Peter Castensen
May 7, 2008
Anticompetitive Mergers and Conduct in Agricultural Markets: A Major Failure of Antitrust Enforcement Policy
Statement prepared for the hearing on
"Agricultural Consolidation and Other Related Issues"
held by the
Senate Judiciary Subcommittee on Antitrust, Competition Policy and Consumer Rights,
May 7, 2008.
By
Peter C. Carstensen
George H. Young-Bascom Professor of Law
University of Wisconsin Law School
975 Bascom Mall
Madison, WI 53706
pccarste@wisc.edu
Introduction
I am honored to have been asked to offer my views on the state of antitrust enforcement in the markets related to agriculture. In nut shell, the government agencies charged with enforcing antitrust law have repeatedly failed to challenge or to remedy the competitive problems that confront American agriculture. The most conspicuous failure has come in merger enforcement where a series of decisions either not to challenge mergers or settle for weak, even anticompetitive, remedies has resulted in increased concentration on both the input and output side of agriculture. In addition and equally troubling, the enforcement agencies have failed to undertake challenges to well documented, anticompetitive conduct affecting farmers and ranchers. Given this pattern of failure, it is time for Congress to give serious consideration to creating alternative means of enforcing the commands of antitrust law.
Over the last decade, I have been particularly interested in issues involving competition in agricultural markets. In 2000, I published an article in the Wisconsin Law Review: Concentration and the Destruction of Competition in Agricultural Markets: The Case for Change in Public Policy, 2000 WIS L. REV. 531. A central thesis of that article was that there are serious problems of market failure in agriculture directly related to the high and increasing levels of concentration in the industries buying from farmers and ranchers. I urged increased antitrust enforcement and also suggested legislative action in addition to antitrust enforcement was essential to restoring competition in agricultural markets. The goal of legislation should be to facilitate the operation of a dynamic market process that is efficient, transparent, open, and fair. Since then, I co-authored with Professors Neil Harl and Roger McEowan an article in defense of the packer ownership ban (The 2002 Senate Farm Bill: The Ban On Packer Ownership of Livestock, 7 DRAKE J. AG. L. 267 (2002)). I have also written about the exploitation of farmers by seed companies that have patents or licensed their patents covering some of the genetic material included in seeds (Post-Sale Restraints via Patent Licensing: A "Seedcentric" Perspective, 16 FORDHAM INTEL PROP. MEDIA & ENTERTAINMENT L. J. 1053 (2006)).
I also have a background in some aspects of these issues. As a government lawyer some 35 years ago, I reviewed the old meat packing consent decree and in the process came to appreciate the context within which Congress crafted the Packers and Stockyards Act. In 1995, I served in Wisconsin on a committee that reviewed and proposed modifications for the regulations governing contracts for vegetables being purchased for canning. I have also done an extensive examination of the grain marketing industry in connection with a study of the famous Chicago Board of Trade decision which is a landmark antitrust case. In addition, my work on the competitive implications of other kinds of vertical distribution arrangements has provided me with relevant background on some of the key issues being considered today. In September 2001, I was one of six invited academic experts in the U.S. Department of Agriculture's Public Forum on Captive Supplies held in Denver, Colorado. I have also been an invited witness at a number of hearings held by this committee and by the Senate Agriculture, Nutrition, and Forestry focused on the issues of agricultural competition.
Overview
Farmers are poorly served by existing market structures and practices. Farmers and ranchers today confront excessive concentration in most of the industries buying and processing agricultural products including those in meat, grain and dairy. The existence of concentrated markets creates the incentive and the capacity for such firms to engage in conduct aimed at exploiting those participants with limited options and to entrench existing market power against the threat of deconcentrating and effective competition.
Free and open markets are generally the best institutional structure for achieving all the important goals of economic policy: efficiency, dynamic growth, equitable allocation of resources, opportunity for all participants. Where markets are unconcentrated with many buyers and sellers, there is a strong tendency for efficient, workable and fair methods to develop as the inevitable outcome of the interaction of many participants all seeking a neutral and open market place.
But no such inherent tendency exists in markets where there is a substantial difference in size between buyers and sellers and the market is also highly concentrated, i.e., there are few firms altogether on one side. Also, if one side has significant and persistent advantages in information or some other important element related to the transactions between buyer and seller, then too such a market is unlikely to experience much pressure for desirable conditions. There is a grave danger that strategic conduct will shape such markets frustrating the goals of an efficient, open, fair and accessible marketplace. This in turn imposes immediate burdens on the disfavored class of participants and ultimately on consumers and the economy as a whole as less efficient production and market transactions take place.
When markets lack the inherent tendencies to create desirable conditions, the law can play a vital role in defining rules for the participants that reduce their capacity to engage in strategic conduct and restore greater balance among the participants. The statute books contain many such laws including ones regulating credit, insurance, product safety, job safety, franchising of various kinds (e.g., gas stations, fast food, automobile dealerships), energy markets and, of course, securities markets.
America's farmers and ranchers are caught in an economic vise. When they seek to buy the various inputs that they need--seed, fertilizer, equipment, herbicides, etc.--they face increasingly concentrated markets and exploitive strategies of producers. When they attempt to sell their products, especially dairy and meat products, they have only a very limited number of buyers who use their buyer power to drive down the prices paid for these products. Despite the recent dramatic increases in the price of grains, those commodity markets are also highly concentrated on the buying side. This bodes ill for the long run ability of farmers and ranchers to receive the full benefit of the market prices that such commodities receive.
The markets for agricultural commodities provide a textbook illustration of how law and regulations can either facilitate or frustrate the accomplishment of the goal of an efficient, transparent, and equitable market context. Antitrust law enforcement over the past decade has failed to deal effectively with either the substantial structural changes or the exploitive and exclusionary conduct manifest on both sides of the production of agricultural products. In addition, the U. S. Department of Agriculture (USDA) has substantial authority to adopt and enforce rules that could ameliorate some of the most serious problems of access, information disclosure, and exploitation. But it has consistently failed to use its authority to facilitate efficient market practices.
The consequence of the combined failure to enforce antitrust law and to fashion relevant market regulations is that farmers and ranchers were and are under-compensated for their production. But at the same time, consumers are paying higher and higher prices for food products because the bottlenecks in the process of moving food from farm to market and on to the consumer have allowed both processors and retailers to exploit both producers and consumers.
Another important general point that informs much of the following discussion is that buyer power needs to be measured with different metrics and its competitive implications understood in terms of the buying side of the market. Contemporary antitrust enforcers have largely failed to appreciate these facts. As a result, the analysis of mergers creating buyer power has been consistently deficient. In addition, the agencies have failed to take necessary action to control anticompetitive conduct by dominant firms in agricultural markets.
In the following discussion I will review the last decade of inactivity in four major agricultural market areas (pork, dairy, commercial crops, and beef) as well as in grocery retailing. The consistent observation is that the enforcement agencies, basically the Antitrust Division of the Department of Justice although the FTC plays an important role with respect to mergers in the grocery store aspect of the market, have failed to perform appropriate analysis of the competitive effects of these mergers and conduct. As a result, markets are worse off today than they were 10 years ago. This has harmed both producers and the ultimate consumers. I will end my discussion with some reflections on the directions that public policy ought to take to remedy this situation.
Pork
One of the most instructive areas to examine is the market for hogs. In 2003, Smithfield acquired Farmland's pork processing facilities and in 2007, it acquired Premium Standard Brands (PSB). The PSB merger consolidated the only two major processors serving the Southeastern United States. The next closest major facility is in Kentucky about 400 miles away. My interviews with agricultural economists who had studied the industry told me that producers faced very substantial costs if they wanted to take their mature hogs to that more distant processor because it is costly to haul mature hogs that long a distance. As a result, hogs in the Southeast were often priced as much as 10% below the price paid for comparable hogs in the Midwest even before the merger.
My investigation received strong confirmation from the RTI study of the pork processing industry that GIPSA sponsored. That study which focused on the period 2002 to 2005 found that there was statistically significant buyer power in the market for mature hogs. This is very significant because this finding antedated the acquisition of PSB. This finding can then be compared with market structure in the pork industry. During the period when buyer power was found to exist, national concentration rose from an HHI of 1042 in 2001 to an HHI of 1334 in 2005 (i.e., before taking account of the Smithfield PSB merger).
What is significant here is that standard buyer side analysis would be likely to conclude that increases in concentration in that range would be unlikely to cause an adverse effect on competition. But here in fact we have direct, sophisticated econometric evidence that lower levels of concentration are sufficient to create buyer power.
Despite knowing that the PSB merger would in fact increase buyer power with demonstrable adverse effect on producers, the Antitrust Division failed to act. The Division claimed hog raisers in the Southeast would not be exploited because they could transport their hogs to other processors, but the closest facility that appears to exist is approximately 400 miles away in Kentucky. Shipping mature hogs that distance is costly and would be rational only if the price offered by the monopoly processor were very deeply below the price offered by the distant buyer. In addition, the Division claimed that farmers providing contract services in the Southeast could somehow switch to providing those services to "independent producers who own their own hog operations in the area." But given a monopoly buyer, these "producers" face the same problem of depressed prices that other smaller producers would confront.
Not only was this decision wrong, but it demonstrates a major failure to understand both the dimensions of the markets for mature hogs and to appreciate that buyer power in fact occurs at lower levels of concentration than the Division associates with seller power. The PSB merger was one that not only "may substantially lessen competition or tend to create monopoly", it was one that did substantially lessen competition and tended to create a monopoly in pork processing in the Southeast.
Dairy
In 2001, the Antitrust Division allowed Suiza to acquire Dean. This combination created the largest fluid milk processor in the count with a market share in excess of 30%. The Division approved this merger without formal objection, but its review lasted many months and involved a substantial revision of the proposed deal. Basically, the new Dean agreed to divest a significantly larger number of milk processing facilities than it had originally proposed. In addition, the press release announcing approval implied that the new firm would not enter into a long-term exclusive dealing contract with Dairy Farmers of America (DFA), the largest dairy cooperative. However, Dean and DFA quickly found a way around that commitment. In addition, Dean refused to deal with independent milk producers who had traditionally been direct suppliers. Even these high-volume, high quality producers were forced to submit to DFA.
National Dairy Holdings (NDH) purchased the divested facilities. DFA, however, was a substantial shareholder in NDH and obtained an exclusive supply contract. The third major milk processor is Hood. Through various means, Hood and NDH, with the blessing of the Antitrust Division, have managed to combine their managements. Only vigorous protests from a few cooperatives have kept Hood from completely embracing the DFA exclusive dealing arrangements.
The divestitures should have provided a means to retain competition in both the buying and sale of fluid milk. The limits on exclusive dealing were also important because DFA was using its control over access to the Suize and Dean processing facilities to coerce other cooperatives into merging with it or putting themselves under its control. Most dairy farms produce Grade A milk suitable for use as fluid milk, but in fact the bulk of that milk is used for other purposes such as making cheese or ice cream. However, to share in the premium paid for milk used as fluid milk, a farmer's milk must be delivered to a fluid milk processor some percentage of the time. In practice, however, most farmers belong to cooperatives or other buying groups and it is the group that must make delivery of some percentage of its milk for some period to time in order for all the members of the group to qualify for participation in the higher price milk pool.
DFA through its ownership links to NDH and Hood and its exclusive dealing arrangement with Dean (and other milk processors) has control of access to fluid milk sales in many parts of the country. The result is a serious problem of access for those dairy farmers that do not want to be part of DFA. In addition, it appears that DFA has engaged in various discriminatory and preferential agreements with the result that many dairy farmers are getting less for their milk than they received when the buying side was more competitive.
The Antitrust Division has an open investigation of the conduct of the milk industry. But the matter has been pending for years without any action. In summary, then, the combined failure of the Antitrust Division to take firm action against consolidation of processors and to challenge the panoply of anticompetitive practices rife in the industry has resulted in serious losses of income and coercion of farmers.
In addition, as Professor Ron Cotterill of the University of Connecticut has documented, the increased concentration in both processing and retailing have resulted in an increasing price spread between what farmers receive for milk and what consumers pay for. Thus, the failure of antitrust enforcement in dairy has resulted in harm to both producers and consumers.
Commercial Crops--Corn, Grain, Soybeans and Cotton
In 1999, the Antitrust Division allowed Cargill to acquire Continental's grain operations subject to some modest divestiture. The immediate result was to increase concentration in the business of buying grain such as corn, wheat and soybeans. Farmers found themselves with less competition at the farm gate for their crops. As in the case of pork, the levels of concentration that resulted are such that there was a significant increase risk of buyer power. Despite the recent increases in the prices for most grains, the point here is that the industry structure facilitates the intermediaries such as ADM and Cargill to extract much of the gain that ought to come to the farmer.
A related problem is the consolidation of the seed industry. There have been a large number of mergers and acquisitions in the industry over the last decade. The result has been a reduction in competition and an increase in the price of seeds. The most recent combination is the most egregious. The Antitrust Division has allowed Monsanto, the leading owner of patents on genetic modifications for seeds, to acquire Delta Pine & Land (DPL), the dominant producer to cotton seeds. DPL does not itself engage in developing genetic modifications, but it is an obviously major partner for such development. As long as it remained independent, it had a strong incentive to support competing lines of development. This both increased the potential for major improvements in genetically modified seed and ensured that competing technologies could be developed. Competition in technology in turn would result in lower costs to farmers and fewer restrictions.
The previous administration had, to its credit, rejected this merger because of these risks. However, recently, the Antitrust Division approved the merger subject to complex regulatory decree that has the effect of limiting technological competition. The decree effectively authorized two firms to continue to compete with Monsanto while excluding a third major developer. The very complexity of the decree demonstrates the anticompetitive potential that this merger creates. Indeed, the number of objections to the settlement filed with the trial court as part of the Tunny Act review is unprecedented. Among the objectors are 13 state attorney generals.
Interestingly, the Division also required that Monsanto allow its licensees in cotton seed production to stack Monsanto's genetics with genetics from other sources. This reflects recognition that Monsanto has and can use such restraints to foreclose competition. Unfortunately, the government only imposed this limit with respect to cotton seed despite the fact that Monsanto apparently uses the same anticompetitive restraints in all its seed contracts. Thus, despite recognizing the anticompetitive nature of the restraints, the government has failed to act to protect corn, soybean and other farmers from exploitation.
An additional problem with the current methods of marketing seeds containing patented genetics is the use of post-sale restraints on the use of the seed. In the case of cotton and soybeans, the restraints Monsanto imposes prohibit saving and replanting the seed. This forces farmers to buy new seed each year at inflated prices because the seed companies do not face competition from saved seed. The argument for the restraint is that Monsanto is entitled to a license fee for such use. That claim, assuming its validity, does not justify a practice that protects from competition the un-patented components in a seed. The alternative approach, used by seed companies elsewhere in the world, is to collect a fee from farmers who save and replant seed subject to intellectual property rights. Unfortunately, the Antitrust Division although fully aware of the anticompetitive effects of Monsanto's policy has failed to challenge it and in fact seems to have excused it as lawful in a brief of the Supreme Court. Subsequently the government has concluded that patent law does not authorize and immunize at least some post-sale restraints.
Beef
The next panel will discuss the competitive effects of the proposed Swift-National-Smithfield merger in detail. The beef packing industry has not seen many anticompetitive acquisitions in the last 25 years. In fact, Smithfield's entry a few years ago with geographically dispersed foothold type acquisitions was a clear plus for competition and may have moved the industry toward some modest deconcentration. In addition, the beef feeding operations that Smithfield also owns are not vertically integrated into its packing houses. Hence, it has had a strong interest in the retention of a viable market for fed cattle. The same would be true of any other owner of those feeding operations if it was not vertically integrated. Of course, such an owner is a potential de novo entrant into the slaughter market in the region near the feeding operation. As long as such a firm stood in the wings, it would put pressure on existing firms to be more competitive. If actual entry occurred, it would stimulate a more competitive market for beef because it would increase the number of competitive buyers in the market. In fact, shortly before deciding to sell out, Smithfield had been in the process of making entry. It halted that plan and thereafter proposed to sell its entire operation to JBS.
I have two observations related to overall antitrust enforcement. First, the existing, pre-merger, level of concentration in the beef packing is substantially greater than in pork processing. In pork we know that buyer power exists. It follows that buyer power already exists in the beef processing market. Moreover, the proposed merger will substantially increase that concentration and create the kind of vertical integration that will make manipulation of the cash market even more possible. Hence, this merger "may substantially lessen competition" in the words of the Clayton Act's prohibition on anticompetitive mergers.
Secondly, in September 2001, a group of experts including myself evaluated the competitive implications of the use of long term contracts by packers to secure cattle rather than the cash market. We disagreed about the competitive implications of this practice overall. But we agreed that no packer should be allowed to use its current cash price at the plant receiving contract cattle as the basis for the contract price. The incentive to manipulate cash prices is obvious, but the more subtle harm is that the buyer for such a plant can not raise the cash price even to get a good pen of cattle because the effect is to raise the price of all cattle coming to that plant that week. Thus the contract suppliers have the benefit of a "most favored nation" system and are assured that they will get the same or better price than the cash price. But this distorts buying practices and harms the cattle feeding business by restricting the flexibility of buyers in the cash market. Moreover, there are a number of alternative bases for pricing contract cattle that would significantly reduce the incentive to manipulate the cash price. Hence forbidding this practice would not undermine whatever efficiencies contract systems might produce.
Despite our expert consensus the USDA has failed to adopt even this simple regulation. Moreover, the Antitrust Division is aware of this practice and other market manipulating practices including collective misstatements about future cash purchase plans and joint withdrawals from buying in the cash market. Yet the Division so far I can tell has not even conducted a through investigation of these anticompetitive practices.
The beef market illustrates forcefully the failure of the USDA as the agency charged with responsibility, like the SEC or CFTC, to ensure a fair, efficient and open market regulator to carry out its responsibilities. But it also shows that the Antitrust Division aware of the regulatory context that facilitates collusive market manipulation and exploitation of producers has failed to take any action to enforce the antitrust law prohibitions on this conduct.
Grocery Consolidation
Another factor that merits brief reference because it is very important in the overall evolution of buyer power in agricultural markets is the increased concentration in the grocery business. As that business becomes more concentrated the retailers acquire greater buyer power and use that power in ways that cause adverse effects on upstream markets. Indeed, one continually reads as a justification for mergers among food processors that they need to combine in order to have bargaining power with retailers. The other side of that power is an increase in their own buying power. They use that power to drive down the prices they pay even as they try to keep up prices with respect to what retailers pay for their goods.
The FTC is largely responsible for enforcing antitrust law in the grocery business. It has failed to take appropriate account of the creation and entrenchment of buyer power in its reviews of such mergers. Although this may seem only indirectly to affect farmers, the reality is that upstream power is reflected back onto the suppliers least able to transfer the impact further. A major error, therefore, in the analysis of buyer power in merger cases is the failure to look for the places where the exploitation of such power will come to rest. If such a focus had been used, the FTC and Antitrust Division would have observed that there is a more substantial risk of adverse effects on competition from mergers creating buyer power.
The Implications of Consistent Failure to Enforce the Law
The history of the last decade of antitrust enforcement related to agricultural markets is sad. As the foregoing summary shows, the results have been a substantial increase in concentration that has and will result in exploitation of farmers and a failure to challenge any of the anticompetitive practices that these firms employ. There is also an equally disturbing failure on the part of the USDA to use its substantial existing authority to protect and promote fair and open market access. It is not surprising that the latest farm bill may impose more direct legislative commands on these markets.
In the late 1990s, the Antitrust Division created Mr. Ross's position. The hope was that this would provide better engagement with agricultural issues. It is clear after a decade that the position carries no authority to oversee the initiation and filing of cases. At best, Mr. Ross provides a phone number to which complaints can be directed. But he has no authority to do anything about those complaints. He is a dedicated civil servant in a difficult position of having to justify and defend decisions over which he has no control. What this tells us is that institutionally the Antitrust Division has not been able to provide the kind of oversight of anticompetitive conduct and mergers in agriculture that Congress expects.
What can be done? First, hearings like this have some utility because they provide to the actual decision makers at the Antitrust Division the signal that there is significant unhappiness with the policy of inaction that they are pursuing. This role can be augmented if the Committee will insist on confidential briefings by the Division about its decisions. Such a strategy will avoid the disputes about confidentiality that arise when it is asked to justify its decisions in public. At the same time, such briefings will make the Committee and its staff much better informed about the merits of the analysis and decisions of the Antitrust Division in agricultural matters. Another alternative would be to have the Government Accountability Office conduct a full review of the key decisions about agricultural mergers in the last 10 year with a focus on the consequences of the actions and inactions of the Antitrust Division.
Second, I would suggest it is time to change the institutional arrangement for enforcing competition law in agriculture. The Grassley-Kohl bill, S 1759, proposes changes in the process that would directly address the institutional failures of both the USDA and the Antitrust Division. Not surprisingly, the Justice Department has made self-serving assertions that it can handle antitrust enforcement without any changes in the current system. As the foregoing review of the last decade of non-enforcement shows, the facts do not support that claim. If enacted, the bill would impose new obligations on the Antitrust Division to develop appropriate enforcement guidelines relevant to agriculture, and it would change the standard for determining the legality of mergers in agriculture to ensure more effective public and private enforcement of the antitrust laws. If that legislation were the law today, the JBS/National/Smithfield merger would never even be under discussion.
If America's farmers and ranchers are to have the benefit of a workably competitive market, there must be a more active and informed oversight of both the practices of buyers and sellers and a much stronger commitment to restoring competitive market structures. In light of the past decade of experience, it is clear that change is required.
