Reply To: Managing Channel Profits: Comment

We welcome Moorthy's Comment (Moorthy 1987). It provides an opportunity to clarify several issues raised in our 1983 paper. That paper was motivated by one concern—channel coordination. We believed that channel members benefit from coordination and that many institutional arrangements, such as quantity discounts, are actually coordinating mechanisms. We developed a simple model to demonstrate our belief. For historical accuracy, McGuire and Staelin (cited by Moorthy) had already independently developed a useful model of duopolistic channel competition. Being unaware of their working paper, our model not surprisingly differed from theirs. Both papers provide important but different insights into channel management. In our 1983 paper, each channel member made one decision. These decisions determined our model's four unknowns, i.e., G, g, p = G + g + C + c, and q = D(p). The process repeated until reaching equilibrium where channel profits, (p — c — C)q — f — F, were not maximized because the price was too high—a classic prisoner's dilemma. Thus, we introduced a function / which transferred channel profits, i.e., (p — c — C)q f— F, between channel members. We sought a transfer function which maximized channel profits, retained symmetry, and divided profits so all channel members were "better off". All channel members retained control over their respective decisions after enacting the transfer function. See Figure 1. With transfers, each channel member made decisions and received the transfer payment until equilibrium resulted. Any transfer function making each channel member's profit a monotone increasing function of channel profits would work. Avoiding complexity, the paper limited itself to linear functions. With linear functions, the manufacturer got k{(p c C)D(p) + k2 Fand the retailer got (1 k\){p c — C)D(p) k2 -f. Note, when k{ = 0, the manufacturer got k2 — F regardless of his decision and, when k\ = 1, the retailer got -k2 —/(here k2 < 0) regardless of his decision. We wanted each channel member to have the incentive to set their decisions optimally. Hence, we required 0 < k\ < 1. We called our transfer function a quantity discount because the function implied a declining average transfer price in quantity. Our function, which Moorthy calls "the Jeuland-Shugan Scheme", implies a transfer of k2 + CD + kx{p{D) — c — C)D from the retailer to the manufacturer. This payment has three parts: a fixed payment, k2: a fixed per-unit payment, CD; and a variable per unit payment ki{p(D) — c — C)D. Moorthy argues the first two parts of this transfer (i.e., a two-part tariff) are sufficient. This conflicts with our paper (p. 254) because we do not allow k\ = 0, as noted earlier. Although this argument appears technical, it is more. In our opinion, Moorthy's view represents a possible but disparate view of the channel. Unlike Moorthy, one of our "four key assumptions" is "symmetry" (Gould 1980). Granted, this assumption is controversial. "Symmetry" implies channel members are symmetric partners who make independent decisions and share the ensuing profits. In our opinion, "symmetry"