieved and approaches used weremixed. The industries of the organizations who contrib-uted are revealed; the individual company names are not,for reasons of con"dentiality.

In all of the cases, a decision was made to outsource ornot outsource independent of the leverage issue. Theissue addressed here is thus the governance of theprice/leverage of inputs used in the manufacture of out-sourced items. For purposes of the following presenta-tion, the customer/buying organization refers to thecompany that is outsourcing an item. The supplier refersto the provider of inputs such as raw materials or compo-nents, for an outsourced item. The contractor refers tothe contract manufacturer that is actually manufactur-ing/assembling the outsourced item.

5. Unsatisfactory current/previous attempts

The "rst group, buying organizations with unsatisfac-tory experience in leveraging outsourcing, are classi"edas such for one of two reasons. First, they may have triedto gain leverage and found that they could not. Second,they may have found that they have lost leverage inoutsourcing, and have not yet found a satisfactory way toregain it.

5.1. Motivational training/products company (MTPC)

The "rst organization is a leader in the "eld of personalmotivation, specializing in public and in-company train-ing, and selling accompanying printed material, video-tapes and audiotapes. It began its business bymanufacturing and selling its own audiotapes. It soondiscovered that `manufacturinga was not its core com-petency, and sought to outsource this business. In orderto continue to get good prices from the suppliers of thetapes and tape cases, MTPC purchased all of the partsthat went into production. These parts were drop-ship-ped directly to the contractor from the suppliers. How-ever, the buying organization received and paid theinvoices for all of these items and kept track of all of theinventory on its books.

This created some accounting problems for a numberof reasons. If any of the inputs were unusable or unac-counted for, inventory counts had to be adjusted accord-ingly. There was a greater danger of being out-of-stockon something due to the physical distance and separationof accountability. Unexpected out-of-stock situationsmight require a rush order, premium transportation, andhigher overall supply chain costs.

When renegotiating its labor contract with this con-tractor this year, these problems were discussed exten-sively. Clearly, such problems were raising transactioncosts, creating excess paper work and ine$ciency, andduplicating the e!orts of parties involved in the supplychain. The discussion and negotiation revealed that thecontractor was buying similar parts for other customers.By combining the volumes, the contractor was able tonegotiate LOWER materials costs than MTPC! Thus, allof the excess work and hassles were actually increasingthe buying organization's out-of-pocket materials prices,as well as its transaction costs. The buying organizationnow has the contractor negotiate prices and buy the rawmaterial directly. It pays cost when the "nished goods arereceived, based on actual materials invoices. It also paysa slight mark-up to re#ect the fact that the contractornow carries the materials, and uses its own capital, ratherthan the buying organization's capital. While this scen-ario had a satisfactory resolution, it serves to illustratethat organizations may over-estimate their leverage inthe market place, and in doing so, create unnecessaryadditional out-of-pocket and transaction cost.

5.2. Heavy manufacturing

This heavy manufacturing company (Heavy) isa leader in the manufacture and sales of farm, lawn andearthmoving equipment. As with many similar US-basedmanufacturing operations that are located in the rust-belt and heavily unionized, it has been outsourcing anincreasing number of its components and operations forthe past "ve years or so. This is particularly true in the

18 L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27

Fig. 1. `Collective procurementa in the automotive industry.

areas of machining, tooling, and other metal operationsthat are relatively labor intense.

Most of the shops to which Heavy outsources aresmall in comparison to Heavy. As a result of all of thisoutsourcing, it has seen a phenomenal, albeit gradual,downward shift in the tonnage of steel that it purchasesand uses in its internal operations. As a result, it is payingcomparatively higher prices for steel used in its internaloperations. In addition, because most of its contractorsare relatively small, the contractors are also paying highsteel prices, even higher than this company is paying forits internal uses.

This situation of higher steel prices has snowballedinto an avalanche. Given the hundreds of millions ofdollars of steel used directly and indirectly each year bythis company, it estimates that there are `tens of millionsof dollars on the tablea in terms of potential savings fromvolume leverage. The company is currently undergoinga study/benchmarking with other companies regardinghow to best manage this situation.

6. Successful leverage

Four examples of successful leverage are illustratedbelow. These are de"ned as successful based on the or-ganization's own assessment.

6.1. Automotive industry (Auto)

The automotive industry provides an excellentexample of how to successfully manage and leveragepurchases with second tier suppliers. While only one ofthe major auto manufacturers was interviewed, welearned that two other automakers were using an identi-cal process based on benchmarking the two had donewith the case study company. The two areas that itleverages are both raw materials: steel and plastic resin.

The leverage relationship is managed the same way inboth cases. The contractor ("rst tier supplier) places itsorder for steel or resin electronically via EDI or theInternet with Auto. Auto in turns transmits the orderelectronically to its steel or resin supplier(s). This supplierdrop ships the order to the contractor, who placed theorder with Auto, and then invoices Auto electronically atthe going price. Auto turns around and electronicallyinvoices the contractor at a `maskeda price. This rela-tionship is illustrated in Fig. 1. The `maskeda price isa "xed price that Auto determines for the year, so thatprice stability is provided to the contractor, and Autoabsorbs all of the variances. More importantly, that`maskeda price does not allow the contractor to see whatAuto is really paying for steel. That eliminates any temp-tation of the contractor to `sharea price information withcompetitors of Auto who also use that contractor, or totry to get that same price from the supplier of inputs for

other buys. If the latter were the case, the supplier wouldbe extremely reluctant to provide Auto this favorableleverage.

Auto says that the biggest bene"ts of this approach arethat it has the ability to control the quality of rawmaterials used by its contractors. It also takes on theburden of managing and resolving quality and other rawmaterial problems of these suppliers. Since Auto has thecontract and the relationship with the steel supplier,Auto will deal with the steel supplier directly. Whenasked whether Auto is concerned that its contractorsmay order extra raw materials to get a better price, itresponded that was not an issue. Auto has a "rm sched-ule which it shares with its contractors, so has a verygood feel for what that contractor `shoulda be buying.

6.2. High-tech ozce equipment (IM)

This success story is an organization in the imagingindustry (IM). It learned its lessons about purchasingleverage from the Japanese, when the Japanese enteredthe US market as "erce competitors at signi"cantly lowerprices in the late 1970s/early 1980s. It discovered that oneof its competitors was paying about 6% less for resin,although it was using only one fourth as much! Therewere two reasons behind that. First, IM was using wellover 100 types of resins, while its competitor was usingunder 20. Thus, the competitor gained considerableleverage through standardization. Second, its competitorcontracted directly with the second tier resin suppliers,

L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27 19

while IM did not. After standardizing its resins, con-solidating its supply base and negotiating directly with itsresin suppliers, IM enjoyed an 11% cost savings over thecompetitor, worth about $5 million a year in bottom linesavings.

In administering the resin contract, IM negotiates withthe supplier on behalf of the contract molders. The con-tractors then order directly from the resin suppliers andIM is out of the loop. When asked whether they do notcare that their contractors know their price, the responsewas, why should we? What about the contractor pressur-ing the resin supplier for a lower price? That is betweenthe supplier and the contractor. If the contractor can geta better price and thus be more competitive, IM felt itwas a win}win situation, and helped build a strongersupply base.

IM uses this same approach on a number of standardelectrical components. However, IM does have con-tractors use some components/parts that are proprietaryto IM. These are often manufactured by OEM's, and thetechnology may be patented or proprietary to either IMor the supplier. In those instances, IM actually purchasesall of the needed component parts and has them shippedto IM. These parts are then drop-shipped to the con-tractor as needed. IM pays for the parts and manages allof the inventory. When asked why the high degree ofprotectiveness with these items, it was due to their pro-prietary nature and their relative importance to IM'score competencies. These were items for which IM didnot generally outsource the assembly to a contractorunless it had capacity issues. In these situations, it did notwant the contractor to have access to any price informa-tion or any extra parts.

IM is very satis"ed with the process it uses to managethese outsourced items. It believes that it is leveraging itspurchasing power to the full extent possible, withoutcreating an undue administrative burden, or unnecessarybreaches of core technologies.

6.3. Industrial/consumer chemicals (Chem)

This organization, a major producer of industrial andconsumer chemicals, polymers and "bers (Chem) viewsthe outsourcing decision, including leverage aspects, verystrategically. One area that it has decided to outsource,while maintaining its leverage, is logistics services. Whileit sees these services as providing some competitive ad-vantage, it believes that it retains competitive advantageand control by choosing all of the service providers:transportation companies, warehousers and so on. Thethird party contractor manages these relationships.

Chem is able to retain its leverage by choosing andnegotiating contracts with the suppliers of service. Chemactually pays all the bills, which are matched via itssophisticated EDI system. Thus, it receives two logistic-related charges, one related to the management fee it

pays the contractor who manage its logistics services, theother related to the actual transportation or warehousingservice performed by the service supplier.

Each year, it re-evaluates whether the contractors whoprovide management should start paying the bills andnegotiating the contracts with the suppliers. It continuesto perform these tasks inhouse for two reasons:1. It gets better information on actual movement pat-

terns from its own sophisticated internal informationsystem.

2. It wants to ensure its leverage/not let others know itscontractual prices.

6.4. Electronics manufacturer (Elec)

This organization (Elec) is a leading manufacturer ofhigh-technology components and products. It is a sup-plier to H-P. The philosophy of this organization is `donot out source any purchasing activity beyond MROa.As the VP of Materials and Supplies pointed out, mater-ials are over 50% of this organization's cost of sales. Assuch, materials purchases have the potential to bea source of competitive advantage.

This organization avoids outsourcing because it be-lieves contractors make money on other "rm's ine$cien-cies. Elec strongly believes that contract manufacturerssuch as SCI and Selectron are `giving awaya the manu-facturing and making their money by marking up rawmaterials and components on which they have signi"cantpurchasing leverage. Elec believes that purchasing is es-sentially the source of competitive advantage for thesecontractors, so manufacturers should be able to leverageit as well.

Elec is quite satis"ed with its purchasing leverage strat-egy of keeping all signi"cant purchases internal. Thisphilosophy was also con"rmed in a conversation withthis "rm's CEO.

7. Mixed success

The next group of organizations have had mixed suc-cess, meaning that there are aspects of their managementand control of purchasing leverage that they are verysatis"ed with, and other aspects that they would like toimprove.

7.1. Oil

The "rst mixed success organization is a leading do-mestic producer of oil and gas. Its situation in outsourc-ing purchasing leverage is a bit unique. First, it decided tooutsource the purchasing and management of its pipes,valves and "ttings to key distributors. It found thatbecause these key distributors (suppliers) serve a numberof customers, they can provide better in-stock service and

20 L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27

Fig. 2. Purchasing leverage in oil industry with distributors.

response from the manufacturers than Oil can. Sincethese suppliers/distributors serve a number of oil pro-ducers, they also get better volume leverage with themanufacturers than can Oil.

For example, Shell, Exxon, Texaco and Chevron alluse the same supplier/distributor, and let this sup-plier/distributor negotiate with the original equipmentmanufacturer (OEM) to utilize the combined leverage ofall these oil companies. Where volume is di!erent amongthese "rms, the supplier/distributor re#ects that in di!er-ential pricing. Oil also feels that it wins because goingdirect to the OEM as it had done in the past was morework and coordination for Oil. The supplier/distributornow gains a good deal of expertise on the purchaseditems, and Oil receives the bene"t of that.

While this sounds like a success, a the problem isoccurring at the next level. Oil outsources the buildingand maintenance of many of its re"neries and pipelines tocontractors. It has made arrangements with its sup-pliers/distributors for these third party contractors tohave access to its contracts. In addition, it has assumed inthe negotiation process that items used by these con-tractors will be part of its annual buy, and has negotiated

its pricing structure with its supplier/distributors accord-ingly. The problem is, it is having a di$cult time gettingits contractors to utilize those supplier/distributor con-tracts. Fig. 2 illustrates this situation. Scenario I showsthe contractors ignoring the supplier/distributor con-tracts and Scenario II shows the contractors properlyutilizing Oil's supplier/distributor contracts.

The lack of contractor compliance with Oil's suppliercontracts is creating a number of problems. First, it ismaking it di$cult for oil to achieve its annual volumecommitments with suppliers. It is hampering its e!orts tostandardize pipes, valves and "ttings for ease of mainten-ance and training. It is also creating higher costs inconstruction for Oil, although the contractors deny this.Oil does write the requirement of using its sup-pliers/distributors into the contracts. However, this ap-proach is not a tradition in the industry. Up to this point,the contractors have been allowed to utilize anothersupplier/distributor on the grounds of faster service,lower price in that region, better quality, etc. This is anissue that Oil is struggling with, and needs to re-solve/manage if its outsourcing leverage e!orts are tosucceed.

L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27 21

7.2. Personal computer industry (PC)

In many ways, this is the richest example due to thewide variety of outsourcing leverage options which thisorganization uses. By all accounts, this organization doesnot have a sophisticated, managed outsourcing leverageprogram. Things have just kind of `evolveda, not all ina manner which pleases PC. Each of PC's approachesand where they apply are summarized below.

7.2.1. Approach for contract manufacturersPC treats contractors very similarly to its own assem-

bly operations in terms of management. It negotiatesprices with all of PC's part/component suppliers, leverag-ing PC's total volume. It then actually orders all of theparts, and has them shipped to its own facilities where itkits them. It pays for all of the parts and keeps the valueof the kits in its inventory. It ships the kits to the con-tractors, who assemble them into "nished product andship them back to PC. It then adds the labor/subas-sembly cost to the parts cost for valuing the "nishedproduct in inventory. (Fig. 3, Scenario I).

When asked why it was done that way, PC replied thatthese contractors were supposed to temporarily "ll-in forshortages of manufacturing capacity. The kitting processis extremely expensive in terms of transportation time,inventory management/holding costs, and delays in cycletime to market. If it appears that these contractors will beused in long-term, PC will get out of the kitting and havethe contractors receive components directly. PC believesthat there is no reason that these should not be turnkeyoperations. However, this approach has been going onfor years in this organization in some areas, with no signof moving toward turnkey.

There is an exception to the desire for turnkey opera-tions. The exception is when there is proprietary techno-logy/components involved. Then PC wants to stay in theloop to ensure these assets are being properly managedand not `slipping awaya. As with others in this industry,they have been the victims of disappearing ship-ments/hijacked trucks.

7.2.2. Parts on allocationPC's policy for parts on allocation is to negotiate the

contract directly with suppliers for all parts used to getPC's volume leverage. It then purchases and pays for allof the parts and has them shipped directly to PC. It thendetermines the allocation of these parts to contractors.These parts are shipped out, but PC retains the parts onits books, also shown in Fig. 3, Scenario I.

When asked why PC follows this procedure of excess-ive handling, the reply was that it wanted to ensure thatits contractors did not order parts in PC's name, andthen use them in another customer's product. Recall thatAuto's sophisticated system could handle this. Thus, PCadds excessive cost and time delays in actually bringingthe components in house, to have tighter management.

7.2.3. Turnkey operationsPC would like to have as many of its contractors act as

turnkey operations as possible, because that reduces thecost, management and time delays to PC. PC manages itsturnkey suppliers as follows. First, it negotiates the con-tract price for all parts/components directly with thesupplier. Then it follows one of three strategies.

In the "rst strategy, where PC is a major part of thecontractor's business, it lets the contractor place ordersdirectly with the supplier, receive shipment and pay itsown bills. It does not worry about whether the contractorpasses PC's price savings on to other customers, or getsa lower price from the suppliers and absorbs the savings.In its opinion, that would just make the contractora stronger, more viable long-term competitor. This is byfar PC's preferred alternative in terms of ease of manage-ment, speed and overall cost of doing business.

The second strategy focuses on situations where PC isnot the major source of business to a contractor, andsome of the items are `on allocationa. Here, PC negoti-ates the contracts with parts suppliers based on its fullvolume. It has the contractors order from PC, and itpasses the orders on to the parts suppliers. The suppliersdrop-ship the items to the contractor, and bill PC dir-ectly. This scenario creates a much greater managementissue in terms of inventory reconciliation, payment andso forth, similar to that faced by MTPC before it out-sourced purchasing to its contract tape assembler. Insome circumstances, as mentioned above, PC will eventake delivery of critical parts and then drop-ship them tocontractors. It has done this because it "nds that, despitethe delays and added costs, it is actually easier for PC tokeep track of what happened to the inventory that it haspaid for if the inventory passes through its facility atsome point. This seems highly ine$cient, and is shown inScenario II.

Yet another scenario is one where the contractor buysparts/components directly from suppliers, and payswhatever that contractor's normal prices are. At the timeof order placement, the contractor must identify thatthose orders are to support PC. The supplier then givesPC a rebate directly, to get the price paid down to PC'snegotiated price. This approach is common where PCrelies both on internal manufacturing and contractoroutsourcing of the same items. Note the similarity be-tween this approach and the one used by Auto. In Auto'sapproach, it took responsibility for billing reconcili-ations. In PC's approach, that burden is shifted to thesupplier of inputs.

PC uses the rebate approach under a number of cir-cumstances. First, it is used when PC does not want thecontractor to have access to its pricing information forwhatever reason. Second, it used when the supplier doesnot want the contractor to know the prices that PC ispaying. This latter scenario is actually fairly commonwith OEMs, and was depicted in Fig. 1.

22 L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27

Fig. 3. Leverage in the personal computer industry.

PC is somewhat dissatis"ed with its current approach,as it has evolved rather haphazardly rather than as a re-sult of strategical planning and execution. Some of thechanges that PC would like to make are: less physicalpossession of inventory/handling of items that will go tocontractors; less paying for/keeping on its books mater-

ials/components to be used by contractors, less involve-ment in the management of materials to be used by thecontractor overall, beyond some price negotiations.There are some internal political factions that believesthat PC needs to touch/own many of these componentswhich may make it di$cult to implement these changes.

L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27 23

Fig. 4. Decision process for managing leverage of input purchases when end item is outsourced.

8. Summary

All of the organizations used as case studies above areextremely successful "rms with a history of growth, prof-itability and market leadership. Yet many di!erent ap-proaches to outsourcing purchasing leverage are used.The sections below present the recommended ap-proaches for managing purchasing leverage while out-sourcing manufacturing based on combining transactioncost theory with the case study "ndings.

8.1. Translating xndings into purchasing leverage issues

Extensive literature searches did not reveal any studiesperformed which look speci"cally at how to best managethe potential loss of purchasing leverage on inputs whenthe manufacture of the end items is outsourced. Theliterature focuses on the role of the purchasing functionin ousourcing (Ellram and Maltz, 1997; Purchasing,1998a). The analysis below assumes that a fundamentaldiscussion has already been made to outsource manufac-turing. The question addressed is how to manage (gover-nance structure) the purchase price of inputs used in theproduction of the outsourced item. The decision ofwhether to be actively involved in the selection andmanagement of suppliers used by the subcontractor is

not speci"cally explored here, as it may or may not bea separate issue.

8.2. Transaction frequency

Based on TCA, purchasing leverage is only possible forrecurrent transactions conducted with the same supplier.If the buying organization does not place signi"cantbusiness with one supplier, there is no leverage to begained. For recurrent transactions, the buying organiza-tion must realistically view its purchasing leverage versusthat of the contractor. This is shown on the "rst node ofthe decision tree in Fig. 4. If a contractor performs similaroperations and purchases similar materials on behalf ofa number of customers, it may be able to get betterleverage and service than the buying organization. Onthe other hand, if the contractor is relatively small, oruses the materials/components only to support this par-ticular customer, the customer (buying organization)may be able to get greater leverage.

8.3. Asset specixcity/purchase specixcity

The second major factor in the TCA focuses on thespeci"city of assets involved, which can also relate tothe speci"city of the purchase in question. If the item

24 L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27

purchased is a unique, non-standard item, there is noreason to conclude that either the contractor or thebuying organization could do better in terms of price. Onthe other hand, the buying organization or the suppliermay not want interference in its relationship with thesupplier of this speci"c component, so may contractdirectly rather than working through the contractor.This increases transaction costs as both the contractorand the buying organization become involved with thesupplier. These additional governance costs must beweighed against the potential reduction in the risk ofopportunism in uncertain markets. This could also bereferred to as economies of scope in purchasing.

If the buying organization believes that it hassuperior talent in the purchasing department in terms ofnegotiating contracts with suppliers, versus thecontractor, it may want to be involved in the contractnegotiation as well. In cases where the contractor hashigher total volume than the customer, the customer maywant to utilize the contractor's leverage. This issue isshown in the second node of the decision tree in Fig. 4.Whether it is possible to leverage this is explored in thenext section.

8.4. Economies of scale

Based on TCA, economies of scale also have an a!ecton whether an item should be made internally or pur-chased from the outside. According to TCA, if the organ-ization is buying a standard item from a contractor whosupplies like items to other customers, the contractorshould be able to achieve the lowest material/componentcosts based on its total leverage. This was the situation inthe "rst case study MTPC. On the other hand, if thebuying organization is spreading its volume for like itemsacross a number of contractors, none of whom purchaseas much as the buying organization purchases in total,the buying organization can likely get the best leverage.Thus, the buying organization should negotiate directlywith the raw material/component suppliers, retaininggovernance over the price relationship with the supplier.This was the case for both Auto and IM. This is shown inFig. 4 as `customer contract directly with suppliera. If,for some reason the company does not want the con-tractor to know prices, it can either mask prices, as wasthe case of Auto, in Fig. 1, or use as rebate scheme, asused by PC. Both of these options require a computersystem that is able to reconcile actual usage with thecontractor's order, to make sure that the customer (buy-ing organization) is paying for the proper quantity ofmaterials/components (masked price) or getting theproper rebate (rebate structure). The implications of theleverage aspect of the outsourcing decision are furtherdetailed in Fig. 5. Fig. 5 provides another way of lookingat the key issues that a!ect the decision of how to bestmanage purchasing leverage when outsourcing.

If there is no concern with contractor knowledge of theprice negotiated for materials/ components, by far themost e$cient approach is to let the contractor orderdirectly from the supplier. IM followed this approach forresins and standard electrical components.

Does it ever make sense to pay for the goods directly,`owning thema while the contractor manufactures?From an e$ciency, accounting and control standpoint,the authors believes it makes little sense. Unless a com-pany's accounting and control systems are so poor thatthis is the only way to monitor usage and payment, it isa waste of resources. If this is the case, a better investmentcould be made in improving the accounting systems.Yet this approach was used by IM for proprietaryparts/components and by PC in allocated parts (Fig. 3,Scenario I) and some turnkey operations (Fig. 3, ScenarioII). If contractors are so untrustworthy that they maydivert parts to other customers, does the fact that thecustomer (buying organization) has paid for those partsdeter this practice?

Perhaps the most ine$cient of all practices is to phys-ically take possession of the goods from the supplierbefore shipping them to the contractor. This createsdelays in time, potential errors, damage, wasted handlingcost and so on. It can be justi"ed if kitting is trulynecessary, as in the case of PC (Fig. 3, Scenario I). Yet PCalso did this for parts on allocation. Why should phys-ically touching the parts improve control over a `maskaor `rebatea situation, where the supplier noti"es thecustomer (buying organization) of contractor activity? Itseems extremely ine$cient, and implies a very tight gover-nance system, beyond simply supplier price management.

8.5. Uncertainty

As posited in TCA, in situations where external andbehavioral uncertainty is low, the market provides themost e$cient alternative. For purposes of this paper,behavioral uncertainty is a function of relative trust andpower, i.e., will the contractor try to take advantage ofus? As market uncertainty increases, some type of obliga-tional contract is preferred, protecting both parties frommarket contingencies. As uncertainties become extremeand unpredictable, so that they cannot be anticipated incontracts, and the risk of opportunism is high, the buyingorganization may want to protect itself by being directlyinvolved in the buying. However, as Williamson (1979)points out, sometimes the third party/contractor is ac-tually better able to absorb uncertainty because it canspread its risk over a number of customers. Thus, if theuncertainty in the market comes from customer demandrather than the suppliers or contractors, then the buyingorganization may want to leave the buying of mater-ials/components to the contractor.

In practice, it seems that uncertainty regarding howtrustworthy the contractor is plays a greater role in

L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27 25

Fig. 5. Relative leverage of customer "rm versus contractor.

outsourcing purchasing leverage than does market un-certainty. In a trusting relationship with a contractor, thebest approach is to work directly with the contractor toenjoy that contractor's lower `volume pricea if the con-tractor has more volume than the buying organization.This was the approach pursued by MTPC, who pays thesupplier cost plus a markup and Oil, for purchases fromdistributors (Fig. 2, Scenario II). The latter pays a lowerprice based on its relative prices than low-volume cus-tomers of its suppliers/distributors, and higher pricesthan higher volume competitors. Thus, it still enjoys its`relativea leverage, but even more so because of thesupplier's/distributor's greater overall leverage.

In a non-trusting relationship with a contractor, theorganization still has options. It can potentially lose itsleverage by just allowing the contractor to handle thenegotiations with the supplier, and hoping it is gettinga fair deal. If the contractor has great enough totalleverage by combining all of its customer's volume, thismay actually be a `bettera deal for the customer thannegotiating with the supplier directly. However, as Elecwarned, this may be where the contractors are making allof their money. Does it really matter if the money is madehere, if the customer is getting a `bettera deal than itwould with direct suppliers negotiations? In this case,

purchasing leverage is no longer a source of competitiveadvantage, because the contractor may pass similar sav-ings onto all customers, unrelated to relative volume. Theresearchers were unable to document any situationswhere this was actually the case.

On the other hand, as shown in Fig. 4, the customercould negotiate with the supplier directly for pricing, andthen use a masked price or rebate structure. Whether thismakes sense depends on the relative leverage of thecustomer, and the comparison of the contractor's willing-ness to share leverage bene"ts of pricing, versus thecustomer's leverage price plus associated transactioncosts of still being involved in the purchasing aspecttransaction.

9. Conclusions

With the continued growth in outsourcing manufac-turing, concern over the loss of leverage on purchasesused in outsourced manufacturing will continue to grow.Clearly, the decision of how to handle the outsourcing ofpurchasing leverage is a complex, multi-criteria decision.Applying TCA to the case studies, it is clear that thepreferred options for managing purchasing leverage as

26 L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27

part of the outsourcing decision depends on: whether thetransaction is recurrent; how unique/speci"c the pur-chased item is; the relative economies of scale of thebuying organization versus the contractor; and the levelof trust between the buying organization and the con-tractor.

The biggest issue appears to relate to trust/the natureof the relationship with the supplier and even more sowith the contractor. As organizations come to recognizethe high level of mutual dependence and bene"ts ofcooperation, purchasing will need less direct involvementin the transactional aspects of outsourcing purchasingleverage.

Further research is needed to address several issuesrelated to outsourcing leverage of purchased inputs foroutsourced items. These include:

f How important is leveraging prices of inputs to theoverall pro"tability of contract manufacturers? If pur-chasing organizations insist upon negotiating theirown prices for purchased inputs, might they end uppaying more in the long run in terms of

fC higher prices for manufacturing expenses from thecontractors;

fC higher transactions costs due to managing the inputcontracts;

fC higher input prices than the contractor would havepaid;

f When competitors are outsourcing manufacturing tothe same set of contract manufacturers, are price andquality still relevant competitive issues, or is the "eld`leveleda on those factors?

f When negotiating with contractors, does the buyingorganization consider its relative cash #ow and "nan-cial position, tax, and other issues in determining

which "rm is in the best position to `owna the produc-tion inputs? Related to this, how do contractors decidewhat the markup should be for be for holding andmanaging inventory?

f The nature of relationships and issues between buyingorganizations and the contract manufacturers con-tinues to evolve. These relationships will provide a fer-tile "eld for research far into the future.

References

Clark, J.M., 1961. Competition as a Dynamic Process. The BrookingsInstitute, Washington, DC.

`Contract Manufacturing,a Purchasing, May 21, 1998b, 124 (8),38}39.

Ellram, L.M., 1991. Supply chain management: the industrial organiza-tion perspective. International Journal of Physical Distribution andLogistics Management 21 (1), 13}22.

Ellram, L.M., Maltz, A., 1997. The Tool of Supply Management in theOutsourcing Process. Center for Advanced Purchasing Studies,Temple, AZ.

Heide, J.B., 1994. Interorganizational governance in marketing chan-nels. Journal of Marketing 58, 71}85.

John, G., Weitz, B.A., 1988. Forward integration into distribution: anempirical test of transaction cost analysis. Journal of Law, Econ-omics and Organization 4 (2), 337}355.

`Outsourcing,a Purchasing, March 26, 1998a, 124 (4), 48}49.Quinn, J.B., Hilmer, F.G., 1994. Strategic outsourcing. Sloan Manage-

ment Review 36, 43}55.Venkatraman, N., 1997. Beyond outsourcing: managing IT resources as

a value center. Sloan Management Review 38 (3), 51}55.Williamson, O.E., 1981. The economics of organization: the transaction

cost approach. American Journal of Sociology 87 (3), 548}577.Williamson, O.E., 1979. Transaction cost economics: the governance of

contractual relations. Journal of Law and Economics 22, 233}262.Walker, G., Weber, D., 1984. A transaction cost approach in make-or-

buy decisions. Administrative Science Quarterly 29, 373}391.

L. Ellram, C. Billington / European Journal of Purchasing & Supply Management 7 (2001) 15}27 27