Outsourcing with Quality Cometition: Insights from a Three Stage Game Theoretic Model Sang Hoo ae Deartment of Economics Clark University 950 Main Street Worcester, M 01610-1477 508.793.7101 sbae@clarku.edu Chung Sik Yoo Deartment of Economics Yonsei University 34 Maeji Wonju Kangwon South Korea yooec@yonsei.ac.kr Joseh Sarkis Graduate School of Management Clark University 950 Main Street Worcester, M 01610-1477 508.793.7659 jsarkis@clarku.edu ril, 008
Outsourcing with Quality Cometition: Insights from a Three Stage Game Theoretic Model bstract Outsourcing decisions by organizations have strategic and oerational imlications. Strategically, understanding the market and cometition is necessary to make effective outsourcing decisions. In this aer we recognize this concern and model the situation where an organization with uality and cost ressures and oerational strategies may arrive at different outsourcing solutions based on cometitor uality strategy traits. We develo a three-stage game-theoretic oligoolistic model based on differentiated roduct strategy and integrating uality exectations of the market. The model is solved for euilibrium oints on rice, outsourcing activity, and investments in uality. The results show that these decision factors are sensitive to market exectations and uality erformance of cometitors. Performance measures based on rofitability and market share results are also resented within this model. Observations and insights are also resented. Keywords: Outsourcing, Quality, Oerations Strategy, Price, Game Theory 1
I. Introduction Early in a roduct s life cycle, with an emerging market, business organizations tyically will erform most oerational functions internally as vertically integrated entities. s the roduct life cycle matures, and cometition increases, organizations will come under ressure to become more efficient. Efficiency goals may be met through asects of cost cutting such as continuous imrovement and business rocess reengineering efforts (itken, et al., 003). Organizational efforts to hel accomlish these tasks include strategic simlification of rocesses and focusing on organizational core cometencies (Prahalad and Hamel, 1990; Quinn, 1999). relatively straightforward method to achieve simlification and efficiencies of rocesses is to offload non-essential organizational activities, rocesses and functions to an outside arty, i.e. outsourcing. This rocess sin-off tendency is reinforced as features of the business resources, routines and activities e.g. technologies and infrastructures in articular - become more commoditized and knowledge about business best ractices diffuses as the roduct market exands and matures (Magnani, 006). Under this develoment, organizations are almost coerced to become more concentrated on selling their core cometencies which demonstrate a definite comarative and cometitive advantage in the roduct market. Even though outsourcing has become a concet with significant definitional dissonance, it tyically is concerned with an issue of roduct uality vs. costs (ron, et al., 008; Reyniers and Taiero, 1995). In this aer, we investigate this classical dilemma in a game theoretic ersective. y setting u a model develoed in the roduct differentiation literature in economics, we analyze a case where a uality cometing oligoolistic firm outsourcing its inuts from cometitive subcontractors decides on key variables sanning over three stages; roduct uality (via R & D investment), the rate of outsourcing, and finally the rice of the roduct. This roblem is modeled and analyzed using a game-theoretic concet defined as Subgame Perfect Euilibrium through backward induction. This aroach to the issue of outsourcing has several advantages over other models for understanding the dilemma of
uality vs. costs tradeoffs. First, this model exlicitly introduces the realities in which there are uality differences of a roduct, manufacturers with brand names comete in terms of ualities as well as in terms of rices. Second, we find an outsourcing strategy is generally adoted by a firm that is less sensitive to roduct uality. That is, its oerations strategy would ut greater emhasis on cost reduction and metrics rather than a uality based roduction and oerations strategy. Hence this firm should be aware of the fact that a ossible degradation of a uality through outsourcing might affect its market share through uality cometition with other oligoolistic firms. This feature of the conseuences of outsourcing may be catured and analyzed by our model. Third, outsourcing may affect a customer s reference and utility. The resonses of customers to outsourcing are incororated using economically modeled customer utility functions. Fourth, outsourcers faced with uality erformance cometition are generally under an oligoolistic market structure, while subcontractors with easy entry and exit who deend on lower wages are under strong cometitive ressures can be evaluated. The remainder of this aer begins with some background discussion that motivates the roblem and issues faced in outsourcing. Then the three stage game theoretic model is introduced. We then rovide some characteristics of the solution to this model. Some initial analysis and insights through an exerimental arametric analysis rovides some initial insights follow. Finally a conclusion with summaries of our findings and extensions for future investigation are rovided. II. Outsourcing and the Quality-Cost Cometitive Environment Desite oular usage of the term outsourcing by media and the ublic, there still remains some confusion over the definition of outsourcing. ccording to hagwati et al (004), during the 1980s, outsourcing tyically meant a situation when business firms exanded their urchases of manufactured (hysical) inuts from outside rather than inhouse. For instance, car comanies may urchase their essential hysical inuts such as tires, brake system, window cranks, assenger seats, and so on from another comany which may have a long term contract with the original firms. Currently, however, in which commodity (whether it is a good or service) information is easily accessible around the globe with a click on a mouse, the meaning of outsourcing exands into a secific segment of the growing international trade in services (ron et al., 008). This 3
environment alters the definition slightly to take a form of urchase of services abroad, either through transactions by firms or with direct consumtion urchases by individuals 1. Delegating a secific roject or a resonsibility to an outside third-arty entity, sub-contractor, a secialized firm, or even an overseas roduction develoment unit, whether it is a roduct or service, is what we will define as outsourcing. It generally occurs for roduction rocess activities. The original roducer would emloy an outside roducer and contracts to make ortions or comonents of a roduct (or some rocess of roduction) using its own resources. There is amle evidence that artial outsourcing is a significant market ractice where it was recently estimated that only the information technology (IT) outsourcing ractices to offshore locations reresented a $178 illion market in 005 (Chakrabarty et al., 006). In the mobile telehone industry, Nokia, Motorola, and Ericsson outsource mobile handsets at between 15%-40 % rates (Economist, 00). Part of the argument for outsourcing is the decrease in roduction costs. Some estimates have shown that effective outsourcing can reduce costs by overall roduction costs by 0-40% er year (Domberger, 1995; Domberger et al 00). While cost saving may be a dominant reason behind the rush toward outsourcing for many firms, otential uality degradation resulting from unscrutinized offshore subcontracting may become a highly detrimental threat as in the case of Mattel in the toy industry. The recent case of mass recalls of Mattel toys rovide an excellent examle of the dilemma faced by an outsourcing firm. The toys were made by Chinese local firms through a tyical offshore outsourcing arrangement, called Original Euiment Manufacturing (OEM). In general, merican toy makers design and rocess a roduct and then transfer a knowledge regarding the roduct to subcontractors in China, buy back the roducts, and sell them to domestic customers with its own brand. Outsourcing via 1 The WTO categorizes four different ways in which services can be traded. Mode 1 refers to arm s-length suly of services with both of the trading artners remaining in their resective locations. This tye of trade has come into existence through IT revolution and unlike goods trade, it could not readily be subjected to customs insection. In Mode, as in tourism and medical care, service reciients move to the location of service roviders. Mode 3 categorizes a case where the service rovider establishes a commercial resence in another country as in banking and insurance. In general, this tye of trade reuires foreign direct investment. In Mode 4, as in construction and consulting, service roviders move to the location of the service buyer. When they discuss the roblem of outsourcing, most economists and olicy makers have meant trade in Mode 1 services. (hagwati et al(004)) See for instance, The Globe and Mail (007). 4
offshore roduction enabled the domestic firms to enjoy low roduction costs with much less managerial underinnings. U.S. roducers did not fully recognize, however, that a lower cost ultimately led to a uality degradation, to such an extent that the final roduct couldn t kee u with the domestic safety standard for the roduct, which haened to grow higher with income; local roducers in China could not see any incentives to follow high uality standards set in the US, since the Chinese government was much less stringent to uality standards and the original roducer in the US did not elaborate to monitor the erformance. The worldwide recall of nearly 0 million lead-ainted toys further damaged outsourcing from China, coming after et food found to contain melamine, toothaste tainted with diethylene glycol and tires that searate at the treads. This is a case where a low uality (cost-driven) roducer underestimated the value of monitoring costs, inducing more euilibrium rate of outsourcing, hence naively enjoyed an exansion of market and high rofit. Part of the outsourcing uality degradation arguments and difficulties not only relate to roduct uality, but other dimensions of roduct uality such as effective customer service and roduct delivery (Tan et al., 1998). This cost/uality tradeoff for outsourcing is a common enough occurrence and concern that articular attention to this strategic decision and tradeoff is warranted (Reyniers and Taiero, 1995; Wadhwa and Ravindran, 007). In economics, outsourcing ractices of firms are characterized as an imortant toic related to the boundaries of the firm within the framework of the theory of organization. Hence there exist many interesting and imortant aers using the theory of transaction costs, contract theory, industrial general euilibrium, vertical and horizontal integration and differentiation, among others to address this issue and suly chain relationshis. The research and theoretical frameworks we focus on in this aer has seen relatively little study. There are a few oligoolistic models using Cournot and ertrand cometition which do relate to our research. Kamien, Li, and Samet (1989) alies an auction aroach in a duooly framework with ertrand (rice) cometition, allowing subcontracting to each other. Shy and Stenbacka (004) exlores the strategic nature of outsourcing under Cournot cometition in the final goods market. Even though these and other aers have investigated outsourcing from a game theoretic ersective, 5
the inclusion of various customer exectations and uality dimensions, have not been comleted. Outsourcing in a three stage game framework has not yet seen any develoment in the research literature. We take advance this area of study by roviding a three-stage game model that rovides significant and robust managerial insight. III. The model To begin the model develoment, we consider a duooly model with vertical differentiated roducts introduced by Mussa and Rosen (1978). It is assumed that customers are heterogeneous in their marginal valuation of roduct uality, or uality exectations which is denoted by v. There is a oulation of customers whose total number is eual to N and each customer buys at most one unit 3. For simlicity, we assume that a uniform distribution of customer tyes and their uality exectations is given by vi ~ U[0, N ]. For a given rice j and level of roduct uality θ j with {, } j =, and reresent two cometing firms where firm is seeking to make an outsourcing decision, the net utility for a customer with uality exectation v i is def i θ j i θ j u( v ) = v + (1). In this aer the uality of each firm is determined through its roduct innovation which is investment in the uality of its roduct. We assume that firm j chooses its uality θ j at eriod one with a research and develoment (R&D) cost 4 def of ( ) ( ) Θ θ j = β θ j θ if j θ, 0 otherwise. The R&D rocess or uality imrovement investments are not always reuired since each firm is assumed to be endowed with a minimum uality level 5 θ, which is assumed to be large enough to cover the market. 6 θ > 3 unit may be reresentative of a single roduct, a lot of roducts, or even a long term contract. The exectation here is that for this unit the uality reuirements and exectations are the same. 4 Research and Develoment costs would be used to imrove roduct uality, this cost may also incur costs for imroving rocess uality through Six-Sigma, Total Quality Management, or a ISO 9000-like rograms. 5 Here θ is assumed to draw constant utility for customers indeendent of their uality exectations. 6 covered market means that the minimal exectations of all customers are met with this level of uality. Some aers adot a vertical differentiated roduct model with covered market. See Crames and 6
Producing final roducts reuires various inuts which can be obtained by different roduction modes. We assume that each final roduct of firm j reuires I inuts which the total number of inuts is normalized to one (cometing roducts are assumed to have eual levels of inuts). For acuiring each inut, two otions are oen to firm j: in-house roduction and outsourcing via outside subcontractors. [ 0,1] r denotes the ercentage of inuts roduced by the outside subcontractors. We assume that outside subcontractors are secialized in roducing inuts for the roducts in a cometitive market with normalized zero unit costs. Therefore, the rice of outsourced inuts is assumed to be zero. On the other hand, the unit cost of in-house roduction of an inut is c for firm j (j=, ). This assumtion reflects the assumtion that in-house roduction is less efficient, more costly, than outsider subcontractors and one of the main factors for firms to outsource their inuts roduction to lower its unit cost of roduction. However, outsourcing also incurs two different tyes of costs, exected uality erformance costs and monitoring costs. To comment briefly on the two tyes of costs, we first assume that the roduct roduced artially by outside subcontractors entails reduction in exected (erceived) uality erformance, denoted by α ( ) def r = αr which is assumed to be constant across all customers. The arameter α catures the marginal imact on the exected uality erformance. We assume that an increase in the outsourcing rate has a negative imact on exected uality erformance due to increased difficulty in managerial coordination and oversight by the roducer or the incomlete nature of contracts on the uality of inuts roduction between the outsourcer and the outside subcontractors. We assume that these costs are the same across all final customers or at least indeendently distributed with the valuation of customers for the roduct. The second tye of cost incurred with outsourcing is a firm s increased monitoring cost denoted by m( r ) = γ r to ensure the uality of inuts roduced by outside subcontractors. It is assumed to be an increasing convex function of the rate of outsourcing. The arameter γ catures the marginal imact on the outsourcing monitoring costs. def Hollander (1995), oom (1995), Ecchia and Lambertini (1997), Maxwell (1998), Wang and Yang (001), and Wang (003). 7
In order to emhasize exlicitly the strategic effect of outsourcing on cometition and uality level we carry out this task within a framework where only one firm with low uality roduct (i.e., firm ) is able to outsource its roduction. Given the outsourcing rate r the customer s net utility becomes θvi+ θ if the customer buys roduct u( v i ) = ( θ v + θ αr ) if the customer buys roduct with the outsourcing rate r (). To hel solve the outsourcing roblem we consider a three-stage game with the following stages: i Stage one: Each original roduction firm ( and ) chooses its R&D or uality investment to imrove its level of uality ( θ j ). Stage two: Firm decides the rate of outsourcing the roduction of its inuts ( r ), thereby firm roduces ( r ) and buys ( r ) ortion of inuts from the outside subcontractors. 1 Stage three: Each firm chooses its rice ( j ) to maximize its rofit. IV. Solving for Euilibrium in the Three-Stage Game To find the Subgame Perfect Nash euilibrium (SPNE), we begin with eriod three. 1. Stage Three: rice cometition with given uality and outsourcing rate In Stage Three, with given uality levels ( θ, θ ) and the outsourcing rate r the two firms ( and ) comete for customers in terms of ricing. Even though there are ricing exectations, it is assumed that due to cometitive reasons the uality exectations are at a minimal level or higher in this uality cometitive environment. When customers make their urchase decision, they choose the otion that yields the highest net utility. We consider the case of θ > θ, without loss of generality, where firm is cometing rimarily on maintaining higher uality than its cometitors and firm will have the otion to further imrove its uality or further comete on ricing. In this 8
case, a customer s otimal choice between roducts or comanies for a given rice and uality level can be divided as follows: αr θ θ v < v αr θ θ urchase roduct from Firm urchase roduct from Firm Thus, for a given rice and uality level, there will be a customer with uality exectations denoted by vˆ that is indifferent between buying roducts from firm and 늿v+ = v+ r. This feature of the self-selection choice of if θ θ ( θ θ α ) customers between roducts is shown in Figure 1. Figure 1 about here Each firm s demand then can be described as = N vˆ = N αr θ θ = vˆ = αr and θ θ (3). Given the marginal cost c, firm, which is assumed will not outsource, maximizes its rofit by Max P ( c) N β ( θ θ) αr θ θ (4a). Similarly, given the reduced marginal cost via outsourcing, firm maximizes its rofit by ( ( ) ) ( ) r c β θ θ γr Max 1 αr θ θ (4b). 9
y solving each team s rofit maximization roblem, we derive the euilibrium rice () and market share () as follows. 1 ( θ, θ,r) = ( αr + ( 3 r) c+ N( θ θ) ) 3 (5a), 1 ( θ, θ,r) = ( 3 c r( α + c) + N( θ θ) ) 3 (5b), ( α c) 1 r ( θ, θ, r) = N + 3 θ θ and ( θ, θ, ) ( α c) 1 r r = N 3 θ θ (6a), (6b). The corresonding euilibrium rofits are as follows: ( r( α c) + N( θ θ) ) 9( θ θ ) ( ) π ( θ, θ,r ) = β θ θ ( ( ) ( )) ( ) r α c N θ θ and π ( θ, θ,r ) = β( θ θ) γr 9 θ θ (7a) (7b).. Stage Two: Choice of roduction mode. How much to outsource? We now turn to firm s otimal choice on its rate of outsourcing. The objective is to simly maximize euation 7(b) with resect to r, which yields the following necessary first-order condition: ( r( α c) ( γr ( α c) N)( θ θ) ) 9( θ θ ) π 9 ( θ, θ,r ) + = = 0 r (8), which gives us 10
and buys a r ( θ, θ ) In other words, firm roduces a( 1 r ) = ( α c) N( θ θ) ( α c) 9γ ( θ θ ) (9). roortion of its roduct inuts in-house r roortion of its roduct inuts from outsourced subcontractors. To ensure concavity of the rofit function with resect to r, we reuire the following second order condition: ( α c) ( θ θ ) π ( θ, θ,r ) = + < 0 γ r 9 (10) From this second order condition we derive ( ) ( ) c 3 γ θ θ < a< c+ 3 γ θ θ (11) which means the marginal disutility of outsourcing of customers on the exected uality erformance should not be too large or too small when comared with the magnitude of the firm s unit roduction cost, to have an interior solution of the otimal rate of outsourcing. lso, incororating the boundary condition of the outsourcing rate ( 0 r 1), we have ( ( ) ( )) 1 c+ N θ θ θ θ 36 γ + N θ θ α c (1). Combining these two conditions from e. (11) and (1) together, we derive the following constraint for the solution. ( ) 1 max c 3 γ ( θ θ), c+ N( θ θ) θ θ 36γ + N ( θ θ) α c (13). 3. Stage one: choice of uality investment and uality level 11
t eriod one, the firms choose the level of uality to maximize their own rofits, which are given resectively by ( ) (( α c) 9γ ( θ θ) ) N ( θ θ) ( α c) 6γ ( θ θ) (, ) ( ) π θ θ = β θ θ (14a) and (14b) 3 9N ( ) (( α c) 9γ ( θ θ) ) ( ) ( c) N( ) ( α c) 9γ ( θ θ ) γ θ θ α θ θ π θ θ β θ θ γ (, ) = The best resonse function can be derived in terms of its own first order conditions. 6 4 d π ( ) 3 ( θ, θ ) 1 α c N c N = 4 N 18 β( θ θ) dθ 9 + + (( α c) 9γ ( θ θ) ) α 9γ θ θ d π ( θ, θ ) 1 = + dθ 9, N 18β( θ θ) ( α ) (( c) ( )) 3 4 ( α c) N (( α c) 9γ ( θ θ) ) (1a), (1b) V. Emirical Investigation and Discussion Due to the comlex algebraic exressions in euations (1a) and (1b), it is mathematically intractable to derive the otimal euilibrium uality choice of the firms in exlicit functional forms. To understand the decision mechanisms of the firms and the results in from the game euilibrium numerically, we run some analyses. In this emirical analysis, we comare the effects on the decision variables (outsourcing ercentage ( r ), rice (, ), and investment in uality rograms ( θ, θ ) ) by the exogenous arameters, α (factor of marginal decreases in uality due to outsourcing), c (unit roduct cost), and γ (factor of marginal increase in monitoring costs due to 1
outsourcing). Tables 1, and 3 summarize these results, where we fixed other arameter values at N = 1, β = 1, and θ = 1, resectively. 1. The effect of the marginal disutility of customers on outsourcing (exected uality erformance) The results in Table 1a and Table 1b show that as the customers become more uality sensitive about outsourcing, their ercetions of outsourcer uality, the otimal rate of outsourcing, euilibrium uality and rice difference between the two firms will be lowered. In this situation, the relative market share and rofit of the high uality roducer (firm ) will tend to increase as exected since uality cometition tends to increase as this customer uality sensitivity increases. This result can be easily exlained through intuition. s the customers become more sensitive about the uality degradation from outsourcing, low uality, low rice roducer seeking through outsourcing will be discouraged. Thus, there will be ressures or incentives to increase uality investments and raise rice. Hence the euilibrium rice difference between the uality oriented () firm and the rice/cost oriented () firm will be lessened. If the low uality roducer be oriented toward uality cometition, the high uality roducer will gain, hence its otimal market shares () and rofits () will be relatively higher as the market s (customer s) uality exectations increase. Therefore, from this reasoning, we can make the following observation: Observation 1 The more the customers are sensitive to outsourcing uality, uality and rice cometition becomes more intense, and the high uality roducer will gain. Tables 1a and 1b about here We also see that customers can exect the rices of their roducts to increase as their sensitivities increase, since the otimal rices for maximizing the rofits for both firms tend to increase. Even though the rices increase for both firms, firm, the cost-driven firm, will lose market share as its rices get closer to the uality leading firm, as shown by the ratio ( ) in Table 1b. Overall, the otimal rofitability of firm will tend to increase in this environment. strategic marketing outcome of this observation is that 13
firm should try to further differentiate itself on uality and reinforce the ercetion in customers that firms that outsource will have oorer uality erformance.. The effects of unit roduction cost difference from subcontractors We have assumed that subcontractors are under cometitive ressure and normalized their unit roduction cost to be zero (guaranteeing a lower roduction cost than the original firms and ). In this situation, increases in the unit roduction cost of by uality cometing firms (firm ) resents the case where outsourcing accrues greater benefit in roduction cost. Tables a and b shows the effects on firms euilibrium decisions of this maniulation. Tables a and b about here These second stage results show a slightly different result from the revious stage. s the unit roduction cost (c) increases, the otential benefit of outsourcing tends to also increase. Hence, the otimal outsourcing rate ( r ) of the low uality, cost-driven, roducer firm () will tend to increase. The euilibrium uality differential (reresented by the ratio ( ( θ θ) ( θ θ) + + ) in Table b), will be higher since there will be more outsourced roducts. In this case, the outsourcing firm, the cost-cometitive, lower uality roducer, will exand their market share ( in Table a) and will have a higher relative euilibrium rofit ( π π in Table b). ut the effects on the euilibrium rices () are less clear. The direct effect of an increase in the unit roduction cost will be an increase in roduct rices. The indirect effect via outsourcing, however, will induce a rice decrease. The relative effects of these two contradicting forces in euilibrium deend on the intensity of uality cometition. The table shows that if the difference of the unit roduction cost is relatively minor between the original roduct firm () and the subcontracting firms, and hence outsourcing savings, are relatively insignificant, increases in the unit roduction cost will bring lower euilibrium differences in rices. This result occurs because outsourcing contributes little to uality in this stage, the lower uality roducer will be more sensitive to rice caused by a cost increase as in-house roduction is much more significant. s the difference of the unit roduction cost 14
becomes greater between the outsourcer and subcontracting firms, the lower uality roducer (firm ) will become less sensitive to an increase in the cost of in-house roduction. This situation will ultimately result in an increase in the euilibrium rice difference. This reasoning leads to an imlicit roosition that there might be a level of the cost incentive of outsourcing resulting in a minimum difference in euilibrium rices in oligoolies under uality cometition. lso, we can observe that as roduction costs increase, the uality oriented firm () will tend to invest more in its uality initiatives θ in Table ) on an absolute level and when comared to firm ( ( θ θ) ( θ θ) ( Table b). + + in Observation s the cost benefit of outsourcing becomes larger (c becomes larger), the euilibrium uality difference will be greater, and the outsourcing low uality roducer will gain. Remark -1 In the initial stage of outsourcing, where in-house roduction is dominant, cost increase in the outsourcer s industry will be associated with lowered rice differences. ut in the stage of significant ercentage of outsourcing, cost increase will bring clear uality difference and hence rice difference. Remark - Whether an increase in outsourcing, resulted from a cost increase in the outsourcer s industry, will bring an increase in rice difference between the high uality roducer and the lower uality roducer, is an emirical and oen uestion. We can ostulate that it rice differences may increase or decrease in difference (the ratio in Table b) amongst the firms may be estimated by emirically finding the minimum level of rice difference over the san of various degrees of outsourcing. Thus, if firm feels that the ricing difference can be imroved if they are not at this minimal difference level by maniulating the outsourcing rate ( r ). If the current stage of outsourcing is believed to be below the rate of that minimum level, an increase in the rate 15
of outsourcing due to the cost benefit will bring weakened rice cometition. Otherwise, the rice ga will be deeer. 3. The effects of monitoring costs It is intuitively clear that as the outsourcing monitoring costs increase (i.e., as γ becomes larger), the euilibrium rate of outsourcing ( r ) will decrease. The results shown in Tables 3a and 3b of the first stage of the three stage game, verify this intuition. These table results show that as uality monitoring costs for outsourcing increase, the otimal rate of outsourcing decrease, leading to a decrease in uality and rice difference. In this scenario, the relative market share of the uality oriented roducer () increases ( in Table 3a) and hence will ultimately gain ground on. This situation occurs as the euilibrium investment in uality by ( θ and ( θ θ) ( θ θ) + + ) actually decrease. nother surrising finding here is that even if monitoring costs increase, the euilibrium rofitability increases slightly. This increase in rofitability seems to occur from increases in euilibrium rices for ( ) and ( ). Observation 3. Higher uality outsourcing monitoring costs will lower the otimal rate of outsourcing, uality and rice differences, and will be beneficial to the uality-oriented roducer. One interesting, but very intuitive oerations strategy imlication from this result is that all other things being eual, industries with a lower level of standardization (higher customization and hence in need of more careful observation) will be less rone to outsourcing and a high uality roducer will ultimately gain. VI. Conclusion. In this aer, we analyzed firm decision roblems of outsourcing in light of oerations strategy and roduct differentiation by setting u a three stage game relating to factors of uality cometition, outsourcing, and rice cometition, resectively. y incororating the effects of outsourcing on customers welfare, uality imrovement (R&D) costs, and uality outsourcing monitoring costs, we were able to derive and observe some interesting results relating to decisions on outsourcing. We clearly observe 16
that there is a conflict and tradeoff associated with roduction cost benefits and uality degradation tyically associated with outsourcing in a uality cometitive market. The introduction of uality into an outsourcing decision framework is necessary in uality cometitive environment, but the results have imlications in most decision environments, even in situations where rice and uality are less tradeoff oriented. y borrowing the basic insights and theoretical develoments from vertical differentiation literature within economics, and by incororating an outsourcing decision roblem of the firms into a classical uality and rice cometition model, we were able to make a number of insights. First, the outsourcing decision will amlify the difference of firms in the relative investments in uality rograms. More recisely, based on our threestage game theoretic model, firms will choose more divergent uality management strategies when faced with outsourcing ossibilities. Second, if customers and the market becomes more sensitive to (ercetions of oor uality increase) with outsourced roducts, uality and rice cometition will become more intense with less euilibrium differences, and the uality-oriented oerations strategy roducer will ultimately gain. Thirdly, as the roduction cost gain of outsourcing increases, mainly through roduction cost increases in an outsourcer s industry, outsourcing will become more attractive and the uality differences will be larger. Price differences at the euilibrium may vary deending on the direct effect of the roduction cost increases on ricing and the indirect effect of cost reduction of outsourcing. Fourth, if outsourcing reuires additional monitoring of the outsourced roduction rocesses, the euilibrium rate of outsourcing will dro, leading to lowered rice and uality differences (hence cometition becomes intense). Within our framework, outsourcing brings exanded market share for the lower uality roducer and hence better ossibilities of rofit gain. lternatively, any decrease (imrovements) in outsourcing monitoring costs (e.g. through additional monitoring technology to imrove efficiency or more careful selection of high uality suliers (ron et al., 008)), imroving market ercetions of outsourcing uality, or increases in roduction costs will increase outsourcing incentives and ultimately benefit cost-oriented roducers by roviding an imroved cometitive (relative market share and rofitability) stance. 17
Well designed outsourcing may allow a firm to be more focused on strategically more imortant sets of tasks and to become more cometent in its core caabilities. ut it is also claimed that excessive outsourcing, without considering long term strategic conseuences, may lead to unduly heavy deendence on subcontractors, undermining its R & D caabilities and causing otential technology redation by those suliers (Wu et al (005)). dvances in our model may reuire that we incororate these other factors as otential strategic costs associated with outsourcing. Our model secifically addresses the issue that with uality cometition in mind, this conclusion may be misleading, the lower uality roducer by introducing and exanding the ossibility of outsourcing may already committed to go for a lower uality accetance of their roduct. This result does not mean that the industry as a whole will suffer from uality degradation. Rather, in fact, the high uality roducer will actually ut more labor for an ugrade, to go for greater differentiation as the uality roducer in this market in the threat of outsourcing cometitors. Hence, our results imly that outsourcing may actually rovide customers with more uality roduct choice than before. For a lower uality cost-driven firm, this claim of higher uality through outsourcing decisions might be very valuable to be accounted for, but for an analyst more concerned about the whole icture of the conseuences of outsourcing, the claim will be surrising, though it may describe the reality well enough. Thus, instead of just redicting whether organizations should outsource we find it is imortant to consider the roblem within a framework of general uality cometition rather than just exlaining and redicting the outsourcing market for the sake of outsourcers themselves. The imlications for oerations strategy are also retty clear here. Organizations seeking to outsource in a uality sensitive market should be wary of customer ercetions and that shifting to cost-based oerations strategies may entail risks. lternatively, for organizations in cost sensitive markets, having too much investment in uality initiatives may be detrimental to market share and rofitability. n aroriate oerations strategy balance is necessary deending on market characteristics, subcontractor caabilities, and organizational cometencies. Our aroach to outsourcing with a three stage game theoretic model to rovides significant insightful results, but there are some limitations, which rovide amle 18
oortunity for future research. In our model, we imlicitly assumed that the ualityoriented roducer does not outsource, but it excludes many interesting cases from strategic interactions among the firms, including some dynamics. This may be a future task of our research. References itken, J., Childerhouse, P., and Towill, D., 003, The imact of roduct life cycle on suly chain strategy, International Journal of Production Economics, 85 (). 17-140. grawal, V. and Farrell, D., 003, Who wins in offshoring, McKinsey Quarterly, Secial Edition : Global Directions, 37-41 ron, R., andyoadhyay, S., Jayanty, S., and Pathak, P., 008, Monitoring rocess uality in off-shore outsourcing: model and findings from multi-country survey, Journal of Oerations Management, 6 (),. 303-31. hagwati, J., Panagariya,., Srinivasan, T.N., 004, The Muddles over outsourcing, Journal of Economic Persectives, 18(4), 93-114 oom,.,1995, symmetric international minimum uality standards and vertical differentiation, Journal of Industrial Economics, 43(1),. 101 19. usiness Week, 003, The new global shift, Cover Story, 3 Feb. usiness Week, 005, Outsourcing innovation, Secial Reort, 1 March. Chakrabarty SK, Ghandi P and Kaka NF, 006, The Untaed market for offshore services. The McKinsey Quarterly, No. /006. Crames, C. and Hollander,., 1995, Duooly and uality standards, Euroean Economic Review, 39(1),. 71 8. Domberger, S., Hall, C., and Li, E., 1995, The determinants of rice and uality in cometitively tendered contracts, Economic Journal, 105, 1454-1470 Domberger, S., Jensen, P.H., and Stonecash, R.E., 00, Examining the magnitude and sources of cost savings associated with outsourcing, Public Performance and Management Review, 6(), 148-168 19
Ecchia, G. and Lambertini, L.,1997, Minimum uality standards and collusion, Journal of Industrial Economics, 45(1),. 101 13. Economist, 00, The fight for digital dominance : secial reort Nokia vs. Microsoft, 3 Nov. 67-9 Globe and Mail, 007, Don t bash China U.S. toy makers are at fault, 3 Set. Monday Grossman, G., and Helman, E., 00, Integration versus outsourcing in industry euilibrium, Quarterly Journal of Economics, 117, 85-10 Grossman, G., and Helman, E., 005, Outsourcing in a global economy, Review of Economic Studies, 7, 135-159 Hilsenrath, Jon E., 004, ehind outsourcing debate : Surrisingly few hard numbers, Wall Street Journal, ril 1, 1 Kamien, M., Li, L., and Samet, D., 1989, ertrand cometition with subcontracting, RND Journal of Economics, 0, 553-567 Magnani, E. (006). Technological Diffusion, the Diffusion of Skill and the Growth of Outsourcing in US manufacturing. The Economics of Innovation and New Technology, 15 (7),. 617-647. Maxwell, J. W., 1998, Minimum uality standards as a barrier to innovation, Economics Letters, 58(3),. 355 60. McCarthy, I. and nagnostou,., 004, The imact of outsourcing on the transaction costs and boundaries of manufacturing, International Journal of Production Research, 88 (1), 61-71 Mussa, M. and Rosen, S., 1978, Monooly and roduct uality, Journal of Economic Theory, 18, 301-317. Prahalad, C. and Hamel, G. 1990. The core cometence of the cororation, Harvard usiness Review, Vol. 68 No. 3, May-June,. 79-91. Quinn, J.. 1999. Strategic outsourcing: leveraging knowledge caabilities, Sloan Management Review, Vol. 40 No. 4, Summer,. 9-. Quinn, J.. and Hilmer, F.G., 1994, Strategic outsourcing, Sloan Management Review, 35(4), 43-55 Reyniers, D., and Taiero, C., 1995. The delivery and control of uality in sulier roducer contracts. Management Science 41, 1581 1589. 0
Shy, O. and Stenbacka, R., 004, Partial outsourcing, monitoring cost, and market structure, Canadian Journal of Economics, 38 (4), 1173-1190. Tan, K.C., Handfield, R.. and Krause, D.R. 1998. Enhancing firm s erformance through uality and suly base management: an emirical study, International Journal of Production Research, 36(l0),. 813-37. Wadhwa, V., and Ravindran,.R., 007, Vendor selection in outsourcing, Comuters & Oerations Research, 34 (1),. 375-3737. Wang, X. H. (003). note on the high-uality advantage in vertical differentiation models, ulletin of Economic Research, 55(1),. 91 9. Wang, X. H. and Yang,. Z. (001). Mixed-strategy euilibria in a uality differentiation model, International Journal of Industrial Organization, 19(1 ),. 13 6. Wu, F., Lee, H.Z., Chu, L.K., and Sculli, D., 005, n outsourcing decision model for sustaining long-term erformance, International Journal of Production Research, 43 (1), 513-1535 1
u, u u = θ v + θ i u = θ v + θ αr i 0 vˆ 1 v uy uy Figure 1. Customer s uality choice/ercetion ranges and utility values
Table 1. Numerical nalysis with resect to customer uality sensitivity to outsourcing (α ) (with c =, γ = 1) α r θ θ 0. 0.81594 1.44 0.6593.01991 0.530677 0.46933 0.911 0.194365 0.4 0.310 1.3419 0.69958.0756 0.566835 0.433165 0.34170 0.1780 0.6 0.189061 1.884 0.731.1197 0.594869 0.405131 0.3848 0.165948 0.8 0.153353 1.56 0.75573.16317 0.616617 0.383383 0.415097 0.15665 1 0.107 1.377 0.77511.045 0.63338 0.3666 0.440867 0.149553 1. 0.094377 1.83 0.79005.440 0.646086 0.353914 0.460788 0.14459 1.4 0.0689 1.41 0.80115.8341 0.65539 0.34461 0.475603 0.140418 1.6 0.0451004 1.6 0.80883.333 0.661747 0.33853 0.485839 0.137811 1.8 0.03033 1. 0.81333.3649 0.66545 0.33455 0.491846 0.13698 0 1. 0.81481.40741 0.666667 0.333333 0.49387 0.13580 π π In this table, θ =0 because of a corner solution to the otimal roblem. This result is not surrising since we generally obtain the maximum uality differences in euilibrium in the vertical roduct differentiation literature due to introduction of a minimum level of uality, θ, or θ j is bounded from below. Table and 3 show the similar results. 3
Table 1b: Numerical nalysis with resect to customer uality sensitivity to outsourcing (α ) ( with c =, γ = 1). α r ( ) ( θ ) + θ θ + θ π π 0. 0.81594.44 1.31656 1.13073 1.4978 0.4 0.310.3419 1.3053 1.30859 1.9175 0.6 0.189061.884 1.8838 1.46834.3048 0.8 0.153353.56 1.7393 1.60836.6506 1 0.107.377 1.5899 1.776.94789 1. 0.094377.83 1.4333 1.8554 3.19417 1.4 0.0689.41 1.674 1.90183 3.38705 1.6 0.0451004.6 1.090 1.95636 3.554 1.8 0.03033. 1.18993 1.98909 3.6086 0. 1.1693. 3.63636 4
Table a. Numerical nalysis with resect to roduct inut manufacture cost (c ) (α =1, γ = 1) c r θ θ π π 1 0 1. 0 1.81481 1.40741 0.666667 0.333333 0.49387 0.13580 1. 0.03033 1. 0.01333 1.5813 0.66545 0.33455 0.491846 0.13698 1.4 0.0451004 1.6 0.0883 1.7509 0.661747 0.33853 0.485839 0.137811 1.6 0.0689 1.41 0.40115 1.91098 0.65539 0.34461 0.475603 0.140418 1.8 0.094377 1.83 0.59005.069 0.646086 0.353914 0.460788 0.14459 0.107 1.377 0.77511.045 0.63338 0.3666 0.440867 0.149553. 0.153353 1.56 0.95573.335 0.616617 0.383383 0.415097 0.15665.4 0.189061 1.884 0 3.131.4441 0.594869 0.405131 0.3848 0.165948.6 0.310 1.3419 0 3.9958.53395 0.566835 0.433165 0.34170 0.1780.8 0.81594 1.44 0 3.4593.59463 0.530677 0.46933 0.911 0.194365 3 0.3441 1.5407 0 3.60657.61486 0.48368 0.516318 0.8837 0.1583 Table a. Numerical nalysis with resect to roduct inut manufacture cost (c ) (α =1, γ = 1) c r ( ) ( θ ) + θ θ + θ π π 1 0. 1.8947. 3.63636 1. 0.03033. 1.754 1.98909 3.6086 1.4 0.0451004.6 1.6198 1.95636 3.554 1.6 0.0689.41 1.5651 1.90183 3.38705 1.8 0.094377.83 1.5554 1.8554 3.19417 0.107.377 1.5899 1.776.94789. 0.153353.56 1.6719 1.60836.6506.4 0.189061.884 1.8105 1.46834.3048.6 0.310.3419 1.3015 1.30859 1.9175.8 0.81594.44 1.3336 1.13073 1.4978 3 0.3441.5407 1.3796 0.93679 1.0606 5
Table 3a. Numerical nalysis with resect to the level of monitoring costs(γ ) (α =1, c= 3) γ r θ θ π π 0.8 0.493564 1.7071 0 3.5181.719 0.40773 0.5977 0.137956 0.015 1 0.3441 1.5407 0 3.60657.61486 0.48368 0.516318 0.8837 0.1583 1. 0.63718 1.437 0 3.65334.7993 0.55307 0.474693 0.83801 0.10705 1.4 0.135 1.375 0 3.6867.91433 0.55165 0.44835 0.30187 0.03177 1.6 0.17953 1.3357 0 3.7088.9993 0.56979 0.43008 0.345939 0.196177 1.8 0.154434 1.3096 0 3.71768 3.04997 0.58308 0.41697 0.365087 0.19017 0.135633 1.915 0 3.7901 3.0934 0.59310 0.406898 0.379869 0.185109 Table 3b. Numerical nalysis with resect to the level of monitoring costs(γ ) (α =1, c= 3) γ r ( ) ( θ ) + θ θ + θ π π 0.8 0.493564.7071 1.54851 0.688399 0.6844 1 0.3441.5407 1.3796 0.93679 1.0606 1. 0.63718.437 1.3051 1.1066 1.34691 1.4 0.135.375 1.6364 1.304 1.5759 1.6 0.17953.3357 1.371 1.3446 1.7634 1.8 0.154434.3096 1.189 1.3984 1.91977 0.135633.915 1.0554 1.4576.0514 6