To Keep U.S. in the Chips, Modify the Antitrust Laws

Formation of U.S. Memories Inc., the consortium of U.S. computer and semiconductor firms to make state-of-the-art DRAM chips, indicates that U.S. firms are finally beginning to use cooperation as a tool for competition. Unfortunately, the consortium is born under the shadow of outdated U.S. antitrust law that will restrict the company’s ability to compete with foreign producers. The new company must endeavor to obtain full value from technology it acquires or generates, while avoiding antitrust policies adopted in a different era for different problems.

U.S. Memories, which apparently will receive equity contributions from seven firms (Hewlett-Packard, IBM, Advanced Micro Devices, Digital Equipment Corp., Intel, LSI Logic and National Semiconductor), will establish facilities to manufacture DRAMs, some fraction of which the investors will be obligated to purchase. Its initial technology will be licensed from IBM.

This is a novel arrangement for U.S. industry, and it indicates a recognition of common interests between semiconductor users and producers, and also among the users. The competitive disadvantage that many U.S. computer manufacturers face vis-a-vis their Japanese competition may be softened by this venture.

But the new firm is already affected and impeded by U.S. antitrust laws. More bad news may be in store if the new company is successful. We see several ways that this venture is affected by antitrust law.

--While the need for such an enterprise has been clear for years, it has been slow in coming in part because managers and lawyers had to wind their way through a thicket of antitrust concerns and constraints. Even preliminary discussions among competitors create antitrust concerns, because it may be illegal for competitors to discuss future investment plans.

--The new venture is exposed to possible lawsuits from other U.S. chip producers. If it wants to migrate into related products like static RAMS, it could face treble-damage lawsuits from specialist firms.

--Investors will not have access to the venture’s proprietary memory technology. We see no sound economic reason for this stance; it may stem from antitrust concerns.

--U.S. Memories’ decision to build new facilities, rather than use the capacity of several of the co-venturers, may also be driven by antitrust uncertainty regarding cooperative contractual arrangements.

Finally, a key reason why firms want to be part of U.S. Memories is because DRAMs are “technology drivers.” Producing them generates new skills and capabilities relevant in other areas. To imprison this technology for antitrust reasons is to handicap the company’s position to contribute to U.S. competitiveness.

Before the patience of U.S. Memories’ partners wears thin, and before they compromise the enterprise in order to steer around antitrust concerns, Congress should modify antitrust laws to ensure that arrangements like this are legal. Antitrust reform is no panacea for the problems facing U.S. industry, but it could lower a major barrier for high-technology firms, particularly small and medium-sized ones.

Bills recently introduced in Congress would reduce treble-damage antitrust exposure now faced by firms that enter into cooperative agreements and consortiums created to develop and commercialize new technologies. Both the “registration” approach of one bill and the “certification” approach of another are desirable.

A registration procedure operates with modest government involvement, but provides modest relief as well. It would reduce antitrust exposure to single damages instead of treble damages.

The proposed certification procedure would involve review by the Department of Justice or Federal Trade Commission, in consultation with the Department of Commerce, to ensure that the cooperative arrangement does not create substantial market power. A cooperative innovation arrangement among firms whose combined market shares are less than 20% would be automatically approved. When combined shares are greater than 20%, the arrangement would be approved only if it improved competitive performance, and the combination was warranted on the basis of scale, complementary capabilities and a demonstrable need for operational and strategic coordination. If a certificate of exemption is obtained, the threat of antitrust damages is eliminated and private antitrust challenges would be limited to injunctive relief.

Certification is a necessary companion to registration because it provides a mechanism for evaluating and approving cooperative arrangements and consortiums involving firms with large market shares. Registration alone will not provide the certainty needed to encourage large-scale consortiums. If Congress adopts only a registration approach, it can expect to see a continuation of the present case-by-case, industry-specific petitioning for antitrust exemptions, which is expensive, time-consuming and unpredictable.

These antitrust law changes are not going to remedy America’s relative decline in key technologies. But at least they will bring the United States closer to the policies of Europe and Japan, and in time, will encourage successful cooperative commercialization of innovation.