Last modified: 01 July 2017
I like folk music. I’ve got a lot of Kate Rusby albums. I can hear the Moseley Folk Festival from my window, and yes, I’ve been. I might even go again this year. So clearly I’m susceptible to folk tales.
So did I fall for a folk tale about Microsoft saving Apple (who then went on to create the iPhone and the WebKit browser engine that powers Chrome)? I don’t think so.
In fact quite the opposite. In responding to my original piece , Sam Bowman at the ASI has found someone who seems to be the tech equivalent of Another Angry Blog and has fallen for it like a born-again Corbynista. I'm going to try and snap him out of it, I hope I am more succesful than with my increasingly left-wing friends.
Dilger, whose piece Sam’s is heavily based on, is clearly a big fan of Apple Computer, and no fan of anyone who criticises them. He repeatedly calls the BBC’s Bill Thompson a liar in a rant that would surely include the phrases mainstream media and fake news if it was written today. In Bill’s excellent response , this is shown to be ridiculous.
Dilger holds back even stronger insults for Paul Thurrott, one of the world’s most-respected tech journalists (and no fan of US and EU anti-trust rulings himself). In the article on which much of Sam’s response is based he calls Thurrott an “ultra right-wing style absolutist fundamentalism with a thick, campy layer of extreme left-wing liberal faux-outrage”.
There are warnings signs all over Dilger’s blog that it should be treated with extreme caution. It reads more like the work of a Mac fanboi than a calm and reasonable account of what happened.
But just as it’s important to respond to the substance of articles on Another Angry Blog or The Canary, so I think that it is important to reply to the content here. And I see four really important points in Sam’s blog.
Sam is right to say that the $150m didn’t really matter in saving Apple. It was the guarantees to develop Office and Internet Explorer on the Mac that mattered. That’s obvious just from the change in share price; Apple’s value increased by much more than $150m. And it is confirmed in sworn testimony from Apple’s Chief Financial Officer .
But this is not “the opposite” of anything that I wrote. Microsoft was aiming to secure its monopoly in office apps and to develop a monopoly on the internet. I say this throughout my piece. And at the same time it was aiming to stop itself from becoming a monopoly in the operating systems market in order to improve its position in a number of anti-trust suits.
It’s very important to note that the anti-trust proceedings ongoing against Microsoft in 1997 were restricted to claims that it was abusing its monopoly of the operating system market to gain an anti-competitive advantage and establish a monopoly in new areas. Microsoft did not have a monopoly of the browser market in 1997 (Netscape was still the larger browser) so there was no monopoly to abuse. Microsoft’s abuse of browser monopoly with Active-X came later.
As Sam says, in his ruling in 1999 the judge says that Microsoft is effectively a monopoly, even though Apple still exists. But it’s not logical to reason from this that Microsoft’s motivation for saving Apple wasn’t to try and convince a judge that it didn’t have a monopoly. It argued this at length in the trial, it just failed to convince the judge.
It seems that Microsoft actually came quite close to convincing the judge too. In section III.2.b of the ruling, the judge explains that it is the disparity in the quality and number of applications that is a large factor in holding back Apple’s OS from competing with Windows.
That is why in saving Apple, Microsoft also committed to addressing this lack of applications in two important areas; office and browser. I don’t believe that this was a coincidence.
I remain convinced that Microsoft saved Apple in large part to increase the likelihood of successfully arguing that Windows was not a monopoly operating system. Their plan may have failed in this instance, but that doesn’t matter.
And I think that there’s lots of good evidence for my belief elsewhere. Microsoft has frequently used the argument “that other operating system vendors are doing exactly the same as it does” in court cases. Point 907 in EC Case T-201/04 is one example, but there are many more. If you’re going to use that argument, and if your legal team keeps thinking that it will work in case after case, it’s in your interest to have other operating system vendors who are doing exactly the same thing. Bundling an Internet Browser with their Operating System for example, or writing specific APIs in their Operating System to make Microsoft Office work better. And that’s exactly what Microsoft ensured by saving Apple.
Microsoft saved Apple, who later developed the iPhone and WebKit (that became Google Chrome), to try and avoid punishment from anti-trust courts.
I’ve read the paper that Sam suggested and there are two massive problems.
The first is that it was published in 1998, before the antitrust case in question was decided, using data ending in 1997, before Microsoft saved Apple. I never claim, nor does anyone that I’ve ever discussed this with, that anti-trust prosecutions had instant impacts in this case. Indeed my whole argument is that anti-trust action in the 90s, and the resulting care that Microsoft took not to crush competitors in the future, is what increased the probability of the development of the iPhone, the success of Chrome, and the erosion of multiple Microsoft monopolies a decade later.
In my first piece I show in detail that anti-trust actions in the 90s failed to stop Microsoft from achieving a monopoly in office applications in that period. This paper agrees with that.
My second issue with the paper is one that I’m much less sure about, but I think might be important. The paper looks at a very large number of events, most of them tiny. And then it argues that if Microsoft’s share price went up in response to news on an antitrust suit, its competitors should have gone down, and vice-versa. It finds the opposite, including in the PC-Software category where it lists 22 companies.
But I don’t follow the logic at all. Of these 22 companies, only Oracle, Novell, Borland, and Lotus were likely to benefit from a more competitive marketplace, as these were the only companies with products that Microsoft was threatening. For other companies, such as Adobe and Macromedia, where neither Microsoft nor Apple had any hope of competing, a strengthening in Microsoft’s positions would also strengthen their position, at least in the short term, by reducing the likelihood of greater development costs (on multiple products) to meet a similar-sized market.
Developers like me love to develop for a single monopoly operating system run by a company scared into good behaviour. It's much easier than developing for multiple incompatible operating systems that are constantly innovating, changing, and breaking each other's features.
The last point is the most interesting.
Lawrence Lessig is a very clever guy who understands technology and the legal barriers and promoters of innovation really well. I’ve long listened to his talks and he helped to found my favourite podcast series “Triangulation” – run by the same company as This Week in Google which I referred to in my last piece. And importantly he was involved in one of the largest US antitrust cases against Microsoft and, yes, as Sam notes he has written since that "I blew it on Microsoft" .
But if you read the piece it is clear that Lessig’s headline sounds a lot more confident than he really is. Lessig is heralding the miracle of Linux, an alternative to Windows’ dominance that he never considered. But in 2007, when the piece was published, Windows held an even more dominant position on the desktop than a decade earlier. Mac market share was lower than in 1997, and Linux was in the same place as it is today, at about 2%.
So why does Lessig call Linux a miracle?
I think that the purpose of his piece is to try and win people like Sam over to an idea that they would otherwise hate. Faced with a new challenge of monopoly internet access to many homes in the USA, Lessig is proposing an alternative to fines, separations, and slow legal proceedings. And with attentions focused on less state interference he swiftly proposes state-run internet access to compete with private broadband providers.
To get a deeper insight into Lessig’s thoughts and motivations I’d recommend listening to this talk by him from the same year on whether “Google is the new Microsoft” . In it he repeats his belief that the legal route is best avoided, while also talking about the huge costs of anti-trust lawsuits to Microsoft, the impetus that it gave to companies to support a wider ecosystem, and the deep changes in corporate culture that it led to at Microsoft.
When Lessig wrote his piece at the start of 2007, the iPhone had not yet happened. Android had not yet happened. Google had not yet released Chrome. The legal process that he joined in with had not yet born any fruit. After a decade, I don’t blame him for calling it a failure.
And yet just two years later everything had changed. The operating system market that he helped intervene in was thriving and competitive, with the USA leading the world.
And today, a decade after Lessig advocated not intervening in fixed broadband and proposed municipal broadband instead, the government still blocks municipal broadband to protect markets. Markets which, in the absence of intervention, still offer only one choice to many Americans, who in turn endure poor service and high costs.
Last but not least, I want to pick up on the point I made that I think was least well understood. My position is that anti-trust intervention can be bad (US interventions against Xerox photocopiers and Kodak film spring to mind and I’m never sure about the UK splitting Sky and Sky Bet). But I think that interventions can also be good, and that often we can’t tell for a long time which they were.
I fully accept that we can never know the counterfactual to intervention. Without intervention Microsoft could have created a complete monopoly, sat back, and collapsed earlier; leaving a vibrant, thriving market much earlier than we got one. We can’t know, and there’ll always be a reasonable argument that we should wait and see.
So I want to give a couple of examples of what damage can be done if we wait and see, and keep waiting, and the alternative never arrives.
On twitter I mentioned buses. In almost every UK city a single bus company has a near-monopoly that they abuse. For thirty years we have failed to tackle this in the UK, except in London which serves as a perfect control for a natural experiment. The result is absolutely clear. Bus use in London has doubled while fares have stayed low. While outside of London bus use has halved and fares have rocketed, leaving our cities congested and unproductive.
I think that software is quite similar to buses, but just in case you don’t, here’s another example to think about.
In 2015 the richest man in the world was Bill Gates, in part from the fortunes made thanks to Microsoft monopolies. The second richest man was Carlos Slim, in part from the fortunes made thanks to his near-monopoly control of telephone and mobile telephones in Mexico.
América Móvil and Telcel have a near monopoly (over three quarters of mobile subscriptions in Mexico). Those who oppose intervention keep saying that Mexico should wait for the counterfactual — the unknown innovation or change in technology that will fix everything.
But so far the change doesn’t seem to be coming and the result is clear. Mexico has the lowest rate of mobile phone subscriptions of large countries in South America. Lower than Peru, lower than Bolivia, lower even than Venezuela.
Maybe it’s worth waiting a bit longer in Mexico, I don’t think that we can ever know. But I think that we need to be open to consider caling time on waiting. The impact from Microsoft anti-trust proceedings, even with their huge costs, are why I think that action is sometimes justified.