IBM, Microsoft and Big Tech Antitrust Folly

From The Wall Street Journal:

The continuing trial of Google, along with lawsuits against Amazon and Meta, have brought antitrust back into the public eye. These suits recall the 1969 case against IBM and the 1998 case against Microsoft, the great antitrust battles of the latter half of the 20th century.

Supporters of aggressive antitrust enforcement think that only antitrust suits prevented IBM from commandeering the personal-computer market and Microsoft from taking over the internet. But that’s an urban legend.

Historical evidence rebuts the claim that the antitrust suit forced IBM to stop bundling application software with its machines, jump-starting the modern software industry. As early as 1966, IBM had already made the decision to unbundle independent of the suit because it could no longer provide the variety of software that users demanded.

In the 1960s, IBM led the mainframe computer industry by offering a product that consumers valued for its technical quality and complementary products, coupled with IBM’s customer service. Does that mean IBM would have swallowed the personal-computer industry if the government hadn’t stopped it?

No. IBM was aggressively developing its PC while the trial dragged on. To create it, IBM had given a unit in Boca Raton, Fla., the autonomy of a startup. The company introduced its first PC in August 1981. The machines were selling briskly by the time the Justice Department dropped its antitrust suit in January 1982.

Yet the company’s mainframe computer business proved a liability in the new market. Existing divisions fought for control of the PC, and executives quickly eliminated the autonomy of the PC unit, assigning the new product to legacy divisions. A legion of agile “clone” makers quickly wrested the market away from IBM forever.

IBM had considered the PC a minor complement to mainframe computers, but the PC ended up killing off the traditional mainframe. The once-dominant company lost $8 billion in the second quarter of 1993.

Microsoft’s story is similar. The company also created a unit with considerable autonomy to create its browser, Internet Explorer. The company pressured its operating-system customers to adopt Explorer and engaged in other contracting practices that would be subjected to sanctions in the final antitrust judgment. By 2001, Internet Explorer had vanquished Netscape Navigator in the “browser war.”

Yet far from exploiting the internet, Microsoft managed its browser as a complement to its operating system, the company’s cash cow. Careful historical scholarship suggests that the company did so not because of the antitrust suit but because of many of the same internal forces that hindered IBM’s PC.

Microsoft disbanded its autonomous browser unit and assigned the new technology to its legacy divisions. The effort to dethrone Navigator had always been motivated by the fear that a browser could replace Microsoft’s operating system, and an independent Internet Explorer unit would threaten the operating-system business in the same way Netscape had.

Microsoft had a brighter future than IBM and continued to dominate the browser market for a few years, but Google, a company not burdened with existing assets and capabilities, would exploit the internet more fully.

If there was an antitrust case from the late 20th century that might have had dramatic consequences for technology, it was the long-running suit against AT&T, which resulted in the breakup of the telephone giant in 1982. Unlike today’s targets, however, AT&T was a regulated monopoly. Its breakup was an act of deregulation in the name of antitrust.

Link to the rest at The Wall Street Journal (Sorry if you encounter a paywall)

As the OP mentioned, in 1969, the government launched an antitrust suit against IBM. Despite a huge court tussle, the suit was finally dismissed by the Reagan administration in 1982.

We don’t know what would have happened had the ’69 antitrust suit had not been filed and consumed a huge amount of time of many smart people at IBM, but by 1982, Microsoft was on a roll with PC-DOS and the momentum toward the personal computer was well underway.

Clones of PC-DOS lead by MS-DOS from Microsoft and clones of IBM’s PC personal computers proliferated like rabbits. Compaq, founded in 1982, became the first major producer of PC clones by 1984 and sold larger businesses very effectively.

Radio Shack’s Tandy PC clones proliferated through about 8,000 store locations in urban and small-town locations across the US, UK, Canada and Australia in the 1990’s, aggressively driving the cost of personal computers down so more families and small businesses could afford them.

16 thoughts on “IBM, Microsoft and Big Tech Antitrust Folly”

  1. Go back to the arguments of the Microsoft trial and every single argument against MS was wrong and 20 years of history have proven Billy Boy Gates 100% in every argument: OS market share is meaningless in the internet age–what matters is hardware and apps; the dominant platform is the internet itself, so an OS without a bundled browser is worthless (even then, Apple, IBM, and Unix workstations all came with a browser in the box) and desktop hardware is now third in importance behind mobile and servers. (Gaming boxes, the target of the Wrath of Kahn, are fourth. Way to go after what “really” matters. The Luddites at the UK CMA at least understand that it is the datacenters that bear watching today.)

    The true history of that lawsuit is too ugly for a family friendly forum but the biggest outcome is that a company that, as a company, was (naively) apolitical with just *one* lobbyist lawyer in DC, is now a Billion dollar a year political powerhouse that can draft senators and congrrsssionsl representatives of *both* parties to set up committee meetings to investigate and reduce funding for their regulatory foes.

    The other (law of unintended consequences!) was the death of Windows CE, the *open* PDA, tablet, and cellphone Operating System. Which opened the door for the Nintendo model of lockdown hardware that Apple and later brought to mobile. And thus begat the current antitrust handwringing over mobile appstores. Note that today, Windows temains an open platform and even the oficial MS store is a distnt second to STEAM. Most people use neither snd just buy their apps developer from the developer, who gets 100% of the money, not 70% or less.

    Long before iphone, but not Nokia, windowsCE powered PDA/CELLPHONES:

    The hanging judge found no justification to break up MS, but hey!, he branded it a monopolist and allowed the politically connected CEOs that ran their companies into the ground to sue MS. That was a win…for them. MS promptly raised consumer OS prices by 30%, matching Apple’s annual $130 “upgrades”. (And people fear that Windows 12 will require a subscription! Whole ‘nother FUD story.) OEM prices stayed the same but corporate user licenses went up. At no time did their revenues or profits ever go down.

    And, ATT? A whole other dumpster fire that caused big disruption, wasted billions, and achieved nothing. Beyond destroying LUCENT, nee Bell Labs. Another time.

    • This is certainly one way to look at it. The “facts” are so murky that it’s almost impossible to say who’s “right.”

      I’d argue instead that the biggest screwup was when AT&T — fifteen years before the first release of Windows stable enough to run a business computer (that is, 3.1) — essentially forced the demise of Unix by not only demanding license fees, but sending cease-and-desist letters to anyone who tried to publish or publicize a substitute user manual. But then, AT&T was under significant antitrust scrutiny at the time itself (more successful — so to speak — than that directed at IBM†), so perhaps irrational behavior was only to be expected.

      † Trivial challenge: Trace the ancestry of Verizon back to the beginning. Without raising your eyebrows at what you discover.

      • One could argue it doesn’t matter.
        We ended up where we started, no?
        What exactly did breaking up AT&T achieve?
        (That market forces wouldn’t have achieved on their own?)

        IBM wasn’t broken up by lawyers but they (unknowingly) broke themselves up soon enough. Don Estridge was their last best hope and once they lost him they were played out. Like dinos they didn’t realize they were done for a decade. A zombie corp.

        The name endures but the relevance didn’t.

  2. But, but, but, now I want to hear PG’s history of his own use of computers over the years. I’ll go first:

    – Commodore VIC-20 when I was just going into high school (with cassette drive and BASIC expansion cartridge);
    – Commodore PETs at school (could program in BASIC);
    – ICON computers (GUI + centralized storage for a network);
    – IBM XT at work in 1987, bought IBM XT for home — programmed whole shell system in BASIC and DOS;
    – Mainframe programming for BASIC, COBOL, FORTRAN in ’87-88 at university;
    – 80486 computer in early 90s;
    – PCs, MACs and MAINFRAMES at grad school;
    – DOS based systems at first “big job” in government in ’93;
    – Switched to full Windows at work in ’93;
    – PC with Win 3.1 at home until late 90s;
    – PC with Windows ’95 and ’98
    – Laptops, netbooks, multiple PCs with each iteration of Windows releases

    I missed punch cards at university by a couple of years.


    • I got to use punch cards my first year at the day job. Fortran IV for the win.
      (We had a summer intern who sneered at Fortran. “Pascal is the wave of the future!” Never mind the millions of lines of proven, useful F4 code. Productive tools, sunk costs. Not even F77 was justifiable.)
      I not only got to use a vintage Tektronix terminal, I got to program for it. Vector graphics!
      I’m sure some folks remember the IBM PC/3270 but how about the PC AT/370? With BERNOULLI cartridges for classified work. Actually faster turnaround than the timeshare mainframes. IBM and CRAY both.

      The branch office had a Wang vintage word processor and somebody told the boss I had a home PC (when the average price of a new one was still $1000). Higher ups were pushing for a full refresh to the latest. He had doubts. I pointed out the money wasn’t coming out of his budget, but he could buy a bunch of PCs and copies of WANG software and be okay but Word Perfect, Wordstar, and MS WORD were better and cheaper. A a $400 piece of software could convert Wang format. The PTB, bought the Wangs, he bought a couple of PCs for the AAs. Then some genius running facilities decided to save a few bucks by shutting off all the building AC’s over labor day weekend..Tuesday, all the Wangs were dead. Our PCs were fine. Within a year everybody started moving to PC. We got there first and best: where other execs got the first PCs in their groups as desk decor, our boss got the last, after our staff was covered.

      Lots more stories from that transition era.
      I never did learn to keep my mouth shut.
      The centralized IT guys hated me.
      But the boss found me useful and he controlled my psycheck. 😀

    • I did not miss punch cards. And before you can ponder whether Fortran IV or Fortran 77 was better, before you can argue about Pascal or PL/1 (or, horror of horrors, Ada), you’ve gotta master JCL. Of course, they’re all inferior to a properly-constructed pure-assembler program (presuming you don’t erroneously try to run a program written for an 8086 on an 8088).

      I’ve got soldering scars from my first “personal” computer — an Altair 5000, run over a surplus ADM3 (not ADM3a) terminal that had at one time graced the Pacific Northwest Bell billing center a few miles away.

      I’m afraid I’m… old.

      • Hot solder on a foot is no fun.
        (I did manage to connect a cherry mechanical keyboard to an ATARI 400 by following a Byte article.)
        Saved a good chunk of cash, though, even if I walked funny for a week.

        Alas, in the end, the sixes lost the war to the eights.

  3. How much of the fall from dominance was due to the fact that IBM and Microsoft backed down from the worst of their monopoly abuse as a result of the lawsuits?

    But both IBM and Microsoft (to a lesser extent) did change over time to play nicer with others, but only after a change in leadership (two generations worth at Microsoft, I wasn’t watching IBM that closely, I know of one generation of change there, there may have been a second)

    • I think there is no question that both IBM and MS backed off on some of their behavior due to the suit.

      In an alternative history absent the feds shaking their antitrust swords, I don’t know that either company would have been much bigger today than they already are. I don’t think the clone-makers would have shut down their operations and that workalike software would have continued to proliferate.

      For me, the only significant casualty of MS tactics during this period was WordPerfect (a moment of silence, please). Once Alan Ashton and Bruce Bastian sold the company to Novell, the handwriting was on the wall.

      WordPerfect and Novell couldn’t have been more disparate organizations. I don’t have the slightest sense whether Novell ever changed, but it was a dog-eat-dog work environment at the time of the sale.

      When Ashton and Bastian signalled they were leaving, all the talent in WP headed for the door as well. That exodus of people with brains and some money fueled a mini-Silicon Valley in the area which is still healthy.

      • Word Perfect blew their lead the same way as Lotus, dBase, and Informix Wingz: they had a best of breed product but did nothing to improve it for years, instead spending all their time porting their one product everywhere: every flavor of Unix, OS/2, mainframes, Amiga, Atari. Even NeXT!
        (Did WP *really* need to run on VAXes when PCs were eating up that msrket?)

        During those same years, MS “divorced” IBM, Gates went everywhere, literally begging them to do native WINDOWS versions. And getting rebuffed. So he had no choice but to do it himself, like THANOS. Also at the same time, the memory chip cartel propped up RAM prices, making the much cheaper 386sx PCs market dominant and with them, Windows 3.1. And with both, as *DELL* suggested, MS OFFICE.

        All downhill after that.

        • To be clear, Dell suggested bundling Office, with their PCs and other clone vendors followed, which took it everywhere “for free”. It was not Microsoft’s idea. Just like pricing ENCARTA at $99 wasn’t their idea, but EGGHEAD’S.

          But that is s different story and lesson. 😉

      • If Microsoft had been able to continue to get all PC manufacturers to pay for windows licenses for all systems, and made the Internet browser definition ‘whatever Internet Explorer does is’ we would not have the standardization that allows for people to use mobile devices, chromebooks, Linux and Macs for real work and made the server side independent of the client side.

        They were well on the way of tying it all together, and we’re only a few years away from when all websites had to have special IE6 work-arounds because of them not following standards.

        • Not all PC vendors paid for a Windows license for every PC they shipped, just the ones who found it *cheaper* than keeping track of every single unit sold, and their configuration, and only paying for those that shipped with Windows. Their idea.

          (Dell tried separate SKUs for Linux boxes vs Windows but sales were so low it didn’t justify the process: keeping track of the SKUs cost more. Plus quality control required booting up each system. That meant two pipelines. Double the QA cost. Then there was the matter of user support. For Linux they had to do it inhouse and it cost more than the license to have 24×7 employees on call. After a while they stopped selling Linux and simply sold OS-less boxes. For $10 *more* than with Windows. Quality control still required booting up the Linux boxes and then wiping the hard drive. Yet again, it was the OEM who chose to pay for licenses they didn’t use.)

          As to IE being non-standard, that wasn’t exactly true. They complied with every committee spec, once ratified, and often *before*. That is how IE5 killed Netscape. People forget that up to HTML4, Netscape ran ahead of the standards body introducing tags and scripting commands as they saw fit. Up to HTML 4.2 all the committee did was rubber stamp the defacto standard, Netscape, and everybody was forced to follow. Then Netscape management got “standards religion” and stopped paying attention to user needs and innovating to make site building easier.
          IE didn’t stop, they made IE faster by tweaking the OS to execute HTML faster (Apple did the same) by moving the render engine down from the app level to the OS execution level. (This opened up all sorts of security exploits, a problem for later.) This integration is what the “trustbusters” latched onto.

          At the app level, MS simply followed the long dragged out committee work on HTML 5 and coded the new tags ahead of ratification whereas Netscape waited to start coding. (Cheaper, I suppose.) Once the spec was signed off on, IE5 was ready for release and site builders wanting to use the new spec only had IE availsble for months.

          The IE 5.x extension after that were, of course, meant to take market share by being more useful and, yes, “suck up all the air”. That is competition. In the end Netscape killed itself by taking their eye off the ball–user needs–and following standards instead of making them. (The same way is how NT ate the Unix workstation market: proprietary is always faster than committee specs.) And that is why the hanging judge had to rule MS didn’t harm Netscape, only *tried*.

          MS tried to kill Netscape but beat them to punch.

          Afterwards, Ballmer also got cheap, cut back on IE staff, and opened the door for CHROME and Google took over.

          Finally, in the lattest example of stupid executives at work, Google too got cheap and open sourced Chrome. MS hss sinced become the largest singlr code contributor to CHROMIUM outside Google and arguably the biggest contributor of *new* features and better, leaner code. Little by little they are taking back control of browser evolution snd becoming more ” Edgy”. 😉

          Today, Edge is faster, leaner, more stable, and more feature rich than Chrome which is today’s “bloated resource hog” by comparison.

          (Note that pure Chromium brats Chrome which is based on it.)

          Competition never ends and under Nadella Microsoft’s hypercompetitiveness is alive and well, but with stillettos instead of sledgehammers.

          The game is still on.
          Most folks just don’t notice.

        • David, during the period from 2001 to 2013 there’s one class of users for whom IE was not only a standard, but the only option: Anyone interfacing with the student loan system. Students and parents. School administrators. Banks. Loan servicers.

          Because critical elements of the program (primarily, but not only, the entire user-data-internal-firewall-and-verification subsystem) ran in ActiveX only. It was impossible to use a non-IE browser because the server-side systems checked for an IE fingerprint and rejected the session as non-secure if they didn’t find it.

          All because a downstream vendor — not M$ — that created the system under contract for the Department of Education built that in. Careless? Intentional? Nobody’s talking.

    • Uh, MS is far from a kinder gentler player.
      They’re just subtler. And sneakier.
      You haven’t noticed how they play the media and politicians? Play along with whatever is trendy:

      Go to their executive pages and you’ll find pictures, names, titles, and…pronouns.
      (Most all bring his/her CIS humans.
      This week? After the Gaza attack on Israel, Nadella issued a statement, straight to the point:

      “There was a terrorist attack by Hamas on innocent citizens of Israel, and that has to be condemned in the strongest possible ways,” Nadella said during an interview with Axel Springer CEO Mathias Döpfner. “Our heart goes out to them and there should be no ambiguity, right? There is no room for violence of that form.”

      Short and sweet as an answer to a question in a prearranged interview.

      Try these two from the recent FTC trial:

      1- Sony paid for timed exclusivity for two ZENIMAX games for the launch of the current generation of consoles. They also sniffed around a new rpg BETHESDA SOFTWORKS (a ZENIMAC subsidiary) was working on. Now all Bethesda RPGs on console for 20 years, have launched and run best on XBOX (the last one, SKYRIM has sold 60 million and counting) so Sony buying the next one away from MS would really have hurt. MS, being on good terms with Bethesda found out. Nadella called up ZENIMAX and asked for a number. $7.5B? Deal!
      And the new game? XBOX exclusive. Forever. It launched las month, bigger than SKYRIM, by far the top seller of the month, instantly the 7th best seller of the year.
      Gates would be proud.

      2- Around the same time Sony was screwing up the new XBOX console launch, ACTIVISION was putting the squeeze on MS, saying that since Sony outsold XBOX 2-1 last console generation, they wanted a bigger cut of CALL OF DUTY sales, 80% instead of the usual 70%. No choice, right? Well, two years later, ACTIVISION was me-too’ed and hit with a California Labor lawsuit. Most everybody in the tech world stayed quiet. They might be next. MS didn’t. The released a press release saying they were appalled at the charges and would be reevaluating their business relationship with Activision. Next dsy, Nintendo followed suit. The stock dropped 30% in days, with no end in sight. A few weeks later, Activision eagerly agreed to sell lock, stock, and CALL OF DUTY to MS for $69.7B. They had called everybody from Sony to Amazon, Apple and Netflix. All were willingto buy parts, usually for stock. MS offered all cash at pre-scandal price. Two years later, the biggest tech deal ever was done.

      Or this:

      3- Last fall, OpenAI released ChatGPT to great acclaim. It soon came out that MS paid $10B for 49% of any profits and full access to the tech. Remember how Internet Explorer came to be? (MS licensed the MOSAIC source code from Spyglass for a one time payment and a percentage of browser sales revenue. But since IE was free, there was no revenue to pay. Mosaic itself was developed at taxpayer expense at MCSA anyway.) At last word, OpenAI is working frantically to monetize GPT while MS is rolling the tech into all their existing products. Oh, and GPT apps only run on AZURE.

      OpenAI may of not survive but MS has already made their investment back.

      Ever watch a good caper movie or show?
      Fun to watch a “mastermind” at work.
      That’s today’s MS, two steps ahead of competitors and softselling their agenda.

      They’re still as hypercompetitive as ever; they’re just less blatant. Slippery.
      Maybe even smarter. But kinder, gentler, “softer”?

Comments are closed.