Privacy, Antitrust and the National Information Infrastructure: Is Self-regulation of Telecommunications-related Personal Information a Workable Tool?
Donald I. Baker
W. Todd Miller
Baker & Miller PLLC
The National Telecommunications and Information Administration (NTIA) has raised a very serious set of issues in its Report entitled, Privacy in the National Information Infrastructure; Safeguarding Telecommunications-Related Personal Information (The Privacy Report). These issues are important to each of us as citizens in a free society in which it is increasingly difficult to hide from solicitors or sleuths. They are also important to us as consumers who may desire a quiet, uninterrupted dinner hour. More important for the antitrust laws, the issues raised by The Privacy Report are critical to providers of consumer-oriented services and merchandise in the United States. Thus, the issues in The Privacy Report touch not only ordinary people's personal lives, but what may be core issues and opportunities for businesses in a dynamic set of competitive markets. In this environment, self-regulation is fraught with antitrust risks.
What The Privacy Report calls "telecommunications-related personal information" (TRPI) is information about individual consumers collected by different suppliers. The specific information collected about any one consumer may differ depending on the supplier's relationship to that consumer. Telephone companies may have one set of information (on a consumer's communications habits), cable companies another set (on entertainment preferences), and credit card issuers a third set (on spending habits) with the information undoubtedly overlapping to some degree. Increasing liberalization of currently regulated markets (e.g., local telephone service) will likely cause these suppliers to want to use the TRPI that they possess to compete in providing a broad range of services, merchandise and opportunities to consumers. This is very important in practical terms: customer acquisition costs can be critical to various TRPI-related businesses (such as on-line computer services), and access to TRPI may be a vital factor in effectively competing for customers.
Stated another way, there is currently a market in TRPI. Companies are willing to spend significant amounts of money to purchase information about the demographics of individual consumers. As telephone companies and others that are currently regulated in their use of TRPI seek to enter new product or service markets, they will likely desire to increase their participation in TRPI markets. Thus, one can anticipate that ever increasing information about consumers will be bought and sold. The basic policy questions presented by this increasingly important TRPI market are easy to state and hard to answer:
- When and how is the initial collector of TRPI (e.g., a bank credit card issuer) to be allowed to use TRPI in merchandising other services to the same consumer? and
- When and how can the initial collector transfer the TRPI to some third party--presumably for payment--and, if so, on what terms?
A rule that sought to prevent the initial collector from using TRPI to merchandise its own products and services would create immense regulatory problems and would likely be deemed one in futility: would we really want to bar the local telephone company from offering selected consumers better services simply because it determined the potential demand by collecting TRPI? On the other hand, what is the extent of use that should be permitted by the initial collector? This becomes a real antitrust issue if the collector has market power for the products/services for which the TRPI is collected. Should it be permitted freely to favor its affiliated enterprises by providing TRPI to them, but denying it to third-party competitors?
Even a rule that prevented the initial collector from transferring TRPI to third-parties or affiliates would tend to enhance the market position of the initial collector vis-a-vis other kinds of competing enterprises that depend on the purchase of consumer profile information. A prohibition on transfer would also tend to entrench the position of current market participants because new entrants--even those that eventually will be able to collect significant TRPI--may not be able to focus their marketing efforts on consumers most likely to desire their products or services. Such a rule could be a matter of antitrust concern if the initial collector enjoyed a substantial degree of market power (as might currently be the case with the local telephone company or exclusive cable franchisee). In any event, it is clear that there will be divergent interests between net collectors of TRPI (phone companies) and net acquirers of TRPI (catalog sales companies and new entrants).
Consumers' interests might also be quite diverse. From a consumer's perspective, the market for TRPI reflects necessary tradeoffs between (a) privacy concerns (including the intrusions and burdens of being inundated with commercial offerings generated on the basis of TRPI) and (b) increased knowledge of potentially desirable goods and services. For many consumers, privacy concerns are paramount: given a choice, they would block all third-party access to their personal TRPI. Others are not greatly concerned by the general availability of such information (but may be concerned depending on the type of information made available and to whom). Many others fall somewhere in the between the extremes.
Balancing these important interests (both consumers' and suppliers') would be immensely difficult for a government regulatory agency acting under a "public interest" standard. Different interest groups and different decision makers would almost certainly have different judgments on the critical details. Importantly, however, the decision would at least be made by a group without any economic stake in the outcome.
The same cannot be said of a self-regulatory body. Quite simply, the same type of "public interest" balancing issues become much more difficult in a self-regulated environment--where different enterprises and different interest groups have seriously different stakes. To start with, who is going to be represented on the self-regulatory body? Collectors (and potential sellers) of TRPI? Potential acquirers of TRPI? The American Civil Liberties Union? Assuming that a reasonably representative self-regulatory group can somehow be created, how would decisions be made? How transparent and open would the process be? Are the group's members likely to be able to come up with reasonably specific standards that can be administered in an even-handed way among enterprises which may have significantly diverse positions and interests?
SELF-REGULATION--A QUICK HISTORICAL LOOK
The medieval guilds were classic self-regulators. The task of the guild was to regulate a particular trade in a particular city or other geographic location. The typical guild apparently spent time and effort maintaining standards among those in the trade, but it also spent a lot of time and effort in defending the guild's turf against outside competitors--from more distant cities or other trades. The protectionist history of guild regulation underscores a fundamental point: self-regulators often combine--and sometimes confuse--self-regulation with self-service.
History teaches us that self-regulation may work reasonably well where the self-regulatory group is open to all who want to participate and the participants have a common interest in maintaining industry standards for honesty, fair dealing, etc.--all for the purpose of encouraging buyers to resort to the goods and services of the self-regulated group.
Self-regulation may also work in a network or market-creating environment, where the participants in the scheme have a common interest in having standards that enable them to connect with or trade with each other and thereby achieve economies of scope and scale. In a network environment, such a process is most likely to work in a competitively neutral way where the technology is relatively stable (e.g., establishing the standard railway gauge) and where the standard does not seriously infringe on the innovative abilities of individual competitors. In the market-creation setting (e.g., the stock exchange), the self-regulatory scheme is most likely to work where the group simply sets forth the rules for the operation of the market, such as how trades are cleared.
In each of these often-successful examples, those involved in the self-regulatory process had a common interest in the success of the common effort and the outsiders (if any existed) would not experience any significant competitive disadvantage from not being involved.
The process does not work so well where the market is changing and the self-regulating group (SRG) faces the risks of competition and competitive innovation from outside the group or from dissenting SRG members. In these circumstances, the self-regulatory process almost inevitably becomes a defense of the status quo against new ideas, new competitive innovations and, worst of all, new competitors.
The self-regulatory process works worst--and echoes the medieval guilds--when it seeks to regulate the output of its members or the prices and terms on which members can serve the public, in circumstances where the SRG members collectively have market power.
Antitrust history brightly illustrates the foregoing points. It is rich with self-regulatory schemes that have failed, but also includes a few examples where self-regulation has succeeded. Most of the self-regulatory antitrust suits are of the "boycott" variety: they involve access to facilities, information, certification or standards set by an SRG. Often the SRG consists of the traditional providers who seemed--sometimes blatantly--to curb outsiders and innovators. Here is a diverse, representative, but far from complete sample:
In British & Foreign Marine Insurance Co., Ltd. v. Ruddy Brook Clothes, Inc., 195 F.2d 86 (6th Cir. 1952), cert. denied, 344 U.S. 816 (1952), the Court of Appeals upheld against a "boycott" claim the exchange of customer loss information among fire insurers designed to avoid fraudulent fire insurance applications: armed with this information, each insurer was still free to make its own decisions, but all had legitimate common interests in having accurate risk information.
In Terminal Railroad Ass'n v. United States, 226 U.S. 420 (1913), the Supreme Court recognized the value of having common interchange arrangements that allowed traffic to flow through the St. Louis gateway (but sustained injunctions to prevent discrimination against outsiders).
In Fashion Originators' Guild v. Federal Trade Commission, 312 U.S. 457 (1941), the Supreme Court struck down a self-regulatory scheme--created by a group of higher-priced dress designers designed to regulate and exclude those who would copy the dress designs of the high-price originators. The Defendant organized a boycott scheme whereby each "originator" agreed not to deal with the outlets to which the "pirates" sold their goods. The Supreme Court struck it down saying that, "the combination is in reality an extra-governmental agency, which prescribes rules for the regulation and restraint of interstate commerce, and provides extra-judicial tribunals for determination and punishment of violations, and thus, 'trenches upon the power of the national legislature and violates the statute'" (citation omitted).
In Goldfarb v. Virginia State Bar, 421 U.S. 773 (1975), the Supreme Court struck down a scheme under which the local county bar association, supported by the state bar, sought to prescribe the minimum prices that lawyers could charge for real estate services. A justification based on ethics and standards was rejected.
In National Society of Professional Engineers v. United States, 435 U.S. 679 (1979), the Supreme Court struck down a self-governing organization's ethical rules that prevented the negotiation over fees for engineering service until after the engineer had been selected for the job. This was an obvious attempt to insulate the Defendants' members from price competition and was justified on the grounds that ruinous price competition would lead to unsafe structures. It was rejected. The Court explained that "the [analysis of restraints under the Sherman Act] does not support a defense based on the assumption that competition itself is unreasonable."
Board of Regents of the Univ. of Okla. v. NCAA, 468 U.S. 85 (1984), involved a self-regulatory effort to maintain a "level playing field" among football-playing colleges. The NCAA adopted a rule that required all football telecasting rights be done through the NCAA, preventing each individual member from going out and selling its rights based on its own merit. The Supreme Court found this an unreasonable restraint of trade, rejecting the whole series of saintly "public interest" justifications (e.g., full stadiums for local games).
In U.S. v. National Association of Broadcasters, 1982-93 Trade Cas. (CCH) ¶ 65,049 (D.D.C. 1982) (consent decree), the Department of Justice challenged a self-regulatory scheme limiting the number of minutes of advertising that a TV broadcaster could run in any particular hour. The Department asserted that this arrangement, regardless of its popularity among politicians, was simply an output limitation that would be likely to result in higher prices for TV advertising. Thus, the SRG was serving its member economic interests while purporting to do politically-acceptable things.
FTC v. Indiana Federation of Dentists, 476 U.S. 447 (1986), dealt with a boycott claim based on the refusal of an SRG to permit its members to transmit X-rays to insurance companies on the ground that it was inconsistent with professional standards.
Allied Tube & Conduit Corp. v. Indian Head, 486 U.S. 492 (1988), involved a certification process in a self-regulatory context. The Supreme Court upheld liability, where the makers of the product (metal pipe) had manipulated the certification process to deny access to the market certification for makers of plastic pipe (including the plaintiff).
U.S. v. Realty Multi-List, Inc., 629 F.2d 1351 (5th Cir. 1980), is a particularly illuminating case. Here, the local realty board, as self-regulator, had determined that only those real estate brokers with full-time practices and offices in the relevant geographic area would have access to the new multiple listing service they had established. The court, in imposing liability on a boycott theory, noted that because part-time brokers, whom the defendants had excluded, were licensed by the state, there could be no inference that they had lower ethical standards. Instead, the court simply viewed the restriction as an effort by an established group to protect themselves from competition. In reaching its judgment, the court quoted the famous passage from Adam Smith's Wealth of Nations: "People of the same trade seldom meet, even for merriment or diversion, but the conversation ends in a conspiracy against the public or in some contrivance to raise prices."
Silver v. New York Stock Exchange, 383 U.S. 341 (1963), involved a challenge to a New York Stock Exchange rule prohibiting direct wire connections between members and non-members. After finding that SEC did not have jurisdiction over this particular exercise of self-regulation, the Court held that the failure to provide a fair hearing to non-members before having direct connections disconnected was actionable under the antitrust laws.
There are many more cases, both in the Supreme Court and more in the lower courts. The short answer is that antitrust courts have often been fairly skeptical of the "public interest" justifications offered by SRG's when they seem to be protecting their members or outside competition or depriving the public of full and open competition among their members.
ANALYSIS OF OPTIONS
Self-regulation over the distribution of TRPI appears to be an unwise idea because TRPI is so competitively sensitive in a highly-competitive marketplace for consumer merchandising and services. If a SRG with responsibility for TRPI echoes history to the slightest degree, it will tend to favor the self-regulated possessors of TRPI over outsiders in the use of this information. See especially Allied Tube, supra. In so doing, it will run antitrust risks for itself and for its members. The credible legal challenges thus presented will tend to slow down and divert the process. There is no body, of which we can conceive, capable of balancing the interests of TRPI collectors (and potential users), TRPI buyers and manipulators, and consumers.
To address these competing concerns in part, The Privacy Report recommends that (a) consumers be notified of and (b) consent to the use of TRPI for commercial purposes. (The Privacy Report, pps. 8-9). The Privacy Report, however, appears also to contemplate the possibility of a SRG with broad jurisdiction to oversee in some way the market for TRPI.
The Privacy Report sidesteps other alternatives to self-regulation even though there seem to be other alternatives to granting broad jurisidiction to a SRG. Indeed, we would respectfully suggest that more market-oriented solutions be permitted to develop.
While we are not aware of any specific studies on this issue, we assume that, in a competitive product or service market that does not contain restrictions on the buying and selling of general TRPI, the rents collected or expended in trading TRPI would be ultimately reflected in product or service costs. In other words, if the product market is competitive, and one supplier has desirable TRPI that it could sell (TRPI-favored supplier) and another does not, the TRPI-favored supplier would be able to (and would) offer a lower price for its product because it also enjoys an income stream from the TRPI. Both consumers and the TRPI-favored supplier benefit--consumers enjoy lower prices and the TRPI-favored supplier is a more effective competitor against the non-favored suppliers. In this situation, consumers could be assured that their assent to providing TRPI ultimately accrues to their benefit.1
Where TRPI is initially collected by an enterprise that does not face effective competition, there would not appear to be any market force that permits consumers to participate in the benefits of the TRPI market. The consumer is thus faced (in a new environment of notice and consent) with the simple trade-off between (a) privacy and (b) lost opportunity for the consumer from use of increased information about products and services. In this situation, the market might increase the economic incentives for consumers to consent to the sale of their TRPI by the monopoly collector. For example, the collector of TRPI might agree to pay for the use of TRPI, in a contractual arrangement that could presumably be modified from time-to-time.2 In other words, the collector of TRPI that wished to sell it to another entity would offer to pay the consumer for the information.
The public regulatory interest is simply that there be accurate disclosure to the consumer about the fact and limits of the use of TRPI and that the consumer consent to such use. Provided that the regulators do not mandate a binary choice for consumers (permitting or forbidding the use of individual TRPI), a comprehensive market for TRPI should develop. In such a market, the consumer might be offered a chance to refuse consent unless he/she receives a satisfactory price.
In sum, there is more than the consent question raised by The Privacy Report. There is a market issue. Because there is a market for TRPI, there is no reason why we should presume that only one side of the market equation--suppliers--should determine how the market for TRPI works. The consumer, as the ultimate creator of such information, should be given more than a simple "yes or no" choice in the matter. Some might have a relatively high price for consent, but others might have much lower prices. This is what markets are about and this is the way markets should work. The antitrust laws are implicated when a private body attempts to dictate what are essentially market determinations.
We believe that a market-oriented approach would minimize the antitrust risks. If there is to be any self-regulation at all, it should be clearly defined and quite limited. This limited role for a SRG could include (a) developing disclosure standards, (b) determining how persons that misuse TRPI would be disciplined, and perhaps (c) collecting and disseminating information about the TRPI market.
For any SRG authorized in the TRPI market, we would make these recommendations to reduce antitrust risks, including:
The self-regulatory body should be broadly representative, and should not be confined to the collectors of or traders in TRPI. Contrast Gibson v. Berryhill, 411 U.S. 564 (1973) (self-regulatory group entirely made up of providers).
The mandate to the self-regulatory group should be as clear and precise as possible and should seek evenhandedness between consumer interests, provider interests, and the interests of new and alternative providers.
Some review of the SRG should be provided. This could be done either through the possibility of legal challenge to certain standards set forth in the SRG's charter or through a regulatory agency.
The Business Review Procedure of the Department of Justice, Antitrust Division should be utilized, particularly if a market-oriented approach is not pursued. See 28 C.F.R. § 50.6. Under this procedure, a proposed course of conduct, including a self-regulatory proposal, may be explained to the Justice Department which is asked to state its present enforcement intentions with respect to the conduct. This process is open and public in which the Department ultimately states a position and makes public the record on which it relies. In the TRPI context, this process would enable diverse and interested parties to participate and to be aware of the outcome. Approval by the Department of Justice under the Business Review process does not assure those engaged in activity of immunity from private lawsuits or lawsuits by state attorney generals, but in practice, it is a reasonable safeguard. Private challenges to conduct approved under the Business Review process have been extremely rare, and none has been successful to our knowledge.
1 There are, of course, interesting free-rider questions. Provided that a large percentage of consumers do not opt out of permitting the sale of their TRPI (which might lower the price paid for such information), those declining to permit such sales still benefit from the broader market effects. If this becomes a significant problem for a competitive market supplier, it may simply "unbundle" its prices to consumers and give a lower price to the consumer who consents to the resale of its TRPI.
2 The questions necessarily raised by such a possibility include: Would the availability of payment significantly increase the likelihood of consent? Would transaction costs be high in relation to the value of the information? In principle, if there is economic value in the information, collectors of such information should be willing to pay for it and consumers should more readily assent to its use.
The National Telecommunications and Information Administration (NTIA) has proposed a framework for dealing with telecommunications-related personal information (TRPI) based upon two fundamental elements-- provider notice and customer consent--implemented through self-regulation. In this paper, we address the antitrust issues that are likely to arise when rival companies meet and agree on the manner in which they will use, or more to the point, not use, TRPI.
We conclude that industry self-regulation regarding the use of TRPI, while not exempt from the antitrust laws, is unlikely to run afoul of the antitrust laws, so long as the rules imposed through self-regulation are sensibly designed to protect privacy and not to limit competition.
SELF-REGULATION INHERENTLY RAISES ANTITRUST ISSUES
Self-regulation, whereby competing firms agree on certain aspects of their conduct, inherently raises antitrust issues. The most straightforward example arises when companies agree to "regulate" the price they will charge their customers, or the territories in which they will each sell. Such conduct is more commonly known as price fixing and market division, activities that are per se illegal under the Sherman Act. The law prohibits such forms of collective conduct because they harm competition and have little or no redeeming value from the perspective of consumer benefits or economic efficiency.
Antitrust concerns can arise even when collective behavior by companies in the same industry bears no resemblance to cartel behavior. For example, there is a developed body of antitrust law that deals with trade associations, which regulates the manner in which associations may provide for information exchanges among industry members or may agree on standards that govern product attributes.
The existence of a legitimate public purpose, in and of itself, has not been sufficient to shield such agreements from antitrust condemnation. Antitrust institutions are inherently suspicious of claims that agreements that are harmful to competition have in fact been adopted to support a lofty public purpose. For example, in National Association of Professional Engineers v. United States,4 the United States Supreme Court rejected a claim that a rule adopted by an association of construction engineers was needed to protect the public against inferior designs that would threaten public safety. The Court said that this health and safety defense constituted "nothing less than a frontal assault on the basic policy of the Sherman Act."5
Some readers may find it peculiar or even perverse that antitrust law can present an obstacle to achieving certain meritorious objectives, in particular the protection of privacy. Our response to this is two-pronged. First, we note that antitrust laws, with their single-minded goal of preserving competition, can indeed be blind, or at least insensitive, to other public policy goals. For example, nonprofit institutions are not exempt from antitrust laws, even when they arguably are pursuing laudable objectives. For example, when several leading universities agreed to standardize the treatment of financial aid, ostensibly to use their financial-aid budgets to help more students (rather than competing to attract certain star applicants), they were still subject to antitrust challenge.6
Our second point is that the antitrust objection to arrangements that promote laudable public purposes is often more apparent that real. Faced with assertions that bans on competitive bidding are necessary to protect lives or that bans on price advertising are necessary to protect consumers, antitrust institutions have developed a healthy skepticism toward claims that agreements that have a tangible adverse impact on competition should be upheld because they promote some other public purpose. It is doubtful, however, that the Supreme Court would have rejected the defense of the ban on competitive bidding as an assault on the Sherman Act had it really believed that competitive bidding leads to collapsing skyscrapers. When faced with a more plausible claim that a financial aid coordination agreement among private colleges was necessary to promote equality of access to higher education, the courts were willing to allow the participating universities to prove the necessity of the agreement to this goal.
Yet more often antitrust institutions--the government agencies and the courts--take the position that although the government may override competition values to promote social or public policy goals, private actors cannot reduce competition for equally laudable purposes. The antitrust laws recognize that the government may promote public objectives by empowering private actors to engage in anticompetitive conduct (including even price fixing), and therefore recognize an antitrust defense based on state action without second- guessing the public policy purpose behind the government action. The state-action defense requires a showing that the anticompetitive collective action is (a) authorized by a clearly articulated state policy to displace competition, (b) embodied in state law, and (c) subject to active supervision by the government to ensure that the collective action is consistent with the state's goals.7
Public officials are not empowered to immunize competitors from the antitrust laws. Although the surgeon general has determined that smoking poses a serious health hazard, the FTC challenged a merger of cigarette makers in 1995, long after the harm caused by smoking had been well established.8 The fact that the competitive harm that the FTC's action sought to avoid was higher cigarette prices, and therefore lower consumption of cigarettes, did not affect the antitrust considerations in any way.
Two more examples involving advertising illustrate the point. Although commercial messages on television are annoying to most viewers, a code of conduct promulgated by television broadcasters to restrict commercial time was successfully challenged as an anticompetitive agreement.9 The antitrust concern was that restricting the output of commercial time was akin to an output-restricting cartel and would lead to higher charges for commercials. Chastened by the experience, broadcasters sought and obtained a Congressional exemption from the antitrust laws in order to negotiate ways of addressing public objections to the amount of sex and violence in commercial broadcasts, even though the competitive harm in such a collective action is not readily apparent.10
The self-regulation approach advocated by the NTIA, like most self-regulation efforts, is thus subject to the antitrust laws. Unless exempted from the antitrust laws by legislation, collective action by competitors is subject to antitrust review and possible condemnation, no matter how laudable the purposes behind it. But antitrust jurisdiction and antitrust violations are not synonymous. The fact that agreements having a laudable purpose fall within the purview of the antitrust laws does not mean that they in fact violate those laws.
ANTITRUST CONSIDERATIONS GOVERNING AGREEMENTS ON THE PROTECTION OF PRIVACY
The point that antitrust jurisdiction is not synonymous with antitrust violations is particularly relevant in the context of private agreements to limit the use of TRPI. Although agreements among competitors on the treatment of customers' private information will not be exempt from the antitrust laws, we do not believe they will in fact violate the law.
The reason for this is simple. The antitrust laws ordinarily condemn only agreements that harm competition and lead to higher prices and lower output. We say "ordinarily" because only a narrow range of antitrust offenses, such as price fixing, are regarded as so pernicious and so unlikely to have redeeming features that they are deemed per se unlawful and condemned without any inquiry into their actual effects. Outside this exception, however, agreements among competitors can only violate the antitrust laws if they can be shown by a challenging party genuinely to harm competition.
Many agreements among competitors do not produce any adverse competitive effects. For example, an agreement among modem manufacturers to support a particular transmission protocol is an agreement among competitors that almost certainly promotes competition by ensuring that competing manufacturers' products will be compatible with each other, thereby increasing the product's usefulness to customers. An agreement among manufacturers of plumbing products to a plumbing code that specifies acceptable pressure tolerances for their products is also likely to promote competition by ensuring customers that compliant products will be safe to use, relieving customers of the necessity of investigating individual products' safety attributes.
It is only when these codes harm competition that the antitrust laws step in. How can such codes harm competition? The codes may be contrived by incumbent manufacturers to exclude newer products that may be cheaper or otherwise superior to their own products.11 For example, incumbent manufacturers of traditional plumbing fixtures may devise a code that excludes the use of a new type of fixture that offers some price or quality improvements over existing products. Manufacturers may also enact codes to standardize product or service features where there are no benefits from standardization, so as to limit the dimensions of competition and thus deprive customers of choice. An agreement standardizing prices is an obvious example. Another example is the code adopted by the Detroit automobile dealers to standardize hours of operation, confining showroom hours largely to the business day, which curtailed the ability of customers to shop for automobiles at nights and on weekends. In addition, standardizing product features may make it easier for producers to collude than if products had more differentiating features. This was the allegation when pasta manufacturers agreed through their trade association to standardize the wheat content of their products.
By reciting these examples of illegal agreements among competitors, we merely intend to illustrate the types of agreements that are likely to raise antitrust concerns. We emphatically do not want to leave the impression that agreements to codify ethical conduct or standardize some attribute of a product or service typically are prohibited by the antitrust laws. To the contrary, the vast bulk of such agreements, and the vast bulk of activities by trade associations, raise no significant competitive concerns. We believe that TRPI codes that are sensibly designed to protect privacy are also unlikely to raise antitrust concerns.
We attempt next to provide some contours for this statement. First, agreements among a broad collection of companies that do not necessarily compete with each other in the sale of any product or service are least likely to raise any antitrust concerns. If the companies do not compete with each other, an agreement among them to maintain the privacy of TRPI cannot, by definition, diminish competition. So if a group of companies in multiple industries--say, automobiles, software, petroleum, and telecommunications--establish a code of conduct (and a logo that goes with it to communicate efficiently their adherence to the code) regarding the ways in which they will use TRPI, this agreement will not raise antitrust concerns because the participating companies are not competing with each other even in the absence of the code and therefore cannot diminish competition between them. As competitors of these companies join the effort, the analysis becomes more complicated, but the predominance of companies from other industries over competitors from any particular industry would be suggestive that the agreement is likely to serve legitimate privacy purposes.
Even if the code is promulgated predominantly by competitors, antitrust concerns arise only if it somehow diminishes competition in the markets in which these companies compete. For example, an agreement among competing companies that sell audio products over the Internet to seek prior customer consent for the disclosure of customer information to other businesses does not implicate any significant aspect of rivalry among the firms. The use of a logo that signifies compliance with such a code may, indeed, promote competition by instilling confidence in potential customers that their personal data will be protected and thereby increasing customer use of, and output within, this channel of distribution.
Our antitrust laws are concerned about agreements that reduce the quantity of goods or services sold, and thus tend to raise price. It is highly relevant, therefore, that certification logos commonly expand the quantity of goods or services by providing useful information to customers related to the goods or services they are buying. Certification logos can operate as an efficient mechanism for signaling to customers important information concerning a product or service that they otherwise may either not acquire at all or acquire only after expending search costs. A common example of the signaling function of a logo involves the Underwriters Laboratory logo. Some customers buy electrical products that carry the UL logo because they know that the logo signifies safety, but most do not know what attributes make the product safe. Similarly, customers may decide to buy a product from a vendor over the Internet because the vendor displays the logo of some privacy-protection code organization, even if they do not necessarily know the details of the privacy protection to which participants adhere. In other words, sensible privacy codes can actually increase output, whereas the harm that the antitrust laws condemn entails a reduction in output.
Despite the pro-competitive aspects of certification logos, agreements on the use of TRPI can give rise to antitrust concerns. Agreements on the use of TRPI that spill over into significant aspects of rivalry among competitors can violate the antitrust laws. For example, an agreement among competitors flatly to refuse to do business with competitors that do not adhere to a specific code on the use of TRPI may violate the antitrust laws. In many circumstances, an agreement among competitors to refuse to deal with another competitor will be viewed by the courts as a "group boycott" under the antitrust laws. Such boycotts are condemned as per se unlawful, without any examination of their competitive impact.12
Although the per se condemnation of such agreements may seem arbitrary--in many cases, a refusal by competitors to do business with a rival is unlikely to have an adverse impact on competition--the likelihood that competition would actually be harmed by such a refusal is particularly great in network industries, in which these issues are first likely to arise. In network industries, the ability to interconnect with rivals is essential to competition. A long distance carrier's service is virtually worthless if it cannot interconnect into the networks of local exchange carriers. A bank's credit card has little value if competing banks that service merchant accounts will not process transactions for it. In other words, in network industries, refusal to deal with a competitor could result in the exclusion of the competitor from the market. Although even an exclusion may not necessarily have an adverse impact on competition--the exclusion of one of numerous providers of a particular product or service may have little effect on prices or output--it is easy to conjure circumstances in which such an exclusion would harm competition.
We have so far posited an industry code that imposes a draconian, and somewhat unrealistic, sanction on competitors that refuse to adhere to a TRPI code, namely a flat refusal to deal. Let us now consider more realistic sanctions. Assume that members of an industry agree to a TRPI code based on the NTIA model, which requires customer consent for the disclosure of TRPI. The companies develop a logo or seal for use by adherents to the code and condition use of the logo on adherence to the code. The denial of the right to use the seal to non-complying companies would not, in itself, raise any antitrust concerns. The seal has value to consumers only if it accurately represents companies' adherence to it. Denying the use of the seal to companies that do not comply with the code that it represents is no more than refusing to participate in consumer deception.
Suppose, however, that the seal or logo is that of the trade association and that the association makes adherence to its TRPI code a condition of membership. The association may want to have membership to signify thoughtful use of TRPI and may therefore seek to exclude from membership those companies that will not adhere to the code. Can the denial of trade association membership for non-adherence to a TRPI code violate the antitrust laws? Unless the association controls access to some inputs that are essential to competition, the exclusion of companies that refuse to adhere would be very likely to survive antitrust scrutiny. To succeed in an antitrust claim, the excluded firm would have to demonstrate that the agreement to exclude non-complying firms from the association actually harms competition by increasing prices and diminishing output.13 To do so, it would have to show that its costs are raised significantly both by the denial of membership and by adherence to the code (for it could avoid the cost of membership denial by adhering) and that these added costs affect prices in the market more generally. Exclusion from a trade association very seldom will lead to such an effect.
As a general matter, we see little of antitrust significance in agreements among competitors to adopt, or adhere to, a code that governs the use of TRPI unless it incorporates some collateral agreement to coordinate market behavior. For example, an agreement to adopt a code requiring the disclosure to customers of the uses to which their TRPI will be made available will not, without more, raise antitrust concerns. Although the antitrust enforcement agencies in the past have maintained a distinction between firms' unilateral decisions to adhere to a code and an agreement to adhere to a code, and placed only the former in the safety zone, it is safe to say that neither situation raises antitrust concerns in the TRPI context because neither implicates a restriction on competition that can be said to lead to higher prices and lower output.
We thought long and hard to come up with an example where an agreement among industry members to disclose TRPI only with customer consent would harm competition. The hypothetical example that we sought would have posited that the use of TRPI obtained without customer consent somehow made a particular competitor or competitors particularly efficient, so that a code requiring such consent would impose significant costs on the company or companies and lead to higher prices. We have not been able to come up with a convincing example of such a situation. Although we cannot rule out the possibility that such a case might arise, so that a code limiting dissemination of TRPI without customer consent would indeed raise antitrust concerns, we think that the possibility of such a situation is remote. And even if the situation were to arise where it could be plausibly claimed that a TRPI code would have that impact, the antitrust inquiry would take into account the benefits of the code and balance them against the claimed harm
How can the benefits be credited when courts are skeptical of public policy justifications? In the case of TRPI, the public policy justification--respect for privacy rights--is likely to coincide with a competitive justification. Recall that a key function of a privacy code is to instill customer confidence in services that implicate privacy concerns so as to increase demand for and output of those services. If respecting the wishes of customers for privacy in fact tends to lead to higher output, the antitrust analysis must weigh this effect against any diminution in output brought about by the TRPI code.
In conclusion, antitrust issues ought to be on the minds of any industry group that codifies rules of behavior for competitors. Having said that, we are very comfortable in concluding that industry codes that are sensibly designed to protect privacy are very unlikely to give rise to antitrust problems.
1 E.g., California Dental Association, 5 Trade Reg. Rep. (CCH) ¶ 24,007 (FTC March 25, 1996); Massachusetts Board of Registration in Optometry, 110 F.T.C. 549 (1988).
2 In re Detroit Auto Dealers Ass'n, Inc., 955 F.2d 457 (6th Cir. 1992), cert. denied, 113 S. Ct. 461 (1992).
3 National Macaroni Manufacturers Ass'n v. FTC, 345 F.2d 421 (7th Cir. 1965).
4 435 U.S. 679 (1978).
5 Id. at 695.
6 United States v. Brown University, 5 F.3d 358 (3d Cir. 1993).
7 California Retail Liquor Dealers Ass'n v. Midcal Aluminum, Inc., 445 U.S. 97 (1980).
8 See B.A.T. Industries, P.L.C., 5 Trade Reg. Rep. (CCH) ¶ 23,733 (FTC Apr. 19, 1995).
9 United States v. National Ass'n of Broadcasters, 1982-83 Trade Cas. (CCH) ¶ 70,845 (D.D.C. 1982).
10 Pub. L. No. 101-650, 104 Stat. 5127 (Dec. 1, 1990).
11 See, e.g., Allied Tube & Conduit Corp. v. Indian Head, Inc., 486 U.S. 492 (1988); American Society of Mechanical Engineers v. Hydrolevel Corp., 456 U.S. 556 (1982).
12 See, e.g., Allied Tube & Conduit Corp. v. Indian Head, Inc., 486 U.S. 492 (1988); American Society of Mechanical Engineers v. Hydrolevel Corp., 456 U.S. 556 (1982).
13 See Northwest Wholesale Stationers, Inc. v. Pacific Stationery & Printing Co., 472 U.S. 284 (1985) (exclusion from trade association that does not possess market power of exclusive access to an essential input does not constitute per se violation).
1. This paper has been prepared at the request of the National Telecommunication amd Information Administration for a report that the NTIA plans to publish and distribute to Administration officials, members of Congress, and the public.
2. Joseph Kattan is a partner in Morgan, Lewis & Bockius LLP in Washington, D.C., where he concentrates his practice in antitrust litigation and counseling. Mr. Kattan formerly headed the Office of Policy and Evaluation at the Federal Trade Commission's Bureau of Competition. Carl Shapiro is the Transamerica Professor of Business Strategy at the Haas School of Business, and Professor of Economics in the Department of Economics, at the University of California at Berkeley. Dr. Shapiro served recently as the Deputy Assistant Attorney General for Economics in the Antitrust Division of the U.S. Department of Justice.