STATEMENT OF COMMISSIONER ORSON SWINDLE
IN THE MATTER OF INTEL CORPORATION
DOCKET NO. 9288
When the Commission accepted the consent agreement with Intel Corporation last month, I said that I would take the opportunity to express my views about it following my medical leave. In this statement I will address issues arising from both the consent agreement and the administrative complaint, from whose issuance I dissented last June. Since we do not have the benefit of a trial record here(1) -- and because the information in hand does not allay the misgivings I have had since the outset -- I hope that public comment on the consent agreement will provide helpful guidance on how to vote once the agreement comes off the public record.
In essence, the complaint consists of an allegation that Intel has monopoly power in general-purpose microprocessors (complaint ¶¶ 4-10, 38(2)); an allegation that Intel engaged in exclusionary conduct toward several customers by cutting off key technical information and microprocessor prototypes in order to coerce those customers to license certain of their intellectual property to Intel (¶¶ 11-37, 39); and concluding allegations that, through its exclusionary behavior, Intel has both illegally maintained its monopoly power in general-purpose microprocessors and attempted to monopolize current and future generations of such microprocessors, in violation of Section 5 of the FTC Act (¶¶ 40-42).
In the first place, there is no doubt that Intel has long bestrode the market for general-purpose microprocessors, but there has also been reason to ask whether Intel's position in the market is as unassailable as the complaint suggests. It is widely recognized that Intel is facing vigorous competition in supplying microprocessors to the segment consisting of personal computers costing less than $1000-- a segment toward which a good deal of consumer demand appears to have been shifting lately. Although Intel has not faced challenges of the same magnitude in the midrange and high-end segments of the business, some have also questioned the durability of the firm's dominance of those segments as well. In the absence of a full-blown adjudicative record that might have proved what Paragraph 38 alleges, available information has not dispelled my questions about whether Intel has monopoly power-- as opposed to just an extremely large market share -- in general-purpose microprocessors.(3)
Second, even if one were to assume Intel's monopoly power, I have misgivings about the theory of violation underlying the complaint. The complaint claims that Intel took action against three customers -- firms whose primary significance to the case, according to the Commission's own documentation, lies in their being manufacturers of PCs, not in their being competitors of Intel in the microprocessor market.(4) What action did Intel take against those customers, and for what reasons?
The Commission's complaint says that Intel cut off the supply of technical information and microprocessor prototypes to Digital Equipment, and demanded the return of information and prototypes already in Digital's possession, after Digital sued Intel for patent infringement (¶¶ 18-19). Intel took similar actions against Intergraph, a customer focused largely on workstations and servers, after Intergraph spurned Intel's demand for a royalty-free license to certain Intergraph microprocessor-related technology (¶¶ 26-29). Finally, Intel cut off technical information to Compaq Computer, which had earlier sued Packard Bell Electronics on the theory that certain Packard Bell computer systems used Intel microprocessors that infringed Compaq's patents -- a lawsuit in which Intel felt an obligation to intervene on behalf of the defendant (¶¶ 34-35). According to the complaint, Intel's purpose in taking these actions was to "forc[e] those customers to grant Intel licenses to microprocessor-related technology developed and owned by those customers" (¶ 13). The alleged effects of Intel's behavior were "to diminish the incentives of those three Intel customers -- as well as other firms that are Intel customers or otherwise commercially dependent upon Intel -- to develop new innovations relating to microprocessor technology" (¶ 14) and to "entrench [Intel's] monopoly power in the current generation of general-purpose microprocessors and reduce competition to develop new microprocessor technology and future generations of microprocessor products" (¶ 39).
At this point I do not have sufficient information to be confident that complaint counsel would have proved these rather dramatic charges. My vote against pursuing the case last June, especially as regards Intel's conduct toward Digital and Compaq, rested in part on my sense that the Commission had not sufficiently considered the grounds on which even a putative monopolist is entitled to withhold aid and comfort from another company that threatens serious harm by suing it or suing a third party on whose behalf the monopolist is obligated to intervene. It was my judgment then, and it remains so now, that one could plausibly view Intel's conduct in precisely such an exculpatory light. If the Commission intended to broadcast some kind of general admonition that a monopolist in these circumstances cannot resort to "self-help" (by, e.g., withdrawing and withholding technical information and prototypes) but must instead hire lawyers and take its disputes through lengthy and expensive litigation, then that is a message to which I most assuredly do not subscribe. On the other hand, if the complaint was meant to tell a narrower, more traditional antitrust story based on harm to competition and consumers -- in this case, harm to innovation in a high-technology industry -- I remain unsure whether even that more modest edifice can rest on Intel's decision to withdraw assistance from a handful of customers who were litigious or otherwise flouted Intel's wishes.(5)
Before I turn to the order, I wish to address one other consideration concerning issuance of the complaint against Intel. Regardless of how one characterizes the dealings between Intel and its three customers -- i.e., regardless of whether one accepts the complaint's claim that Intel used its monopoly power to unfairly gain access to intellectual property developed by those customers -- I do not believe that the complaint spells out an especially coherent theory of how those dealings harmed consumers. Consumer welfare is the touchstone of antitrust enforcement, and the "public interest" standard of Section 5 of the FTC Act embodies considerations of consumer welfare. In the absence of clear evidence of how Intel's dealings with Digital, Intergraph, and Compaq could have adversely affected consumers, one can question the very basis for issuing this complaint -- and for injecting a government agency into the dynamic workings of a fast-moving, high-technology industry.(6) I look forward to any public comments that deal with the likely harm to consumers stemming from the misconduct alleged in the Commission's complaint.
Regarding the proposed order itself, some observers have characterized it as having achieved whatever objective prompted the Commission's suit against Intel. I am not so sure, in part because of my uncertainty (discussed earlier) over what message the complaint was meant to communicate and in part because of the very terms of the order. In fact, given my reservations about the merits of the complaint, I would be more concerned about the order -- comprising a difficult-to-enforce mandate to "sin no more," with a major proviso and some significant exceptions -- if it seemed likely to impose real and significant restrictions on Intel.
I expect the proposed order to present possible enforcement difficulties because, among other things, its basic prohibition (order ¶ II.A) commands Intel not to take certain adverse actions against microprocessor customers with regard to "Advance Technical Information" "for reasons related to an Intellectual Property Dispute"(7) and not to "base any supply decisions for general purpose microprocessors upon the existence of an [Intellectual Property] Dispute." No matter what may motivate Intel's future decisions whether to furnish technical information and microprocessor prototypes to customers, it is extremely doubtful that Intel is going to create any kind of record that will enable the Commission to ascertain whether such a decision is "for reasons related to" or "base[d] . . . upon" the one ground made impermissible by the order -- an intellectual property dispute. Exacerbating the impact of Paragraph II.A's subjective language are two further paragraphs that allow Intel to withhold advance technical information from customers (order ¶ II.B.2) or make product supply decisions (order ¶ II.B.3) based on "business considerations unrelated to the existence of the [Intellectual Property] Dispute" -- further verbiage that appears to make order enforceability hinge on difficult inquiries into the state of mind of Intel decision makers. I hope that my pessimism is unwarranted, but the key terms of the order seem destined to enmesh the Commission in expensive, and perhaps intractable, enforcement proceedings if Intel is ever suspected of violating it.(8)
I end where I began -- searching for information to help me decide whether I now have reason to believe that Intel violated the law and, if so, whether I can support this consent order. I genuinely look forward to receiving public comments both supportive and critical of the settlement and the underlying theory of violation. I hope that the considerations spelled out in this statement will be helpful to those preparing to submit comments to the Commission.
1. Were we considering this matter at the conclusion of an adjudicative proceeding, I would of course base my analysis strictly on information in the adjudicative record. In the absence of such a record, I am compelled to rely on other sources of information.
2. Unless otherwise indicated, all further citations to paragraph numbers refer to the administrative complaint.
3. In their statement, my fellow Commissioners -- citing complaint counsel's pretrial brief as support -- assert that "Complaint Counsel said it would offer evidence that the sub-$1000 segment was a small and relatively unprofitable portion of the market . . . and that at the high-end, many computer manufacturers have been abandoning their proprietary microprocessor designs in favor of Intel's . . ." Moreover, according to my colleagues, "[i]n the market as a whole, Complaint Counsel contended that Intel's share had grown, not shrunk, and was in the range of 80% or more." Statement of Chairman Robert Pitofsky and Commissioners Sheila F. Anthony and Mozelle W. Thompson at 2 n.1. I do not disagree that these are propositions that complaint counsel aimed to establish. My point is simply that I have not yet been persuaded by the evidence in the Commission's possession -- as distinguished from complaint counsel's representations and contentions -- that Intel possesses monopoly power in the relevant market.
4. Of course, both Digital (the developer of the Alpha microprocessor) and Intergraph (which developed the Clipper chip prior to 1993) were not only Intel's customers but also -- at least to the extent that they were able to chisel away at Intel's alleged monopoly -- its competitors in the microprocessor market. The Commission's complaint, however, is couched almost entirely in terms of Intel's allegedly anticompetitive behavior toward three victims that needed Intel technical information and prototypes so that they could build computers. And although press releases do not necessarily reflect the official views of the Commission (in the sense that the complaint does), both the June 8, 1998, FTC press release that announced the issuance of this complaint as well as the March 17, 1999, release announcing the Commission's acceptance of this consent agreement spoke almost entirely in terms of Intel's conduct toward its customers. Even if Digital and Intergraph can be characterized as Intel's present or erstwhile competitors -- thereby giving this matter more of the character of a traditional monopolization case -- the Commission has consistently placed far greater emphasis on the supplier/customer relationship between Intel and its alleged victims.
5. My colleagues characterize Intel's conduct as "coercion that forces customers to license away patent rights on unfavorable terms" (Statement of Chairman Pitofsky and Commissioners Anthony and Thompson, supra n.3, at 1), which begs the important question whether Intel was truly engaged in such coercion or was instead defending against attacks by its alleged victims. Regarding my doubts about whether Intel's alleged conduct (and its anticompetitive effects) could have been proved, my colleagues state that these allegations "are consistent with our knowledge of the industry and with common sense . . ." Id. It bears repeating that my concerns arise from the state of the evidence underlying the Commission's allegations. I take little comfort from -- indeed, I am not sure I fully understand -- the notion that monopolization allegations are "consistent with our knowledge of the industry and with common sense." I do not -- as my colleagues suggest (id.) -- "demand certainty" about the facts at issue, but I do look to the strength of the evidence rather than to what a litigant's pretrial brief might promise to deliver (id. at 2).
6. I note that to the extent this case is depicted as involving harm to Intel's competitors (rather than to its customers), that would tend to attenuate further any theory that Intel's conduct threatened harm to consumers.
7. All italics in this paragraph are added.
8. Presumably in response to my point about the difficulty of order enforcement in this case, my fellow Commissioners note that "[c]ertainly the order could have been made more certain in its application by, for example, requiring Intel to deal with all comers on identical terms, regardless of circumstances or the credit-worthiness or other characteristics of would-be customers." Statement of Chairman Pitofsky and Commissioners Anthony and Thompson, supra n.3, at 2. There is nothing in my statement to suggest that I would favor an order drafted along such rigid, mechanical lines. My only point was that, in its current form, the order against Intel could present formidable enforcement problems.