Three months after holding a hearing on Google’s search engine business practices, Senators Kohl and Lee have written a letter to FTC Chairman Leibowitz urging a thorough investigation of the company.  As anyone with even the remotest interest in the subject knows, the FTC has had such an investigation underway for some time now, and it is undoubtedly the most high-profile antitrust issue currently on the agency’s agenda.  Thus, the only purpose for such a letter would seem to be to apply political pressure on the agency for what is, essentially, an antitrust law enforcement matter.

Most worrisome, the letter contains hardly a mention of what is in consumers’ best interests, which should be the focus of antitrust enforcement.  Instead, while the Senators write it is not their intention to protect any specific competitor, their arguments are based on the complaints of several competitors who testified during a committee hearing they sponsored to hear those complaints.

We hope and trust that the FTC is undertaking a thorough investigation based on antitrust law as opposed to bowing to pressure from elected officials.  This will increase the likelihood of a result that is truly in the interest of consumers.

In theory, the antitrust laws do not penalize size, but it seems that virtually every firm that has become dominant in the technology sector—IBM, Microsoft, Intel, and now Google—ultimately becomes the subject of a major antitrust action.  The FTC started its investigation of Google formally last week and Paul Rubin and I wrote a piece on it that was published in Forbes.com.

We discuss the problems with antitrust action in high tech industries and, specifically, the nature of the complaints against Google:

Some websites are complaining that Google is manipulating its search results to advantage its own products and disadvantage its competitors. They want search to be “neutral.” But what does “search neutrality” mean? Does it mean that search engines should rank websites randomly?

Google’s market position was earned precisely because it found a way of ranking search results that is more useful for consumers, and it will quickly lose that position if someone can find an even better ranking algorithm. Before Google, the Web was much less useful precisely because search engines did not rank results in a way that consumers found informative. “Neutrality” could return us to that world.

Also problematic are the possible remedies the FTC could impose if it finds Google has violated antitrust law:

Google’s most valuable asset is its search algorithm, which is secret and constantly being refined. The secrecy of the algorithm is an integral part of its value because there is an entire industry trying to game it in order to achieve higher rankings. Would the FTC ask Google to reveal its algorithm so that the FTC lawyers and their technical advisors can try to determine how to make it neutral?

It is quite possible that the FTC investigation will not lead to further action because thus far there is no publicly available evidence that Google has violated the antitrust laws.  Let’s hope that the investigation doesn’t divert too much of Google’s attention and resources from what it should be doing—improving its current products and developing new ones.

The Department of Justice appears to be in the final stages of its review of Google’s acquisition of ITA Software.  Several travel sites, some (but not all) of which use ITA, oppose the deal.

Google is reportedly willing to honor ITA’s existing contracts with customers and to renew them.  Some of those customers who oppose the deal now want Google also to make available upgrades to the ITA software.  DOJ is reportedly considering challenging the deal if Google does not make such a commitment even though 6 other companies produce and market travel software.  In fact, the three biggest travel sites—Expedia, Travelocity, and Priceline—do not use ITA software.

Former DOJ chief economist Bruce Owen, in a recent paper, “Antitrust and Vertical Integration in ‘New Economy’ Industries,” prepared for a TPI conference, “Antitrust and the Dynamics of Competition in High Tech Industries,” found that the empirical evidence shows that vertical integration is generally welfare enhancing, even when market power is present.  This suggests there is a high bar for blocking an acquisition such as Google’s acquisition of ITA.

Owen made a more general observation that is perhaps even more noteworthy.  Merger reviews focus on the specific transaction under review, but the more important effect may be the signals that the authorities send about how they will view future transactions.  These signals are incorporated into the risk assessments and investment decisions of potential acquirers (e.g., Google) and acquirees (e.g., ITA).

By making it difficult for Google to improve its search engine product by incorporating better travel search features the government is sending a signal to large companies, particularly in the tech sector, that it is going to make it difficult for them to improve their products, at least by acquisition.  Other things equal, this reduces the potential acquirer’s value.

The signal sent to ITA and other potential acquirees is that it is going to be more difficult for them to be acquired.  Making a major exit strategy more difficult reduces the expected payoff to venture capitalists and other investors and, hence, their willingness to risk their capital.  Requiring Google to make its upgrades available to competitors would certainly diminish the value of ITA to Google.  Google might walk away from the deal altogether or go through with it and let DOJ sue.

Even if Google accepted the condition and closed the deal, being required to make software upgrades available to competitors would presumable reduce the incentives to upgrade.  Such a condition also raises the question of how the upgrades would be priced and whether the Justice Department would become involved in pricing decisions.  (This pricing issue arises even if Google commits only to make the existing product available.)

Some improvements might be Google-specific.  Would Google have to make those improvements available, possibly compromising proprietary information?  Would a court decide which improvements were specific to Google and which were more generally applicable?  Such determinations could easily turn into quite a mess.

The most important consideration for the antitrust authorities is the effect on consumers.  Enhanced travel search capabilities that are part of the Google search engine have the potential to produce significant benefits for consumers.  Those benefits, and broader benefits that could result from other tech acquisitions down the road, may be lost if DOJ kills the Google-ITA deal by putting too many conditions on it.

Steven Pearlstein writes in the December 14 issue of The Washington Post that the antitrust authorities should not allow Google to make any more acquisitions.  His argument seems to be that Google is too successful and too big and therefore we need to slow it down and allow others to catch up.

Pearlstein’s approach is at odds with antitrust economics and antitrust history and would likely harm both consumers and innovation.

Pearlstein notes that “It is worth remembering that aggressive enforcement of the antitrust laws has been a crucial part of the history of technological innovation in this country, enforcement that allowed AT&T to be supplanted by IBM, IBM by Microsoft and Microsoft by Google.” This is a nice sounding sentence, but there’s no evidence that it is true.

A recent paper by Robert Crandall of Brookings and Charles Jackson of The George Washington University prepared for the Technology Policy Institute’s project on Antitrust and the Dynamics of Competition in High-Tech Industries examined in detail the major high-tech monopolization cases of the last half of the twentieth century:  IBM, AT&T, and Microsoft.  In each case, “the ultimate source of major changes in the competitive landscape appears to have been innovation and new technology – technology that was apparently not unleashed by the antitrust litigation.”

• IBM lost its dominance because of the advent of personal computers, something the government never envisioned when it filed its suit in 1969.  Thirteen years later, the government asked the court to dismiss the suit because it was “without merit.”  Since then, IBM has transformed itself from a computer equipment manufacturer into the service company that it largely is today.
• AT&T was never supplanted by IBM.  The Justice Department’s antitrust decree and the FCC’s regulatory actions shaped the competitive landscape in telecommunications.  More important, however, was the development of high-speed Internet access and platform-based competition from wireless and cable TV, which occurred long after the consent decree.
• Finally, whatever the merits of the Microsoft case, it likely had little effect on the emergence of Google.  Microsoft remains dominant in desktop operating systems, but faces competitive threats from emerging technologies – smartphone operating systems, cloud computing, and virtual appliances.

Pearlstein suggests “it would probably be counterproductive to prevent Google from using its money and talent to expand into new areas.”  He believes, however, that the government should not allow “Google to buy its way into new markets and new technologies.”  In another paper prepared for the TPI project, Bruce Owen of Stanford (and former chief economist for the Antitrust Division) notes that while “the law distinguishes ‘organic’ (internal) integration from integration by acquisition…from an economic point of view there is no good reason for the distinction.”  He also points out that “Empirical evidence that vertical integration is harmful is weak, compared to evidence that vertical integration is beneficial – even in cases where market power appears to be present.”

Finally, according to Pearlstein, “by swooping in and buying these promising firms, Google forecloses on the possibility that they might be purchased by companies such as Microsoft or Facebook, which could use them to mount a serious challenge to Google’s dominant position.”  It would be news to learn that Microsoft and Facebook don’t have the financial wherewithal to compete with Google in the marketplace for acquisitions.  Indeed, such bidding wars can be very beneficial to innovation by increasing the return that entrepreneurs realize on their investment.  The $750 million price that Google paid for AdMob resulted from a bidding war between Google and Apple.  An antitrust policy that takes the biggest players out of the market for small companies would reduce the potential returns to innovation and therefore would reduce the amount of innovation.  Competition to acquire innovative new companies is a major aspect of competition that the antitrust authorities should take into account when deciding whether to challenge these acquisitions.

It has been the practice of some prominent Silicon Valley companies (and perhaps other companies as well) to refrain from “cold calling” employees of companies with whom they collaborate.  This seems like it could be a reasonable practice since companies may be less willing to enter into beneficial collaborations if they think their best employees can easily be stolen away.  In any event, the companies never agreed to refrain from talking to employees who expressed some interest in changing jobs (that would not be a cold call) or to refrain from hiring any of their collaborators’ employees, so the effect of these cold-call agreements was likely small.

Nevertheless, the Antitrust Division of the Department of Justice opened an investigation because it was concerned that the do-not-cold-call agreements were anticompetitive and had the effect of depressing wages.  That investigation is now complete and six major companies—Adobe, Apple, Google, Intel, Intuit, and Pixar—have settled with the DOJ and essentially agreed not to  enter into do-not-cold-call agreements for a period of five years.

We don’t know whether this is a good or bad settlement, but neither does the Division, because it didn’t look at the economic consequences of the do-not-cold-call agreements or of the settlement on wages or on competition.  Instead of looking at the evidence, which the Division would have done had it evaluated the case under a “rule of reason,” the Division viewed these agreements are per se violations of the antitrust laws.

These agreements could be anti-competitive, but they may also could be pro-competitive, for at least two reasons:  They remove a potential impediment to pro-competitive collaborations between companies.  They may also serve as some balance to California’s weak do-not-compete legal regime.  (Strong do-not-compete employment provisions are not enforceable in California).

The Division should have viewed this as a “rule of reason” case and evaluated both the potential pro- and anti-competitive effects of do-not-cold-call agreements.  Expanding the scope of per se antitrust violations in the absence of analysis can interfere with productive collaboration, efficient employment practices, and, ultimately, innovation in these critically important high-tech industries.