فن آوری اطلاعات، طراحی سازمانی، و قیمت گذاری انتقالی

We show how information technology affects transfer pricing. With coarse information technology, negotiated transfer pricing has an informational advantage: managers agree to prices that approximate the firm's cost of internal trade more precisely than cost-based transfer prices. With sufficiently rapid offers, this advantage outweighs opportunity costs of managers’ bargaining time, and negotiated transfer pricing generates higher profits than the cost-based method. However, as information technology improves, the informational advantage diminishes; the opportunity costs of managers’ bargaining eventually dominate, and cost-based methods generate higher profits. Our results explain why firms generally prefer cost-based methods, and when negotiated methods are preferable.

In the absence of competitive markets for internally traded products and services, why do some firms base transfer prices on internal costs and others allow managers to negotiate prices?1 Allowing managers to bargain over transfer prices has several drawbacks: the time that managers spend bargaining is lost to other productive activities; if negotiations take too long, the firm may miss some market opportunities; and if haggling leads to ill-will between managers, the firm may suffer from sub-optimal decisions and conflict (see Brickley, et al., 2004; Kaplan and Atkinson, 1998; Simons, 2000). Faced with these issues, why do firms ever let managers bargain over transfer prices? And what economic characteristics predict the preferred pricing method? In any large firm, local managers have private information—superior knowledge of local conditions, business processes, and potential cash flows. Other parties in the firm rely on local managers’ reports, but there are two important problems with this. First, a local manager may distort information; if truthful reports are desired, top management must provide appropriate incentives. Second, transferring local knowledge may be costly—it may be difficult or impossible for the local manager to fully describe the precise links between local information, the multitude of local decisions, all of the available alternatives, and the potential cash flows. And it may be difficult or impossible for other parties in the firm to quickly understand and act upon the local manager's reports. We show that it is the ability to communicate knowledge to top managers and others that is a key factor determining the preferred transfer-pricing method. Specifically, the more difficult it is to transfer the local knowledge of the supplying division, the more attractive is negotiated pricing. The costs of transferring local knowledge to top managers and others vary widely across divisions, across firms, and over time. A number of factors influence these costs: the nature and complexity of local knowledge; the cultural, educational, and on-the-job training backgrounds of divisional and top managers; the necessity for rapid responses to changes in local conditions; the firm's size; the technology for communicating local information; and the geographic reach of the firm (Brickley, et al., 2004; Christie, et al., 2003; Demsetz, 1988; Jensen and Meckling, 1992). We use the term “coarse information-technology (IT)” to refer to the limitations of the firm's formal and informal information systems when transferring local divisional knowledge to top managers.2 There is a continuum: in the extreme “perfect-IT” case, it is possible for the local manager to quickly and costlessly report on everything of local economic relevance (top managers must still provide truth-telling incentives if they desire truthful reports); at the other extreme, the local manager cannot provide any relevant reports whatsoever (reporting incentives are then not an issue). Of course, neither extreme is realistic: all firms rely on divisional reports, yet no reporting technology fully communicates everything about local conditions. In the extreme “perfect-IT” case, cost-based transfer pricing works well. Divisional managers are given contracts that provide them incentives for truthful reporting (this, of course, costs the firm but cannot be avoided). Top management then receives all the relevant information about the supplying division, computes the relevant outlay and opportunity costs of internal trade, and sets the price equal to the total relevant cost. The buying division's internal-ordering decision is then optimal from the firm's point of view; but since top managers could simply impose this internal-ordering decision, it is not clear why the firm is organized into profit centers in the first place. Coarse IT provides an answer: if it is not possible to communicate all the relevant local information, two profit centers and cost-based transfer pricing guarantee superior decisions (Vaysman, 1996). If local managers cannot fully and quickly report on everything of relevance to local cash flows, top management must rely on a concise but incomplete estimate of the supplying division's costs to optimally set the cost-based transfer price. Even with contractual incentives for truthful reporting, top management cannot get all the information necessary to compute the relevant cost of internal trade. At best, the cost-based transfer price is set to equal top managers’ estimate of the relevant cost (given their necessarily incomplete information). In Section 3 of the paper, we show that, since this cost estimate is by necessity sometimes too high and sometimes too low, the buying division's internal-ordering decision is sub-optimal. On the other hand, when divisional managers bargain over the transfer price (and thus, effectively, over whether to trade internally), top management can use divisional-profit-based compensation to ensure that the managers come to a price agreement only if it is in the firm's interest to transfer internally. The internal-trade decision thus incorporates the managers’ local information better than under cost-based pricing. This is the informational advantage of the negotiated method. We compare this advantage with the organizational and opportunity costs the firm faces from managers’ bargaining. 3 To capture these costs, we use a multi-period offer–counteroffer bargaining model. We show that top management can use compensation schemes and bargaining rules to guarantee that divisional managers agree on a transfer price quickly. Negotiated transfer pricing is then superior to the cost-based method when offers and responses happen rapidly (this is formalized in Theorem 2 and Theorem 4). We also consider the implications of improvements in a firm's knowledge-transfer systems. As IT improves, expected profits increase under both methods. But they increase faster with cost-based pricing: IT improvements enhance the supplying manager's ability to communicate local knowledge and, consequently, top management's ability to set transfer prices that approximate the relevant costs of internal trade. The informational advantage of the negotiated method decreases with IT improvements, and is eventually outweighed by the opportunity and organizational costs of managers’ bargaining. Thus, with sufficiently fine IT, the cost-based method generates higher profits than negotiated transfer pricing; we document this in Theorem 3. An important caveat is that our study does not address two important practical issues. First, to simplify the analysis, we ignore the effect of taxes on transfer-pricing method choice. Income taxes, tariffs, and domestic content laws influence a firm's transfer-price decisions, while national tax laws and international treaties constrain the firm's choices.4 Our results apply directly to situations where the tax effects are not an overriding consideration for internal trade, either because the trade does not cross tax jurisdictions, or because the firm maintains a system of transfer prices for performance evaluation separately from tax transfer prices (note though that in the latter case the taxation authorities can argue that the tax transfer price has no legitimate business purpose and take legal action to use the internal system for taxation purposes; see Ernst and Young, 1999; Springsteel, 1999). Where tax effects are important, our model can be extended to include these effects, along the lines of Baldenius et al. (2004). Second, when a firm is implementing a cost-based pricing system, divisional managers may expend time and effort attempting to influence the system's design. But, as is typical in standard models of mechanism design with private information, it is optimal for the top management to disallow any influence activities by proposing the cost-based system to the managers in a “take-it-or-leave-it” fashion. Our analysis of optimal contracting under cost-based pricing thus does not admit the possibility that there are opportunity costs of managerial time from system-design-stage influence activities.

A challenge for multi-divisional, vertically integrated firms is to delegate operating decisions in a way that encourages divisional managers to use their private information in the best interests of the firm as a whole. This is complicated by restrictions on the flow and availability of information within the firm. That better and cheaper information leads to improved profits is recognized; less obvious are the implications of IT improvements for organizational design. Our results show that a firm's informational environment is a key factor determining its transfer-pricing method. When coarse IT prevents divisional managers from transferring local knowledge about the firm's production process, we show that properly structured negotiated transfer pricing has an informational advantage over the cost-based method: with the right incentives, managers agree on transfer prices that approximate the costs of internal transfer more accurately. Consequently, when opportunity and organizational costs of bargaining are sufficiently small, the informational advantage of the negotiated method generates higher firm profits. However, improvements in IT reduce the informational advantage of the negotiated method, because the value-destroying internal trade under cost-based pricing diminishes. With sufficiently fine IT, the cost-based method provides higher profits than negotiated transfer pricing because, all else equal, the remaining opportunity and organizational costs of managerial bargaining outweigh the profit destroyed by inefficient internal trade under cost-based transfer pricing with coarse IT. Our results have several empirical implications. First, the results help explain survey evidence in Vancil (1978) who found that: (i) firms comprising unrelated businesses use negotiated transfer pricing significantly more than cost-based transfer pricing (34% versus 15%) and (ii) vertically integrated firms operating within a single product market use cost-based transfer pricing significantly more than negotiated transfer pricing (49% versus 12%). Managers in firms comprising unrelated businesses are more likely to have diverse educational, functional, and training backgrounds; the cost of transferring divisional knowledge to top managers is thus higher than in vertically integrated firms in a single product market. Vancil's evidence is thus consistent with Theorem 2 and Theorem 4. Second, our results imply that firms with the following characteristics are more likely to prefer negotiated transfer pricing: (i) understanding divisional operations requires specialized education or training; (ii) divisional operations are physically located far from the headquarters; (iii) divisional operations are in quickly-changing environments requiring rapid responses; (iv) lack of sophisticated enterprise–resource–planning (ERP) systems; and (v) large firm size.18 Third, Theorem 3 offers testable predictions. The most direct prediction is that the ratio of firms with cost-based transfer pricing to those with negotiated pricing is higher after the 1990s IT boom than before. During the 1990s, many firms invested heavily in internal information systems, benefited from increased computing power per manager, and used ERP systems to streamline communications between different functional lines in large firms (Cairncross, 2000; Sircar, et al., 2000). The investments in communication and reporting systems lowered the barriers to internal communication. Theorem 3 predicts that this reduction in knowledge-transfer costs increases the attractiveness of cost-based transfer pricing. Fourth, the results can predict preferred pricing methods in modern business-to-business retail alliances. Any links between two companies’ information systems are likely to be coarse because the benefits of complete sharing of information will be outweighed by associated proprietary costs. If so, a testable prediction of Theorem 2 and Theorem 4 is that the transfer prices in these alliances are more likely to be negotiated than to be based on the supplier's cost. Finally, our results have practical implications. All else being equal, firms with low costs of transferring divisional knowledge earn higher profits with cost-based than with negotiated transfer pricing. Firms with high knowledge-transfer costs earn higher profits with the negotiated method, but must strive to minimize organizational conflict and the opportunity cost of managers’ time devoted to bargaining; these important implications for job design must be explored in future research. Given the widespread use of transfer pricing in practice, surprisingly little evidence exists that identifies preferred methods. Moreover, no empirical work on transfer pricing has been guided by hypotheses generated from the results and economic intuition of theoretical modeling. Our results, as well as those of Baldenius et al. (1999) and Baldenius (2000), can form a framework of testable hypotheses that can be used to guide empirical tests of the theory.