Mergers Efficiencies in the United States
By William J. Kolasky (U.S. Deputy Assistant Attorney General, Antitrust Division) and Andrew R. Dick (Acting Chief, Competition Policy Section, U.S. Antitrust Division):
Mergers can enable firms to secure a number of distinct types of efficiencies. The principal categories of efficiencies are: allocative, productive, dynamic, and transactional. This appendix describes and distinguishes these four efficiencies and explains why there often is a close interconnection between them in antitrust analysis. Perhaps most notably, transactional efficiency frequently serves as an essential facilitator to achieving allocative, productive, and dynamic efficiencies.
At the most general level, a market is said to achieve “allocative efficiency” when market processes lead society’s resources to be allocated to their highest valued use among all competing uses. In the context of market exchanges between consumers and producers, the allocative efficiency principle can be restated somewhat more specifically to say that the value of a product in the hands of consumers is equalized “at the margin” to the value of the resources that were used to produce that product. This intuitive “equality at the margin” condition ensures that an economy maximizes the aggregate value of all of its resources by placing them in their highest valued uses. Starting from an efficient market allocation, if a firm were to produce one additional unit of the product, the resource cost to society would exceed what consumers would be willing to pay for that last unit. Total social welfare thus would fall as a result. By the same token, if the firm cut production by one unit, the loss that consumers would suffer would exceed the value of the saved resources in whatever alternative use they were deployed. Again, total welfare would fall as a result.
Antitrust policy looks to the process of market competition as its principal means for promoting an efficient allocation of society’s scarce resources. Economic theory formalizes this principle in the First Theorem of Welfare Economics, which identifies a set of very general conditions under which a competitive market process will guarantee the efficient allocation of resources. In the long run competitive equilibrium, the market price is just equal to firms’ incremental or marginal cost. Marginal cost reflects not only directly observable costs of production, distribution and marketing but also the relevant opportunities foregone when a resource is used for one purpose rather than for some other purpose. (Hence, the term “opportunity costs” used by economists.) From society’s perspective, it represents the total cost of the resources consumed in producing, distributing, and marketing an additional unit of a particular commodity rather than employing those resources in their next best alternative use. Thus, when output is expanded to the point where price is just equal to marginal cost, the marginal value that consumers place on a good — which is the amount that they are willing to pay for the good — is just equal to the marginal value of the resources used in the good’s production. In the long run equilibrium, monopoly fails to achieve this allocative efficiency criterion established by the model of perfect competition. This follows from the fact that the monopolist’s price exceeds long run marginal cost. From society’s point of view, the marginal value placed on the good produced by the monopolist is greater than the marginal value of the resources used in the good’s production. Society therefore could be made better off if the monopolist deployed additional resources to expand output up to the point where price and marginal cost were equalized. Antitrust policy embodies this general principle by favoring competition over monopoly and (more) perfect competition over imperfect (or oligopoly) competition.
One way that merger can promote allocative efficiency arises in the context of a vertical merger to address the “double markup problem.” If a manufacturer and a distributor both enjoy some degree of market power, each firm will find it profit-maximizing to add a monopoly markup to the price that it charges. As a result, consumers will face a double markup. Understanding that it has some influence over price, the manufacturer will set a wholesale price that equates its marginal revenue to its marginal cost. Because the manufacturer faces a less than perfectly elastic demand, the wholesale price that it sets will exceed its marginal cost of production, thus producing an initial allocative inefficiency. Downstream, the distributor will treat the wholesale price as its relevant marginal cost of business. Also enjoying market power, the distributor will set a retail price above its marginal cost, resulting in a second allocative inefficiency. Note, however, that the distortion caused by the second markup is compounded because it is applied to an already supra-competitive wholesale price. Contrast this with the case of an integrated manufacturer-distributor. The integrated firm will “charge itself” only the actual marginal cost of producing the good and will extract its market power only at the stage of selling to the final consumer. Consumers facing the double markup will buy less than when there is an integrated manufacturer-distributor. As a result, they are worse off.
Collectively, the manufacturer and wholesaler also earn less profit than they would if they were integrated. This foregone profit provides a strong incentive for the firms to merge to promote allocative efficiency and thereby increase their joint profits. If the integrated firm produces as efficiently as the separate firms, then integration makes both producers and consumers better off. Even if the integrated firm is somewhat less efficient than its constituent parts, the desirable effect of eliminating one of the markups may outweigh this negative effect. If a merger is impractical or barred for other reasons, a variety of vertical contracts may offer alternative means to mitigate allocative inefficiencies from the double markup. Vertical contracts can be structured by the manufacturer to induce its distributor not to restrict input further and thereby (at a fixed wholesale price) cut further into the manufacturer’s own margin. Examples of vertical contracts that can promote this objective are maximum resale price maintenance, quantity forcing (placing a minimum sales quota on the distributor), and two-part pricing that sets the wholesale price equal to the manufacturer’s marginal cost of production and then charges a lump sum franchise fee.
A second efficiency concept is productive efficiency. Production is said to be efficient when all goods are produced at the minimum possible total cost. An equivalent way of phrasing the productive efficiency criterion is to say that there is no possible rearrangement or alternative organization of resources (such as labor, raw materials, and machinery) that could increase the output of one product without necessarily forcing a reduction in output for at least one other product. This restatement highlights the principle that firms’ choices involve explicit trade-offs between competing demands for scarce resources.
Mergers (as well as joint ventures and other cooperative practices) hold the potential to increase productive efficiency in a number of ways, including by fostering economies of scale, economies of scope, and synergies. The first way that mergers can increase productive efficiency is to move firms closer to the optimal scale of production for their industry. Ascertaining the optimal scale for a firm can be done using a number of types of information, including comparisons of actual production costs for firms of different sizes, engineering estimates of probable production costs for enterprises of varying sizes, and comparisons of rates of return on investment. George Stigler pioneered a much simpler and economically more intuitive method, however, which he coined the “survivor principle.”(182) Stigler’s survivor principle is based on the simple intuition that active competition among firms for scare resources — both within an industry and across industries — inevitably will drive firms towards the optimal or efficient scale of operations. Under competition, inefficiently scaled firms will be driven from the market either by exit or by acquisition. Mergers play a very important role in this competitive process by reorganizing the ownership and use of economic resources among firms to achieve efficient productive scale. Combining the operations of two firms may reduce duplication, allow fixed expenditures to be spread across a larger base of output, permit firms to reorganize production lines across plant facilities to achieve longer production runs and reduce switch over costs, lower inventory holding costs, and more finely specialize the use of resources such as skilled labor. Each of these merger rationales can facilitate firms’ efforts to reach an efficient scale.
Some economists and antitrust practicioners argue that antitrust agencies should, as a general practice, be skeptical of treating achievement of economies of scale as a merger-specific efficiency.(183) According to this view, firms generally can reach their efficient scale of production by purchasing additional inputs through market transactions or developing them internally (if the firm is sub-optimally small) or by shedding surplus inputs or machinery in secondary markets (if the firm is sub-optimally large). Because these unilateral changes in firm scale do not necessarily induce the exit of a direct competitor, they are sometimes thought to offer the same gains in productive efficiency without the risk of diminished competition attendant to a merger.
There are a number of practical reasons, however, why internal expansion (or contraction) sometimes may be a significantly costlier means than merger to increase a firm’s productive efficiency.(184) First, mergers may hasten the speed with which firms can expand their scale to exploit economies of large scale production. Mergers may provide the acquiring firm with ready access to existing inventories or supply contracts for important inputs as well as access to additional plant capacity that can quickly be brought on-line. Second, adding new capacity in a market with static or declining demand may place sufficient downward pressure on price to make internal expansion unprofitable. In this situation, neither of the merging firms might be likely to expand its scale in the near future absent the merger. Third, the construction of new capacity may create social waste if duplicate resources at the acquired firm eventually wind up being scrapped when they are removed from competition rather than being merged into a single firm. When any of these conditions is present, mergers may be a privately or socially less costly means to reap economies of scale and enhance firms’ productive efficiency.
A second way that mergers can increase productive efficiency is to enable firms to exploit economies of scope. Economies of scope are said to exist when it is cheaper to produce two or more products together rather than separately.(185) Economies of scope can be quite substantial. For example, one study of the economies of scope achieved by General Motors from combining its production of large cars with small car and truck production estimated that the firm saves 25% in total operating costs relative to splitting the two operations.(186) There are many potential sources of economies of scope. One of the most common is the use of common raw inputs. For example, it is commonsensical that book publishers exploit economies of scope by producing both hardcover and soft-cover editions from the same manuscript, and that automobile companies exploit economies of scope by producing multiple car models that use many of the same input components. Another important factor contributing to economies of scope is technical knowledge about producing and selling related products. Information about one product may be directly relevant for other closely related products. For example, knowledge about how to market steel bars efficiently (such as knowing where customers are located and their purchase habits) could assist the firm in marketing steel sheets. Similarly, knowledge about the techniques to manufacture steel bars efficiently (such as knowing how to operate blast furnaces and where to obtain a reliable supply of pig iron) could make the manufacture of steel sheets more efficient. In these situations, it will tend to be more efficient for a single firm to produce and market both steel sheets and steel bars.
The principle of economies of scope, by itself, however, does not necessarily imply that the products should be produced by a single firm. In theory, economies of scope might be exploited by locating the related production lines sufficiently close to one another to facilitate exchange between separate firms. In practice, however, exploiting economies of scope frequently hinges on achieving transactional efficiencies made possible by having the related production lines brought under common management. Merger is one way to achieve this important nexus between productive and transactional efficiencies. To make this point more tangible, consider steel manufacturing as an example. Iron ore is first melted down into pig iron in a blast furnace; the molten pig iron is then processed in a steel-making furnace and turned into slabs or sheets of steel. It is conceivable that two separate firms, side by side, could specialize with one making pig iron and the other making steel, while a pipe would carry the molten pig iron between the two firms. These firms would be highly reliant upon one another, however, and the risk that either firm could exploit or “hold up” the other would introduce substantial transaction inefficiencies.(187) High transaction costs frequently explain why a firm typically will bring in-house all of the products for which substantial economies of scope exist.
A second illustration of how the achievement of productive efficiencies can hinge on achievement of transactional efficiencies is given by the example of economies of scope flowing from common production or marketing knowledge. In principle, knowledge could be bought and sold in the market, thus avoiding the necessity to house the production or marketing of (say) steel bars and steel sheets under the same corporate roof. In practice, however, market transactions of information can be highly costly, inefficient, and subject to opportunism. This observation may explain why a single firm often produces closely related products, and it identifies another important potential source of efficiencies from mergers.
A third way by which mergers can increase productive efficiency relates to synergies. Synergies are defined as cost savings (or quality improvements) that flow from the close or intimate integration of specific, hard-to-trade assets. Joe Farrell and Carl Shapiro have identified several examples of synergistic efficiencies.(188) One involves efforts to improve interoperability between complementary products. Suppose that one firm produces word processing software that is easy to use but has very limited graphics capabilities, while another firm produces a desktop publishing program that is powerful but difficult to use. Many consumers elect to use the word processor to quickly prepare text files which they then cut and paste into the desktop publisher for formatting. Differences in the programs’ file formats and other incompatibilities, however, make this a second-best solution for consumers. By merging their operations, the two firms could synergistically improve the interoperability of their products by developing a seamless interface between the text and publishing software modules. A second source of synergies involves the sharing of complementary skills. One firm may have developed and perfected a superior approach to manufacturing a product while a rival may have built an extensive and well-organized distribution network. Some form of cooperation — whether a merger, joint venture, or licensing agreement — could allow the two firms to synergistically integrate their respective manufacturing and distribution skills to produce and sell their product more cheaply.
A third efficiency concept is dynamic efficiency, which concerns itself with market processes that encourage innovation to lower costs and develop new and improved products. Whereas allocative and productive efficiency can be viewed as static criteria — holding society’s technological know-how constant– a more dynamic view of efficiency examines the conditions under which technological know-how and the set of feasible products optimally can be expanded over time through means such as learning by doing, research and development, and entrepreneurial creativity. Static efficiency principles favor market equilibria characterized by short run cost minimization and zero profit conditions. The dynamic efficiency principle, most closely associated with Austrian economist Joseph Schumpeter,(189) instead suggests that the short run costs associated with allocative and productive inefficiencies stemming from market power can more than be offset by benefits from encouraging dynamic efficiencies through “creative destruction.”
Schumpeter disputed the traditional view that perfect competition spurs invention while monopoly retards it. Schumpeter stressed the advantages enjoyed by larger firms to finance substantial research and development activities and to appropriate the benefits from their investment and learning across a larger scale of operations. At the same time, Schumpeter did not think that the comparative advantage of large firms in innovation would provide them with a secure or impregnable position in the market. Schumpeter believed that innovation was a continuous process and that no single firm would gain more than a transitory monopoly from invention in the face of a constant supply of new ideas and innovations from its other large rivals. This continual competition would prevent markets from departing too far from the benchmarks of short run allocative and productive efficiency, while the pursuit of temporary monopoly positions would encourage firms to expand technological frontiers and push out new product boundaries that would allow society to achieve in the long run still greater allocative and productive efficiencies.
Embracing a Schumpeterian view of competition, economists Gary Roberts and Steve Salop have argued in favor of applying a dynamic framework for assessing claimed merger efficiencies.(190) According to Roberts and Salop,
Efficiency improvements are not static, one-time-only events. Rather, they occur as part of a rich dynamic process in which efficiency improvements are introduced for private gain but then frequently stimulate competition that creates significant spill-over benefits for consumers. Mergers can speed the pace of technical progress and reduce prices by facilitating innovations that initiate technological diffusion and induce competitive innovations.(191)
Roberts and Salop have elaborated on the link between dynamic efficiency and competition:
The dynamic framework provides a far more realistic account of the manner in which merger efficiencies increase competition. In particular, the dynamic framework recognizes that cost savings achieved by a newly merged entity generally will diffuse at least partially to competing firms over time. As this diffusion occurs, the aggregate cost savings multiply. The diffusion also should enhance competition and increase the likelihood that firms will improve consumer welfare by passing the cost savings on to consumers in the form of lower prices.(192)
Like allocative and productive efficiencies, achievement of dynamic efficiencies can be facilitated by antitrust and other public policies that permit efficient transactions in support of invention. To illustrate, dynamic efficiencies require the establishment of an incentive system to allow inventors to appropriate returns sufficient to make the inventive activity worthwhile. Establishing and protecting ownership rights to the fruits of inventive activity is thus essential. Harold Demsetz has pointed out that “the problem of defining ownership is precisely that of creating properly scaled legal barriers to entry.”(193) Patent protection provides one type of scaled barrier that balances the appropriability of inventions to generate necessary returns to firms against the speed of diffusion of the benefits that consumers derive from invention. Likewise, antitrust policy seeks to determine appropriately scaled entry barriers, for example, by governing the conditions under which inventors can use non-compete provisions to restrain licensees from competing against them, or by assessing the circumstances under which research joint ventures that restrict competition among actual or potential rivals may be necessary to generate dynamic efficiencies.
The fourth and final category of efficiencies is labeled transactional efficiency. It is the broadest category of efficiencies, and as alluded to earlier, it frequently facilitates firms’ efforts to achieve allocative, productive, and dynamic efficiencies. The basic insight offered by the school of thought known as “transaction cost economics” is that market participants design business practices, contracts, and organizational forms to minimize transaction costs and, in particular, to mitigate information costs and reduce their exposure to opportunistic behavior or “hold-ups.”(194)
Oliver Williamson has argued that the critical dimensions of transactions are uncertainty, the frequency of recurrence, and the extent to which participants in market exchange make investments in transaction-specific assets.(195) Asset specificity creates the ability and the incentive for parties to engage in opportunistic behavior. Many business relationships require that one or both parties invest in an asset that is highly specialized to their transaction. An example of a transaction-specific investment would be the construction of a pipeline connecting an oil refinery to an isolated distribution terminal. Because the value of the asset is much higher in its intended use than in its next best alternative use, the parties are locked into their relationship to a significant degree. Neither buyers nor sellers can turn to alternative partners without incurring a substantial loss. By the same token, however, each party can take advantage of the other by attempting to obtain more favorable terms than had initially been bargained. Buyers can refuse to purchase unless the price is reduced, while sellers can refuse to deliver unless the price is increased. As a result, the value of the specialized asset over and above its next best alternative use can be appropriated by opportunistic behavior or hold-ups executed by one or both parties to the transaction.
The frequency that transactions recur also guides the selection of institutional arrangements for governing interactions between market participants. When transactions take place only infrequently, explicit contracts or close integration between companies will usually be unnecessary except in the presence of highly specialized assets. If transacting parties expect that they will maintain a continuing relationship, however, they may rely on implicit or explicit mechanisms such as long term contracts, performance bonds, and reputational sanctions to protect their returns from investments made in physical or human capital specialized to their transaction.
Finally, uncertainty or incomplete information about how the value of resources in their alternative uses may change over time affects how transactions can be efficiently structured. Information is incomplete for the simple reason that it is not costless to generate and communicate. Rational consumers and producers will invest in becoming informed only up until the point where the marginal cost of information equals its marginal value. Because the marginal cost remains positive, it follows that the marginal benefit of information also is positive and hence rational economic actors remain incompletely informed. A corollary of this principle is that, in general, it will not pay market participants to fully insure themselves against risk by designing a complete set of contingent contracts. Instead, market participants will often rely on other methods such as those mentioned earlier, including reputation, repeat dealing, structured incentives, performance bonds, and third party (court) oversight in order to protect their specific investments.
Given uncertainty, the existence of transaction-specific investments, and varying frequencies of market interactions, parties will design contracts, create joint ventures, or propose mergers to minimize these transactions costs for any given level of economic activity. The pursuit of transactional efficiency explains why firms choose to consolidate some activities under common management and direction while leaving other activities to market-based transactions. Applying the concept of allocative efficiency to transactions, economic Nobel laureate Ronald Coase offered an early theory of merger activity when he wrote that “a firm will tend to expand until the costs of organizing an extra transaction within the firm become equal to the costs of carrying out the same transaction by means of an exchange on the open market or the costs of organising another firm.”(196) Coase’s simple yet powerful insight helps us understand why, as mentioned earlier, we frequently observe goods whose production exhibits economies of scope being produced by a merged firm rather than having firms attempt to capture scope economies through market transactions. The risk of opportunistic behavior in this setting raises the cost of market transactions relative to within-firm organization.
Transactional efficiency also helps explain a variety of other business practices and market structures. For example, firms that wish to cooperate on research projects may choose to form a joint venture — or in the limit, merger — rather than rely on arms-length transactions. Joint ventures and common ownership can help align firms’ incentives and discourage shirking, free riding, and opportunistic behavior that can be very costly and difficult to police using arms-length transactions. The pursuit of transactional efficiency also can help explain why firms may adopt various vertical contracts such as exclusive territories and resale price maintenance to help mitigate free riding and principal-agent costs.(197) Lastly, the concept of transactional efficiency has been applied to analyze the market for corporate control in which the threat of hostile takeovers can lessen shareholders’ costs of transacting with professional managers to ensure that they act in the interest of the company’s shareholders.(198)
Notes and References
182. George J. Stigler, The Economies of Scale, 1 J. Law & Econ. 54 (1968).
183. Joseph Farrell and Carl Shapiro, Scale Economies and Synergies in Horizontal Merger Analysis, 68 Antitrust Law J. 685 (2001).
184. William J. Kolasky, “Lessons from Baby Food: The Role of Efficiencies in Merger Review,” Antitrust, Fall 2001, at 82-87.
185. John C. Panzar and Robert D. Willig, Economies of Scale in Multi-Output Production, 91 Quarterly J. Econ.. 481 (1977).
186. Ann F. Friedlaender, Clifford Winston, and Kung Wang, Costs, Technology, and Productivity in the U.S. Automobile Industry, 14 Bell Journal of Economics 1 (1983).
187. The classic discussion of this problem is Benjamin Klein, Robert G. Crawford, and Armen A. Alchian, Vertical Integration, Appropriable Rents and the Competitive Contracting Process, 21 J. Law & Econ 297 (1978).
188. Farrell and Carl Shapiro, op. cit.
189. Joseph A. Schumpeter, Capitalism, Socialism, and Democracy (1950).
190. Gary L. Roberts and Steven C. Salop, Efficiencies in Dynamic Merger Analysis, 19 Offprints of the World Competition 5 (1996).
191. Id. at 7-8.
192. Id. at 7.
193. Harold Demsetz, Barriers to Entry, 72 Am. Econ. Rev. 47, 49 (1982).
194. For a recent application of transaction cost principles to antitrust rules and analysis, see Paul L. Joskow, Transaction Cost Economics, Antitrust Rules, and Remedies, 18 J. Law, Econ., & Org. 95 (2002). A helpful exposition of the meanings and sources of transaction costs appears in Douglas W. Allen, What Are Transaction Costs?, 14 Res. in Law & Econ. 1 (1991).
195. Oliver E. Williamson, Transaction-Cost Economics: The Governance of Contractual Relations, 22 J. Law & Econ. 223 (1979).
196. Ronald H. Coase, The Nature of the Firm, 4 Economica 386, 395 (November 1937).
197. See, for example, Benjamin Klein and Kevin M. Murphy, Vertical Restraints as Contract Enforcement Mechanisms, 31 J. Law & Econ. 265 (1988) and G. Frank Mathewson and Ralph A. Winter, An Economic Theory of Vertical Restraints, 15 Rand J. Econ. 27 (1984).
198. See, for example, Eugene F. Fama and Michael C. Jensen, Agency Problems and Residual Claims, 26 J. Law & Econ. 327 (1983).