Network Feedback? "Them that's got shall get, them that's not shall lose, so the Bible says, and still it's news." - Billie Holiday, God Bless the Child It seems likely that Billie Holiday was referring to Matthew 13:12. Jesus had been in a boat in the Sea of Galilee, speaking to a crowd that had gathered on the beach. After Jesus told the crowd the parable of the seeds, the disciples asked him why he preached so obliquely. He replied that the disciples were chosen to know the secrets of heaven, but the people in the crowd weren't. "For to him who has will more be given, and he will have in abundance; but from him who has not, even what he has will be taken away." Similarly (but darker, perhaps) consider this passage from verse #53 from the Tao Te' Ching: "While the Court is magnificent, the fields have become barren."

       Lady Day's description of network feedback dynamics (at right) is indeed still news and it supports Shapiro and Varian's contention that network economies operate under long recognized principles (IR-1). But why do the strong get more, while the weak ones fade?

       Because information products are cheap to reproduce (IR-3) and are amenable to network based distribution, the economics of those networks become a driving force in the overall performance of the digital information economy. Hal Varian says, "Economists have a fairly narrow definition of a network in this context. We talk about network externalities."

       "So, think of a fax machine. The value of a fax machine depends on how many other fax machines there are. Think of e-mail. The value of e-mail to me depends on how many other people who I correspond with have e-mail." There's a similar relationship between complementary sides of the digital information economy-the content/delivery system relationship. "The value of a Web browser depends on how many Web servers there are," says Varian. "The value of a DVD disk depends on how many DVD players there are and vice versa. The value of a player depends on there being lots of content. So it's that kind of effect that lies at the heart of a lot of these networked industries. Sometimes it's telecommunications where it's very direct and natural, but sometimes it's player/content externalities."

IR-4.1 Extending Metcalfe's Law

       "The value of a network is the square of the number of people connected to it." This is known as Metcalfe's law for Bob Metcalfe, the inventor of Ethernet. Shapiro and Varian say it's more of a rule of thumb than a law. But in any event the idea demonstrates why "the strong get more." If a network of 10 is worth $10, a network of 100 is worth $1000. Count the increase in possible linkages as a network grows, and you'll see why this is a reasonable assertion. People join the strong network, because they can see they'll get more.

Bob Metcalfe: "Square the People"

       Part of the appeal of Metcalfe's law is the way it recognizes that people remain the core resource, even in the often intangible realm of a digital network. But if information is the economic good that brings people to the network, doesn't information have a role to play in creating value on a network? I asked Hal Varian what he thought about that.

       Varian's initial reaction was to say, "No, because we really think of the network as connecting individuals. There's a sense in what you say that is right, that the network is really connecting people to information repositories and it's worth putting your information up on the net if there are a lot of readers and it's worth connecting to the network if there's a lot of content. In that sense I'd say there's a similar feedback. But in the simplest form it's a person-to-person kind of a thing."

       But exploring the idea further, Varian described a potential extension of Metcalfe's law into the complementary relationship that can arise between humans and an information source, such as an online database, saying, "It's not so different from the DVD and the DVD player. I'll buy a DVD player if there's a lot of DVD players out there. So you've got to have that kind of linkage to make it work. Even if you didn't have person-to-person communication. If it was just person-to-database sort of communication the value of connecting to the network would depend on the amount of databases out there, and you're willing to put the database up if there are a lot of people to display your content to."

IR-4.2 Networks Breed Standards Battles.

       The best way to make money in the info-economy, it seems, is to own the rights to something that everybody needs to participate in the dominant network. Shapiro and Varian write extensively about the push and pull of network lock-in; information producers try to get consumers locked in to their proprietary technology and consumers try to avoid getting locked in. At some point, it all comes down to standards, and who controls them.

       Inevitably, some players (especially strong ones) are going to want to go after control of the entire standard. It's not unusual to see an alliance of smaller players form to do combat with the big player.

       It's not necessarily a winner take all situation, and negotiated outcomes are commonplace. Wars are expensive (and are often followed by famines). "We saw a standards battle with 56 kilobit modems," says Varian. "We saw the start of a standards battle with HTML between Netscape and Microsoft, and then both backed off because I think both realized the standards battle had the potential of limiting the size of the network in that case. We see the battles going on with CDMA-GSM telephone systems last year. They settled the battle. They now have a standard protocol for third generation wireless and everybody is sort of counting their money in that pot of gold."

       Hal Varian has looked into many of the standards disputes of the past two centuries, from rail gauges (where incompatibility takes on military significance) to the rise of color television. HDTV has been a battleground for over a decade, and no one has yet emerged a winner, if making a profit is the test of success. I asked Hal Varian whether he thought these standards battles are won or lost primarily in the political realm, or whether it turned mostly on the inherent credibility of the competing technologies?

       Varian says that when you talk to technologists, they're convinced that "the worst technology always wins." He agrees that there is some impact from path dependence, an economist's term for the advantage held by the first player in a field. But the fields of competition are littered with early players who are now extinct. WordStar had a 100% market share; VisiCalc had a 100% market share, but they were upset by other technologies that moved faster and better to create extra value. So, it's not inevitable that the first mover is going to win, but the first mover certainly has the advantage. So, the basic answer to your question is, the politics, the expectations, the marketing, the branding, they're all tools that you use to try to create this kind of market dominance. None of them are, by themselves, decisive. But it's getting all those tools together that help you to win the battle.

IR-4.3 Monopolies in the Information Age.

       In Information Rules Shapiro and Varian argue that the industrial economy is characterized by oligopolies that tend to be fairly stable over a period of time while the information economy tends to be characterized by short-term monopolies. If that's true, what are the implications for governmental policy? Are we applying industrial era legal rules to an information economy that is fundamentally different?

       According to Dean Hal Varian, "The question that's on everybody's mind is this issue of whether Microsoft is going to be this long-run monopoly or if it's going to be succeeded by something else. To hear them talk, they're just hanging on by the edge of their fingernails, and to hear the other side talk they're entrenched and impossible to overturn." Varian (wisely, no doubt) demurred at the prospect of resolving this issue, but Judge Jackson won't have that luxury. So I pressed the issue a bit, and Varian explained that he sees technological innovation (and therefore, economic growth) to be moving beyond the PC, and therefore the latitude given to a Microsoft during the transition to another technology is the critical issue in terms of governmental intervention.

       "I think you see these industry structures-take the PCs as an example, where they come out, and they kind of become commodified and now the whole pressure of innovation has moved on, into areas like handheld devices, which seems to be the hot thing now," says Varian. "In five or six years, I think the handheld devices will settle down and be commodified and the push of innovation will be somewhere else. This has been true, not just in information technology, but it's been true in automobiles, it was true in washing machines, on and on and on."

       Varian says that the issue, "about Microsoft or any other company that has a dominant market share, is not that monopoly, per se, but whether they're able to extend their monopoly power in one market into these other markets and retard the pace of innovation, or the benefits that accrue to consumers from competition in the non-monopolized market." Government anti-trust policy does not make it illegal to be a monopolist. What is illegal is to attempt to monopolize. In the Microsoft antitrust litigation he says that the government first worked to establish that Microsoft has an effective monopoly in the desktop PC market, and that "they used that monopoly in an unfair way to extend their dominance into a new market, which in this case was the browser market." Varian says, "The issue isn't whether Microsoft controls the desktop market. What makes for an antitrust violation is use of that control in certain ways to try to extend their monopoly to other areas."

IR-4.4 Networking in a Regulated Future.

       While the existing legal framework remains appropriate for antitrust litigation, Varian wonders if more extensive government intervention in the information economy may be coming. Will it become necessary for the government to mandate interconnection agreements, standardizations, or open APIs? Right now, he says it's too soon to tell.

       Looking back at the network building of the late 19th and early 20th century, Varian notes that governmental intervention played a significant role. "If we look at Marconi, it's a great example. Marconi had a network in wireless. He refused to interconnect with any other network. The Wireless Interconnection Act of 1912 forced him to do so." Varian says that was justified on public safety grounds after maritime accidents, in particular the spectacular crash of the Titanic. Look at the railroads. Lots of law built up around the railroads concerning whether you had to carry the other guy's freight and what the arrangements would be. There were similar cases that came up with the telephone. We had a history where they became a regulated monopoly, for about eighty years I guess, and then were broken up with parts of them regulated and parts unregulated. So, it's been the case historically that when you have these strong network effects there have been governmental interventions in various forms."

       "If history is any guide," says Varian, "we'll see similar forces at work in the future. The question is, 'what is the minimal set of governmental interventions that can help you to achieve the desired ends?' Because in each of these examples, of course, there have been downsides to all that governmental intervention."

A Global Web In the U.S. we tend to think of our own federal government as being the source of most regulatory intervention, but the World Wide Web is just that, and just about every other new network is international in scope.

       I asked Hal Varian what he sees as the future for international involvement in regulating digital information networks in an age, not just of information, but of globalization as well.

       International involvement in network economies isn't new, but the scope of international regulatory intervention on the Internet is going to expand. Varian says, "We have international standards in telecommunications, electricity, and some of these other kinds of networks as well. And, believe me, there's going to be international regulation of the Internet. Its just naïve to think that there won't."

       "First of all there's international standards, at the engineering level," says Varian. "There's a constant and ongoing effort to do that." Will it end there? It seems that Varian expects to see more attempts at global intervention. He sees the possibility of interventions like content management (e.g., labeling a la pix) or legal standards for digital signatures.

       "Right now the U.S. is working with industry self-regulation. Europe is working much more on a regulatory mode, but the Internet is global and there's going to be some conflicts between these two approaches," says Varian. Let me give you one that's just coming out now. In the U.S., Amazon uses 'one-click' shopping. Everybody thinks one-click shopping is great. There was a great boost in Amazon's sales completion rate when they shifted to one-click shopping. In Europe, it's illegal for a merchant to retain your credit card number, in many countries; Italy is one. In France, it's legal but there are so many requirements that it's not very practical. Businesses in those countries argue that they're being disadvantaged by this difference in the law."

       As Varian points out, there's probably no good reason for the Italian government to maintain its proscription against merchants possessing the credit card numbers of their customers. However, not every cross border conflict in the global network economy will be so painless. Particularly where one nation's economic or cultural identity is at risk, globalization will not necessarily be welcomed by all.