Guide for the Perplexed

The interactions among technology, economics, and law can be both intricate and confusing. Stated simply, technology constrains what we can do, legal rules change the alternatives’ relative costs, and economic incentives indicate which alternatives we are likely to choose.

Mature industries based on well-understood technologies tend to change slowly. Our understanding of the underlying technologies generally leads to stable, reasonable laws and regulations that enable effi­cient companies to earn attractive profits without violating important public policies. Industries undergoing rapid, radical technological change rarely exhibit that type of stability. New technologies introduce new opportunities. Some attempts to exploit these new opportunities may yield enormous rewards, while others that appeared to be equally pro­mising may lead to spectacular failures. Meanwhile, the existing legal environment—by definition a legacy of a different technological era—may be ill-equipped to deal with this brave new world. The inher­itance is almost certain to include laws that prohibit beneficial advances, laws that permit harmful exploitation, and even laws that somehow manage to do both simultaneously.

In the late 1990s, technological advances in software, micropro­cessors, and communications disrupted conventional thinking. These advances catapulted the Internet from an esoteric research tool into the preeminent global infrastructure for commerce, entertainment, and com­munication. New technological possibilities seemed to arise every day, and every new technology spawned promises of obscene riches. Profes­sionals immersed in relevant areas of economics and law suddenly had to rethink their most basic assumptions and attitudes; technology had invaded their realms. Technologists had to adjust many of their own assumptions and attitudes; hordes of lawyers, economists, financiers, investors, marketers, regulators, lobbyists, legislators, and onlookers invaded their laboratories.

Now that the situation has calmed down—at least a bit—an assess­ment of the technologies, the economic theories, and the legal doctrines that emerged during the recent information revolution is possible. Unfor­tunately, disentanglement is rarely easy, and translating esoteric, nuanced arcana into plain English is always a challenge. The information sector’s emergence from the academic temples onto the front pages forced copy­right lawyers to learn about oddities like “data encryption standards” and “file transfer protocols.” Software specialists learned that much of their collegial behavior violated copyright law and often crossed the line into the even more ominous “misappropriation of trade secrets.” Econ­omists had to confront the relationship between intellectual property policy (which most of them understand intuitively) and intellectual prop­erty law (which many of them find baffling). The general public—which had no technical, economic, or legal background—had to assimilate all of these ideas, along with academic concepts like equity valuation,

network economics, industrial organization, antitrust leveraging, monopoly maintenance, transaction costs, software design, user inter­faces, Ponzi schemes, crowd psychology, and accounting conventions.

The sheer complexity of this kaleidoscopic information sector drove many to passive reliance upon the pronouncements of “experts.” On one hand, there’s something comforting about waking up in the morning to discover that Microsoft has automatically downloaded a software upgrade for you, that Merrill Lynch has found a new stock for your port­folio, and that Congress has changed the copyright laws to keep some important industries functioning as usual.

That quest defines our first true challenge. We must glean the key con­cepts from these fields, in plain English, and explore the ways in which their interactions shaped the information sector—and thus increasingly shape our lives.¹⁶ Some concepts emerged from the temples of industrial organization (IO) economics and competition (or antitrust) law. The rel­evance of these concepts to the Microsoft trial is obvious: Microsoft stood accused of violating the antitrust laws. It helps to know why these laws exist, whom they protect and how, what the government believed that Microsoft had done wrong, and how Microsoft justified its behav­ior. These laws’ relevance to other aspects of the information sector, though, is just as significant—if somewhat less direct. Day traders and Internet investors spent a great deal of time reading about the Microsoft trial. These readers invariably asked themselves two questions: Why won’t the government leave Microsoft alone? and How do I identify and invest in “the next Microsoft”? Answers require at least some familiar­ity with the concepts of market structure, its natural workings, and the abilities of certain players to interfere with its smooth functioning.

These same temples also house the secrets that will allow the Internet to make us all rich. Their inquiries into “transaction costs,” or the ele­ments of pricing that owe less to the cost of production than to the cost of distribution, reveal the areas in which new business models oriented around the Internet can help both consumers and producers—while raising the ire of traditional distributors. This revelation sets the stage for the battles surrounding both Linux and Napster—and likely the entire future of the information sector.

But while IO and antitrust are the disciplines best suited to explain these concepts, they are also both established fields, first developed to study railroads, manufacturing, and smokestack industries. Much of the New Economy literature contended that old avenues of inquiry were inapplicable to the dynamic, creative, information sector. While that lit­erature may have oversold the novelty of the situation, the shift from an industrial to an information economy did introduce a number of new concepts into these once stodgy fields, most notably “network econom­ics.” In a network industry, rational consumer behavior turns selected popular products into entrenched standards. A single player who owns that standard becomes very powerful, often able to disrupt the workings of the marketplace and retard innovation in ways that serve the owner’s proprietary interests. One question that lingers among erstwhile dot-com advocates is why Microsoft was so successful while the dot-coms failed so miserably. Certainly, part of the answer lies in managerial competence and product quality. But those answers are only partial. Platform software (e.g., Windows) and many Internet spaces exhibit aspects of network growth. But only platform software allows the exploitation of consumers—which translates directly into high profit margins. Microsoft understood this distinction. The dot-coms did not.

So far, so good. But these ideas all deal with markets.

Two critical ideas from artificial intelligence (AI) and software engi­neering provide some answers: translation and modular design. The “translation chain” embodies the various tasks that software developers tackle. It refers to the many steps required to enable those of us who speak human languages to communicate with silicon chips that notice when voltage levels change. One of computer science’s truly marvelous accomplishments has been its incorporation of increasing amounts of the translation task into the computer, a chain of translations that links volt­ages to bits, to numbers, to words, to programming languages, to user interfaces, and eventually to us. Not too long ago, it took years of spe­cialized training for a person to communicate with a computer. Now even preliterate children can use mice and touch screens to “talk” to microchips. Web browsers that incorporate color, pictures, and sound were an important link in that chain. But they were only the gloss built atop generations of previous work.

The conceptual links in the translation chain lead naturally to modular design. Software engineers understand that the most effective way to improve our computing experience is to link a new software module to the previous generation’s interface. Graphical browsers, pictures, and sound hardly appeared out of nowhere. All of these innovations—like the innovations of preceding generations—arrived first as modular add­ons or as separate application programs that consumers could run on top of existing systems. But these add-ons were often clunky and full of bugs. Software developers eventually worked out the bugs, integrated the new ideas into the main software’s translation chain, and presented slick unified products.

In most generations of software, engineers were responsible for inte­gration decisions. They decided when an add-on product had matured enough to incorporate it into the translation chain. Sometimes they were right, and people bought their new integrated products rather than the old modular ones. Sometimes they were wrong, and their companies lost sales and profits to their competitors. But the Microsoft trial introduced a new twist. Suppose that strategic marketers, rather than software engineers, made the integration decisions? And furthermore, suppose that those marketers were in a position to foist their integrated products on the public whether consumers wanted them or not? Should consumers care? Should the government care? Should the law care? One of the most perplexing questions to emerge from the Microsoft trial thus remains: When is product integration a good thing? This question is likely to reappear in a number of legal battles over the next few years.

Translation chains and modularity are important concepts in the design of information products. But the very notion of an information product raises its own set of complexities. After all, software packages, Web sites, databases, and even music files are little more than interest­ing ideas that are now easy to copy, easy to circulate, and easy to broad­cast. How can anyone truly “own” this type of product? And more to the point, why would anyone want to? How can anyone turn a profit by investing in product development and then watching it circulate freely? These questions lead us back into the realm of law, specifically IP law. IP law long has motivated invention and creativity by allowing inventors and authors to retain certain basic rights in their creations’ use and distribution. In fact, the U.S. Constitution empowered Congress to create a body of laws that provided the appropriate motivation. Con­gress responded by codifying the laws of patents and copyrights.

The constitutional imprimatur notwithstanding, these IP rights create an odd sort of property. Their owners can’t really watch them or police their use; many people violate IP rights with impunity. Usually all the “owners” can do is file an occasional lawsuit—and even then, an infringer’s inability to pay often ends the litigation prematurely. The ability to sue, in turn, suggests that IP owners also can offer a valuable license defining the terms under which they promise not to sue; that’s where they make much of their money.

But the information age has placed an enormous strain on the precepts of IP law. With each passing year, more and more products become digital bit strings, and more and more people learn how to infringe (par­ticularly copyrights) in the privacy of their own homes. These changes infuriate the rights holders and threaten to erode the profitability of their operations—and reduced profitability implies reduced motivation implies reduced product development. Put bluntly, the ability to swap music files could reduce the amount of new music—at least according to one school of thought. This issue arose only at the periphery of the 1990s information sector, as the motivation behind an important change to the copyright laws known as the DMCA of 1998, and as the undercurrent of the battle over Napster. Its importance, however, grew quickly—and it will grow even further as the information sector spreads its tentacles throughout the economy. IP law thus promises to become the informa­tion sector’s next great battleground.

Pillars of the three temples—the legal concepts of antitrust and IP, the microeconomic theories of IO, and the computer-science principles of artificial intelligence and software engineering—formed the framework for the information sector. We must therefore consider these foundations before we can truly understand the information sector’s story.