بازبینی برون سپاری استراتژیک

This paper analyzes a sequential game where firms decide about outsourcing the production of a non-specific input good to an imperfectly competitive input market. We apply the taxonomy of business strategies introduced by Fudenberg and Tirole (1984) to characterize the different equilibria and find that outsourcing generally softens competition in the final product market. If firms anticipate the impact of their outsourcing decisions on input prices, there may be equilibria where firms outsource so as to collude or to raise rivals’ costs. We illustrate our analysis using a linear Cournot model.

Outsourcing has become a widespread phenomenon in the industrialized world in recent times. Examples of industries where outsourcing is a key feature of the organization of production abound: aircraft, cars, computers, mobile phones, audio/video systems, mechanical watches, and so on. Casual evidence suggests that information technology (IT) and related business services are regularly contracted out in a large number of industries (Domberger, 1998), and econometric studies assign a prominent role to outsourcing in various industries.1Yet it is probably fair to say that the industrial organization literature on outsourcing is relatively thin.2 In a recent paper in this journal, Shy and Stenbacka (2003) have provided the first intra-industry analysis of the strategic incentives that oligopoly firms face when outsourcing the production of inputs.3 In their model, differentiated Bertrand duopolists can either undertake investments into in-house production facilities for a specialized input, or they can buy that input from a subcontractor, but at higher variable cost. Hence, if firms do not want to bear the fixed cost of investing into in-house production facilities, they have to incur higher marginal cost for sourcing the input over the market. The present paper analyzes a similar trade-off, but differs from Shy and Stenbacka in three crucial aspects. (i) We focus on outsourcing the production of a non-specific input rather than an input specifically tailored to the needs of final good production. The equilibrium market price of this non-specific input will depend on the vertical structure of the industry since a firm’s outsourcing decision affects input demand (and possibly also input supply), as suggested by the literature on successive oligopolies. We feel that it is natural to account for changes in input prices, even though the previous outsourcing literature has largely ignored them. 4 (ii) We analyze sequential rather than simultaneous firm decisions about the production mode. In this setting, we study the conditions under which the first-mover may adopt outsourcing to raise rivals’ cost 5and when outsourcing may serve as an instrument of collusion in the final product market. (iii) We propose a reduced-form approach towards analyzing outsourcing decisions to accommodate for various forms of product market competition (including Bertrand competition with differentiated products, as in Shy and Stenbacka). With the reduced-form profit functions in place, we adopt the taxonomy of business strategies introduced by Fudenberg and Tirole (1984) to provide a general discussion of the potential strategic outsourcing equilibria. Finally, we illustrate our analysis using a linear Cournot model as a specific example. Our main results are the following. First, in contrast to Shy and Stenbacka, we find that there may be asymmetric equilibria where one firm buys the input from an existing input market, whereas the other firm produces the input internally. The difference stems from the fact that we consider a non-specific input good, whereas Shy and Stenbacka focus on a specific input good. In their setting, equilibria with only one firm outsourcing cannot occur since this firm would both have to cover the entire fixed cost of the supplier and face higher marginal costs than with in-house production.6 In our setting, asymmetric equilibria may emerge, as the non-specific input good may be bought from an existing market without having to cover the entire fixed cost of any given supplier. Second, asymmetric equilibria are typically driven by the changes in input prices associated with outsourcing. It is thus crucial to incorporate input price effects into the analysis of strategic outsourcing. To see this, consider an asymmetric equilibrium where the first firm strategically abstains from outsourcing so as to induce outsourcing by the second firm. The rationale of the first firm’s behavior is straightforward: by preventing an initial increase of the input price, the second firm is induced to outsource and thus to increase its own marginal cost. Unsurprisingly, there is no reversed asymmetric equilibrium where the first firm outsources so as to prevent outsourcing by the second firm: The first firm is not willing to prevent outsourcing by own outsourcing, as it will benefit from the second firm’s marginal cost increase associated with outsourcing, provided that it has not already outsourced. Third, there may be a symmetric equilibrium where the first firm’s outsourcing induces the second firm to outsource. In this equilibrium, firms successively outsource production to the input market to raise their marginal costs, thereby softening competition in the final product market. That is, a “wave” of consecutive outsourcing decisions may serve as a collusive device. Interestingly, there may also be a symmetric equilibrium where the first firm does not outsource to avoid triggering outsourcing by the second firm. This equilibrium may emerge if the softening of competition generated by a wave of outsourcing decisions is insufficient to compensate the first firm for its marginal cost increase. In sum, our analysis suggests that outsourcing generally softens competition in the product market, irrespective of whether product market decisions are strategic complements or strategic substitutes. In the symmetric equilibrium where both firms source the input good over the market (the outsourcing wave), the softening of competition is collusive in nature. In the asymmetric equilibrium where only the second firm outsources, the softening of competition is strategically induced by the first firm, but the second firm’s marginal cost increase is self-inflicted. Finally, the first firm will strategically avoid triggering an outsourcing wave if the softening of competition associated with outsourcing is insufficient to increase its profit. The remainder of the paper is organized as follows. In Section 2 we introduce the basic setup of our analysis before we discuss the various candidate equilibria in Section 3. For illustrative purposes, a linear Cournot model is presented in Section 4. Section 5 concludes.

We have shown that firms may strategically outsource the production of an input good to an imperfectly competitive input market to soften competition in the final product market. More specifically, if firms face a trade-off between bearing the fixed cost of in-house production at low variable cost and avoiding this fixed cost (but incurring higher marginal cost) and if, in addition, input prices vary with the industry’s vertical structure, outsourcing may serve as an instrument of collusion or raising rivals’ cost. Our analysis of a reduced-form model has demonstrated that (depending on parameter values) three different types of equilibria may emerge: (i) There may be an asymmetric outsourcing equilibrium where one firm produces the input good in-house whereas the other acquires it from the input market. In this equilibrium, the first-mover follows a Puppy Dog strategy and strategically abstains from outsourcing to not prevent outsourcing by the second firm. By preventing an initial increase of the input price, the second firm is induced to outsource and thus to increase its own marginal cost. (ii) There may be a symmetric equilibrium where both firms outsource. In this equilibrium, the first firm decides to outsource to trigger further outsourcing by the second firm (the Fat Cat strategy). Firms successively outsource to mutually raise their marginal costs, thereby softening competition in the final product market. That is, firms may generate a wave of consecutive outsourcing decisions to collude in the retail market. (iii) There may be another symmetric equilibrium where none of the firms outsource. In this equilibrium, the first firm does not outsource to avoid triggering outsourcing by the second firm (the Lean and Hungry Look strategy). Intuitively, this equilibrium may emerge if the softening of competition generated by a wave of outsourcing decisions is insufficient to compensate the first firm for its marginal cost increase. The first firm will then strategically prevent successive outsourcing. Our results are based on a reduced-form analysis and apply irrespective of whether product market decisions are strategic substitutes or strategic complements. They suggest that it is crucial to account for the price effects of outsourcing both at the downstream and at the upstream level of the industry to understand better the economics of outsourcing decisions. In fact, upstream price effects associated with strategic outsourcing might be even more important in vertically related industries with specific (rather than non-specific) input goods. Future research will have to address this question.