IBM’s entry into the PC market can teach us a lot about the risks involved with the sort of structural separations recommended this week by the House antitrust subcommittee. 

On Tuesday, the House antitrust subcommittee issued its majority staff report and recommendations, produced after an extended investigation of digital marketplaces that was especially focused on Google, Facebook, Amazon and Apple. The report itself is a beast, clocking in at 449 pages—and we still don’t have the minority report—and will bear careful study.

The report makes recommendations in three categories: restoring competition in the digital economy; strengthening the antitrust laws; and strengthening antitrust enforcement. I think this grouping is telling, as it suggests that there are real boundaries to what antitrust can be used for and that we will likely need to look outside of antitrust if we are looking to address the issues that these companies raise. That further suggests a view that antitrust isn’t some infinitely pliable doctrine that can be bent and twisted into some broader mechanism of social or industrial planning.

And, of course, we need not do that. Within constitutional limits, Congress can pass whatever legislation it pleases, and we think of that as democracy in action. Congress is hardly constrained by the views of Sen. Sherman and however a prior Congress defined antitrust more than 130 years ago. The House antitrust subcommittee majority is calling for more direct regulation of these digital marketplaces. Its top two recommendations for new regulations focus on structural separations and line of business restrictions, as well as rules that prevent discrimination, favoritism, and self-preferencing. I think those are quite related and it would be useful to talk through a particular example to see how those ideas might actually operate.

I will note what is missing in the report: IBM. IBM does make a brief appearance in the report (p. 252), but only to help calibrate Amazon’s enormous market capitalization. From the report’s perspective, IBM has vanished from the scene. Given IBM’s once towering position in the computer marketplace, that is remarkable, but that is not actually my focus.

“Had we blocked IBM from entering the personal computer market, we would have lost a firm with enormous capabilities, and also the firm that seems to have been able to turn the personal computer market into a serious market.”

Instead, I want to focus on IBM’s entry into the personal computer market on August 12, 1981. The personal computer market was quite young and still unformed and it was uncertain as to where would go. Apple was an early entrant, as was Tandy and Commodore. But everything suggests that IBM’s entry into the market completely altered the trajectory of that market. IBM’s entry validated the market. Its sale of personal computers soared, and eventually the IBM PC became a platform on which the industry was based. As clones emerged, that market got away from IBM and it would never lead it again, but the actions of IBM built the market.

Replay exactly what would have happened had we had in place, at the time of IBM’s entry into the personal computer market, the report’s rules on structural separation and self-preferencing. Historical counterfactuals are just an exercise in intelligent guessing, but we can tease out how the contemplated restrictions would have blocked IBM’s entry. I take it that “IBM” as such—meaning IBM acting in its own name— could not have entered the personal computer market. There was no more valuable name in computing than IBM, and that name stood as a guarantee of technical competence and the resources required to ensure that a product would perform as promised. Any new computer company would have been decidedly at a disadvantage in competing with “IBM” in the new personal computers market. So have IBM enter the market not as “IBM” but as PCCO, a name as about as beige as the box that IBM released.

I don’t think that IBM could finance the new PCCO. The report (p. 378) says that “firms can use their supra-competitive profits from the markets they dominate to subsidize their entry into other markets.” IBM clearly was financing its entry into the personal computer market using its profits from mainframes. And I assume that we would have blocked IBM from using its vaunted sales force to help sell PCCO machines. Or perhaps we would have insisted that IBM sales distribute all personal computers—PCCOs and those of its competitors—on a neutral basis to avoid concerns about self-preferencing.

This is a vision of lowest-common-denominator entry: you identify the entry conditions for new firms to a market and you try to restrict the successful existing firms to compete on the same terms. I get why the firms that competed with IBM in the old days and that now compete with GAFA would find those types of restrictions attractive. They are focused on their own ability to make money and don’t otherwise have a direct stake in consumer interests and broader social issues. Those firms don’t care if new markets like the PC are stunted because we have blocked the best firms from competing.

The real question is whether that is good for society. The US has been noteworthy compared to Asia and Europe in building these great tech firms. Had we blocked IBM from entering the personal computer market, we would have lost a firm with enormous capabilities, and also the firm that seems to have been able to turn the personal computer market into a serious market. The tech world of today had flowed from that point of origin. 

Moreover, we have blocked entry before. The 1956 AT&T final judgment blocked AT&T from the commercial computer market at a point where it should have been a strong competitor. IBM ended up dominating the mainframe market. Perhaps that would have happened anyhow but the fact that perhaps its strongest natural competitor was sidelined may have played a role as well.

Entry is one of the most important events that happens in the economy, so the House antitrust subcommittee majority report is right to focus on these issues. The structural separation regime imposed on AT&T almost certainly shaped the mainframe computer market and would have changed the creation of the PC market. You should ask whether you know how those regimes would have played into other examples such as Apple’s entry into the MP3 player market with the iPod and the phone market with the iPhone. Again, Apple, even in its hobbled state in 2001, certainly had powerful advantages compared to many of the other wannabes in the MP3 player market.

We can play through each of these cases for GAFA one by one (and should) but I am not at all sure that the committee has learned the right lessons of history here.

Top photo by by Ruben de Rijcke, via Wikipedia [CC BY 3.0]