Managing monopolies and dominance in the Net age

Guest blogger Mike Shatzkin is Founder and CEO of The Idea Logical Company, where he has focused on supply chain and digital change issues since 1979. Mike has spoken at and organized publishing industry conferences all over the world. He recently launched The Shatzkin Files blog. One of Mike’s several books, The Ballplayers, forms the core of

Our thinking about “monopoly” may need to be recast in the Internet age. This is a complicated question to consider and we need to start gathering some good minds around it.

Network effects were noticed before there was an Internet. Both the phone company and the electric company were networks, and it became clear about a century ago that everything worked better for everybody if they WERE monopolies and everybody was hooked up to the same network, not competing ones. So phones and electricity became regulated monopolies, with prices and other behavior, including mandated service levels, controlled. Whether because of a changing ethos or because things became more complicated, or both, “competition” has been introduced in both spheres over the past two or three decades. With debatable results.

Amazon’s dominance — which is not a monopoly but which certainly looks like unassailable hegemony in the world of online bookselling — can be largely attributed to brilliant execution and maintaining a tight focus on serving the customer. But part of their success at eliminating meaningful competition for online book sales has to do with the nature of the Internet. Online likes one winner in many spaces because it serves the users better NOT to fragment aggregations. If Amazon’s reader reviews were spread over 1000 web sites, they wouldn’t be as useful to the consumers. And their recommendation engine thrives on data; fewer customers would mean less helpful recommendations for those customers remaining, and the concentration at Amazon means less useful recommendations come from all their retailing competitors. This is an edge that may not stay with the retailer forever, though, because the playing field for information about books is being leveled by social networking sites. That’s why Amazon is investing in them.

This tendency to concentration makes it urgent for publishers to get into niches and start trying to own them while they have legacy advantages. If the history of the Net so far is any guide, each information and interest niche will end up being owned by a very small number of players; often it will boil down to one. We seem to have been pretty fortunate with the dominant players (perhaps we should call them “monopoly threats”) that have emerged so far, among them: Amazon, Google, ebay, Craigslist, wikipedia, and a now-emerging Facebook. They’ve executed well and kept their eye on the stakeholders they serve. They, so far, have been more benign dominators than were Microsoft and AOL, two big winners on the previous go-round.

But government is going to have to start thinking through public policy toward what will be an increasing number of dominant players in many niches. “Breaking them up” will seldom be a sensible answer: they often became monopolies because of natural causes and effects and if they were broken up, they’d recreate themselves (as the phone companies have been doing.) But leaving things to the market when there is no market is just trusting to luck.

The futility — or self-defeating nature — of “breaking them up” is obvious if we consider the possibility that Facebook or LinkedIn develops a dominant or near-monopoly position. The success of both sites, and their value to their users, is directly related to the number of people using them.

So breaking up a monopoly or dominant position built on the competitive advantage of a robust database tied to network effects would damage the users. That’s certainly not the objective for society trying to control monopolies.

One possibility is that we should allow combinations of suppliers or users when they are faced with a monopoly. Another is that stakeholders should be identified and, somehow, through mediation or legislation, their legitimate interests should be defined and somehow protected. But we have enough understanding of the Net and network effects now to see that our legacy thinking about monopolies, how to live with them and how and whether to manage them, may need some reconsideration. And we need some people who are NOT lawyers or economists joining those who are in the conversation.

A recent mailing list discussion about this produced a wide range of opinions. On one extreme is the notion that all market dominance is bound to be exploited and the government’s inability or unwillingness to control Microsoft is held up to have allowed damaging predatory behavior (pointing, first of all, to the demise of Lotus and OS/2.) On the other side are those who say the market and the rapid changes fostered by technology are all the control we need. That side too, can point to Microsoft and its current challenges as proof that no government hand is needed, even in the Internet age.

Certainly, one can see the possibility that Google will be challenged by vertical search (and just plain vertical sites) and that Amazon’s dominance in physical book distribution online will be hard to extent to ebook distribution (a topic I will explore further in an upcoming post on my own blog).

tags: , ,
  • Very interesting question, indeed. I’ve always had the feeling that the systems offered by the ‘big 4’ (amazon, facebook, ebay and google) are evolving according to user’s interest (see how useful and interesting are all the Gooogle ‘something’ services!). But I am also very concerned about the other side of the coin: we are so dependent of them, and they now so much about us! This gives them a huge power, hopefully not serving evil objectives!

    I see in interoperability and standards a technical response to that kind of concerns: users want an integrated front-end, they don’t care about underlying systems. For example, they find pretty useful to have a single ‘identity’ (e.g. login-password-freinds network combination), but they don’t require Facebook to be a monopolistic service provider to do that!

    So this could be a regulatory response to the potential evil use of internet monopolies: to obligate platform owners to be open to each others (through e.g. Open Search, Open Social, APIs, feeds, etc.). This way, users will be able to choose providers according to other criteria that the ‘critical mass’ effect.

  • Mike,
    There’s only one telephone cable running up the road to my house, and there’s a reasonable case that the cost of the ‘last mile’ justifies that monopoly versus multiple competing cables. If not the cable, then nobody would argue for multiple competing roads systems running up to my house. So there’s at least some ‘natural monopolies’ because the waste of scarce resources that competition would entail would be huge.

    But that one telephone cable (for which I am beholden to a unresponsive expensive monopoly) connects me to a network of a billion computers, and several billion telephones – a network which is not dominated or controlled by a single monstrous monopoly. It is provided by an uncountable number of companies that are both competing with others as well as co-operating – out of self-interest – to make the network work.

    The Internet monopolies that you are talking about have essentially zero demand for scarce physical resources. Competition would waste essentially nothing. There is no justification at all for them being ‘natural monopolies’ in the sense that this is for roads or the last mile of telephone cable. Bearing in mind that it is in the huge self-interest of the would-be monopolists have to encourage the idea that their monopoly is natural or inevitable, we should be extremely skeptical of the idea.

    Your point that networks effects and databases increase the propensity for monopolization is extremely valid. It’s not the first such effect. The economics software production and the sociology of software applications also tends towards monopolies – as so successfully exploited by Microsoft. Chip manufacturing has similar economics but, luckily, memory manufacturing didn’t fall into the hands of a monopoly. It is interesting to contrast the changes over the last quarter century in PC Memory versus PC operating systems. Driven by competition, the memory in a typical PC has increased in size by factor of several thousand while the price has dropped by by a factor of more than 10. At the same time, in the hands of a monopoly, the price of the Microsoft operating system has increased by a factor of about 20. Almost none of the current magnificent achievements of computing could have been achieved with that 10,000 improvement in price/performance of memory. I cannot even imagine what wonderful advances we have forgone because we didn’t competition driving similar improvement in PC operating systems.

    I must agree with you that, looking back at the effects of monopolies over the last century, attempts to artificially introduce competition after a monopoly is place are mixed at best, as are attempts to regulate them. Enron’s monopolization of the Californian Power Utilities in the name of “deregulation” has to be the best example of the worst. But looking back of that history, there’s nothing that seems appealing to say “let’s do it again”.

    Instead I look at the early days of the Internet – the days when the techies were in control. The Internet is the epitome of the “Network Effect” – and it includes databases (such as the DNS system) that are widely distributed and not owned or controlled by any monopoly while being completely integrated and transparent to the users. The Internet was largely built on what we would now call “Open Source” – which has proved the equal or better of the monopolistic software model – and, of course, the whole thing id founded on electronics and software – technologies with as close to zero replication costs as we are every likely to get. We have the technology and the models to follow to maintain real competition – not artificial competition or bureaucrats pretending to know what competition would produce – if we demand it.

    But we aren’t going to get it with the “pacifist” acquiescence to the internet monopolization that you espouse. We’re gonna get screwed … again!

  • Dave,

    I think we’ve established in other forums that we agree that you can’t effectively just break up monopolies; sometimes you had to figure out how to manage them.

    The extreme cases make the point and you didn’t deal with them.

    I will predict that LinkedIn, Facebook, and Flickr are going to become monopolies. And they’ll be monopolies of the most threatening kind (in a way) because they will harvest very personal data about just about everybody.

    But, of course, people will be there precisely BECAUSE everybody is there. And their positions will be much more entrenched than either Google’s or Amazon’s, because the switching penalties to any other choice will be much greater. You’ll be boxed in.

    So, what do you do about that is my question? And how do you distinguish the cases where you’re going to do that?


  • Mike, I wanted, in this forum, to be clear that the best way to deal with a monopoly is to stop it happening in the first place. That sets the expectation that there is no good answer to your valid and important question – how to deal with one that exists – (but that doesn’t stop us looking for the least worst answer).

    This is all about the dynamics of power. Once “they” have the power to act in their own interests and against “our” interests, there is, by definition, nothing “we” can do about it. (Whoever “we” and “they” are.). Once they have the power, their agreement is necessary for any change, and they are only going to agree to changes that further interests.

    There is no technical reason why Linkedln or Facebook should be closed systems. There is no reason why their databases should be physically a single database owned and controlled by a monopoly. There is technical reason why a social network should not be spread over a physical network with competing elements and providers, who cooperate out of self-interest.

    The websites of Facebook, Linkedln, Amazon, and Google are, in effect, “user agents” in the same sense that a Browser or an email client is a “user agent”. User agent software helps a human deal with a computer network. You have a choice of user agent that is not dictated by the website you wish to visit or source of the email you wish to receive.

    The websites of these monopolies have combined the user agent with the backend. I must use the Facebook user agent (website) to interact with the Facebook database. I must use the Amazon user agent to interact with the Amazon distribution system. I can’t use the Amazon user agent to order books through Ingram or B&N.

    The power of the WWW comes from that separation of user agent (the browser) from the backend (a website). Imagine what the web would be if Microsoft could have combined its user agent (windows) with its backend: a web in which 90% of people could only visit one website: Microsoft’s.

    So my answer to what to do with those monopolies is to force independence between the parts. Separate Amazon’s website from its distribution. Allow other websites to use its distribution and other distribution system to get orders from its website. To allow facebook’s website (and other others) to work with Facebook’s database (and others). To allow Google’s searching and ranking algorithms to work with competing algorithms and with other ‘user agents’. That is to say to allow modular competition.

    But that’s just saying that the best to fix a monopoly is not to have a monopoly. Competition is good for everyone – except the monopolist, but once they have the the power to be a monopoly, they are going to decide to be one and to remain one.

    So does it matter what we think we should do with monopolies — because it is not going to happen?

  • My purpose here was to start a conversation I think needs to happen; not to manage or dominate it because I’m not qualified to do so.

    It seems to me there are two problems with the solutions you are proposing. One is that hobbling a “prospective” monopoly requires prescience. What if you had “hobbled” MySpace before Facebook made clear that IT was the real monopoly threat?

    The other is that the Amazon and Google technologies were built with huge risk-laden investments. (Yes, I know: so where Microsoft’s and the railroads’. That’s a legitimate concern on both sides.) If entreprenurial investments are required to be open to competitors, what would that do to innovation? Why not just chuck copyright and patent law?

    This exchange only serves to confirm my notion that we have an increasingly challenging problem to think about that goes beyond my humble powers to add or detract. (It’s a good month for quoting Lincoln.) Doing nothing doesn’t seem right. What you’re proposing also doesn’t seem right. And trusting to the market where there is no market, as I said in the original post, is clearly not right.

  • Falafulu Fisi

    Hey folks, there is no such thing as monopoly, there is only a better competitor and worse ones in a free market. In fact the only monopoly that exists today is the government created entities, such as Post Office and that type of monopoly is unearned. The type of monopolies that Amazon, Microsoft and others had achieved are earned. Here is an excellent article on earned monopoly:

    Drop the Antitrust Case Against Microsoft

    Monopoly thru voluntary trade is earned and it is something to be applauded.

  • Mike -you initiative is welcome. I’m trying to help – especially as I’m a quite familiar with the sound of one hand clapping.

    Falafulu Fisi: it is hard to believe the author of the referenced op-ed piece (written in 2002) ever read the Antitrust Findings Of Fact (from 1999)

    The last line of it is:

    Microsoft has demonstrated that it will use its prodigious market power and immense profits to harm any firm that insists on pursuing initiatives that could intensify competition against one of Microsoft’s core products. Microsoft’s past success in hurting such companies and stifling innovation deters investment in technologies and businesses that exhibit the potential to threaten Microsoft. The ultimate result is that some innovations that would truly benefit consumers never occur for the sole reason that they do not coincide with Microsoft’s self-interest.

  • Mike,
    I think you are successfully defining conditions for a solution.
    1. It must not require prescience
    2. It must not hobble innovation

    I believe one satisfies #1 by applying the solution to all companies and circumstances. No one can predict which company is going to be successful (or a would-be monopoly).

    The economic theory of innovation is that a company that innovates has a monopoly because it is, by definition, doing something that nobody else is doing. It makes “extra” profits from that monopoly (Apple sells iPhones at high prices to the people who value being the first with a new toy). The extra profits attract competitors who imitate or otherwise compete with the innovation. Competition lowers the price (or volume) the innovator can get, so the reward for innovation declines when it is no longer an innovation. The company must find a new innovation if it wishes to reap more reward for innovation.

    Without competition the innovator can extract monopoly rent forever and has no incentive for more innovation.

    Patents and copyright extend the life span of the innovator’s monopoly. The effect is a trade-off between providing sufficient incentive to invest in trying to innovate and the length of time before there is an incentive for new innovation. There is no objective determination of the length of the extension, and Power is successfully using its power to arbitrarily extend it. Given that power is already making huge monopoly profits, the effect is to reduce innovation not to increase it.

    I remain skeptical that there is any solution to the monopoly problem. The only hint of a solution I have seen to set the incentive correctly. We cannot expect a company to do anything but maximize its self-interest. That’s it’s job.

    Instead of hobbling it, we (“society”) should arrange the incentive so that the behavior that maximizes their self-interest is the behavior that “society” wants. To do requires that a monopoly is NOT the most profitable wayto run the business. If businesses see the creation of a monopoly to LOSE money (perhaps because people will hate them or because the government will interfere or whatever), then they will go out of their way to avoid creating a monopoly. It will be in their self-interest to ensure that there is real competition — and I would expect them to be better at doing that than a government bureaucrat or me.

  • Yuhong Bao

    “To do requires that a monopoly is NOT the most profitable wayto run the business. If businesses see the creation of a monopoly to LOSE money (perhaps because people will hate them or because the government will interfere or whatever), then they will go out of their way to avoid creating a monopoly. It will be in their self-interest to ensure that there is real competition”
    You could try that, not that I’d recommend it though. In fact, I am not even sure that it is even possible.

  • Yuhong Bao

    Regulating the monopolies once they are created can be done however, and is certainly a good idea.
    “We cannot expect a company to do anything but maximize its self-interest.”
    Well, that is another can of worms altogether, a big part of which include the short-termism of Wall Street.

  • Yuhong Bao

    What the government also can do is to ensure that the monopoly is granted and defended leglimately, not by ways such as vendor lock-in.