D. Economists ponder standards

Path dependency, increasing returns and network effects butt in — their sudden ubiquity in business is discussed — an example from the field is used to illustrate how 'standards' are now a common element in business strategy.

1. Economists sometimes ponder apples, and when they are not under the apple, dreaming like physicists of Newton(s), they ponder qwertyness as they type on it. If one were surveying the possible scholarly approaches to standardization, one would discover the economic mini-consensus that goes by the names "path dependency, increasing returns and network effects." However, even outside of such a survey, the rhetoric and the ideas of this consensus have begun to permeate the worlds of internet commerce and were common references in the trade and quasi-academic literature of standardization— less in healthcare, but present with the internet nonetheless. While the tiny abstract machine of network effects makes instant sense to many actors in the field, it is perhaps only convincing as a scholarly endeavor for mathematical economics— the deliberate choices of actors may not be reached by the simplistic means they propose, but the justifications for those decisions often follow their lead. It appears here in brief explanation of its importance as a narrative justification for actors in the internet economy— particularly for the general importance of standards to the actors in my field and their marketing tools.

2. The literature on increasing returns and path dependency has increasingly returned in the last twenty years, largely a result of a few key articles that made the object lessons contagious. Beginning with W. Brian Arthur's 1983 working paper "On competing technologies and historical small events: the dynamics of choice under increasing returns" [Arthur83] and Paul David's 1985 article "Clio and the economics of QWERTY" [David85] economists learned departures from single-equilibrium solutions via the mathematics of probability in stochastic processes (in Arthur's case) and via the "essential" role of chance historical events (in the case of David). The game is played largely by matching model to metaphor and the play of chance and necessity (Arthur cites Monod as chief inspiration) in the service of systemic change enthrall to the point of concern. As Charles Sabel puts it in his critique of path dependency, (see below): "and so the dismal science finds a new way to be dismal (Section 2, paragraph 6)" [Sabel95].

3. For these economists, standards adopted for any reason, call it chance or necessity, can have a profound effect on markets. David types QWERTY on his Apple because a few manufacturers learned the hard way that once a population of typewriters learns by using one thing, the prohibitive cost of changing keeps them coming back[1]. Path dependence thus emerges as the economist's genuflection before history, even as history is reduced to the chance event. QWERTY, or more mythically VHS, is not the superior technology that neo-classical welfare economics demands the market favor. Inferior technology has achieved 'lock-in' and it is economically disadvantageous— read impossible— to improve. To save a neo-classical synthesis, these events fall outside the closed system: they are network externalities. Externalities are a familiar beast of economic theory, but combined with path dependency and the "complexity theory" of stochastic processes, it becomes a question of a few wrong moves making a monopoly of mediocrity. Conveniently, the price system still functions in this theory, as long as history is simply accident, as long as history is only "cleopatra's nose." QWERTY is an accident for the economist, albeit an especially significant one, updating and reversing the apiarian party-line with a fable of the butterfly from chaos theory. Virtue no longer rests in vice (self-interest), market failures require correctional institutions to return vice to its hive. [2]

4. Economists Margolis and Liebowiz, vigoruous opponents of this theory's implications for classical economics, critique Paul David on historical grounds by unearthing every last bit of information about the Dvorak keyboard that an economist playing historian can uncover. Their critique uses brute force, but makes its point: historical small events are arbitrarily common, which historical event is important can still be investigated by standard classical and neo-classical economic tools. Charles Sable critiques path dependency in much subtler and broader sense. He offers counter examples (such as the choice to switch from a typewriter to word processor, clearly a larger switch in sunk costs than from QWERTY to Dvorak) to prove a thesis about the mutual intelligibility of different choices. Sabel points towards the pragmatism of economic institutions that he identifies elsewhere, in order to show that the existence of "increasing returns" as an assumption in the economies of today, is countermanded by a pragmatic deliberation on the goals of economic activity.

5. Curiously, the word and the concept "standards" is absent from the most theoretical of this work, even from Sabel's critique. Perhaps this is because for the economist who studies the market as nature, standards are still yet culture. Or perhaps it is because standards are seen as complex but unproblematic technical decisions. They can and must be included under the heading of regulation and law, since supposedly deliberate process determine them. They may then have an effect on a market or an economy, but they are certainly not an essential component. On the contrary, when theorists of increasing returns pick examples, they call them "technologies," "solutions," "adoptions." The impossiblity of this inversion is signalled syntactically in Arthur "The standard technology problem in economics was figuring out the economic circumstances under which a new, superior technology might replace an old inferior one, and how long this process might take." [Arthur83] Problems can be standard, given, in need of transcendence— but technologies should not. Technologies are not arbitrary conventions, they are rational evolutions. The quickly told story of VHS and Betamax often serves to illustrate: VHS gained an early market lead that led to the purchase of more VHS players and to the distribution of major films more commonly on VHS. With every decision to go VHS over Beta the "network" (the sunk costs of the VHS market) gained incrementally in inertia until it became disadvantageous for someone to buy Beta initially or to switch from VHS to Beta. The rub is that, accrding to the story, VHS is "an inferior technology" whose success depended on some set of chance events that led to VHS cassettes being more widely distributed. VHS was at no point declared a standard by any governing body, other than perhaps internally by JVC, and yet it quickly became the de facto standard for the industry, requiring that any corporation that wanted to participate in the lucrative business of videocassettes would have to use this standard, and more importantly that any existing competitor would be driven out, not on price, and not on quality, but on the "network externality" of incompatibility — a market failure that drives economists batty. Of course the criteria by which "inferiority" is measured is never explored— though it is largely immaterial, since the key problem for the economist is that competition and equilibrium are disturbed by problems associated with the existence of increasing returns—which is also not questioned— which leads to lock-in and market failure. Until the work of Carl Shapiro, Garth Saloner, Joseph Farrell, and Michael Katz (responding in various ways to the proposals of Arthur and David) which looked at idealized standards decisions, the role of standards was assumed to be unproblematic.

6. This contrasts starkly with the strategic use of 'standards-talk' in the justification of decsions in both my fieldsite, and the internet economy at large. These descriptive economic theories quickly become normative when funneled into the array of publications with wider than library circulation. What has come to be called the "winner take all" economy (and when Adrian Gropper refers to "the impact of standards beyond healthcare") is the vernacular version of this economic mini-consensus. Path dependency and lock-in are memorable fables, and a convenient explanations for both success and failure, but when combined with the object lesson of "network externalities" (and especially the word 'network', which even though it does not refer only to technical networks, is constantly made to bear the weight of every amateur explanation of the 'New Network Economy') they become explicit strategy for the participant in economic life. Add this self-monitoring management behavior to the already complex attempt to create standards and what used to be a game of coordination played with implicit gains, becomes a game of monopoly with explicit support from the science of economics.

7. This introduces a whole new level of publications, from trade journals in the world of standards as well as the glossy journals and tipsheets marketed at CIOs and other executives that disseminate this langauge. These sources range from the cotton candy punditry of Gilder, Negroponte, Forbes, Fortune, Fast Company to the expertise of someone like Carl Cargill, former standardization strategist for Netscape, to various quasi-academic management handbooks such as Peter Grindley's Standards Stragtegy and Policy: Cases and Stories (whose introduction is sub-titled "Winning with standards") [Grindley95]to trade and quasi-academic publications, in particular, StandardView. All of these sources preside over what they experience and communicate as a change from a purely technical concern to a newly commercial one (though not yet political, unless, ironically someone breaks a law, especially a patent or copyright law).

8. In 1998-9, along with the Open Source movement (q.v. section H below), the jargon of increasing returns and network effects are officially the Buzzwords du Jour. Carl Shapiro and Hal Varian published Information Rules: A strategic guide to the Network Economy [Shapiro99] and Kevin Kelley, futurist, wacko, and ex-editor of Wired has published an even more watered down version called New Rules for the New Economy : 10 Radical Strategies for a Connected World [Kelly99]. The emphasis on 'rules' is primarily a sort of auto-immune response to the extraordinairy quantity of speculation that circulates about the radical transformation and functioning of the internet economy. Shapiro and Varian say on page one: "Technology changes, economic laws do not," in an attempt to leverage the ailing legitimacy of economic theory and the sexy academic glamour of Harvard Business School onto a handbook for capitalists who are tired of hearing that "everything has changed" or that you can "make your own rules." This handbook gives digestible versions of all of the writings on increasing returns and path dependency and offers sagely advice on what that means for Your Business. It finds its way into venture capital firms, the justice deparment's case against Microsoft, and the everyday language of business plans and strategy abstracts. While few of these books and annoucements offer any real detail about the technical structure being put in place, they do at least sense that standardization (not just mechanization, or formalization) have a significant role to play— and do a fine job of converting anyone who needs only a little convincing.

9. Consider— as an example of this process— an article[Gropper99] that Adrian wrote— the details of which are less important than the insistence on standards, and especially the distinction between DICOM [DICOM is explained in more detail in section F] and Internet standards:

The Internet has demonstrated the value of standards. Cost, ease-of-use, functionality and growth all benefit from the widespread adoption of standards. It's hard to imagine filmless radiology without standards to replace the film, film jacket, flash-card and light box.

10. Rarely does an editor let such repetition pass, but in this case it is par for the course in an industry where such statements are only slightly less common than the demands for standardization that precede them. Adrian's article also tries to differentiate within standards, which we will come to shortly, in order to train CIOs and PACS buyers about their importance:

DICOM is essential. DICOM however, was not designed to the same standards of accessibility, security and cost-effectiveness as the Internet. As a result, neither PACS nor teleradiology can be designed exclusively around the DICOM standard. The DICOM standard will continue to evolve, but in those areas where the Internet offers an alternative standard, DICOM may not evolve fast enough to compete.

11. The language of evolution and competition should not obscure the fact that the difference was one of design. "DICOM however, was not designed to the same standards of accessibility, security and cost-effectiveness as the Internet." The conotational slippage deserves the focus. A standard (the internet standards) "designed to the same standards" captures the blur of history, economics and technical design that differentiates DICOM from Internet. What standards are these? Implicitly high standards of accessibility, security and cost-effectiveness? Why wasn't DICOM privy to these same "standards" when it became a standard? What peculiar dissemination is signaled by this homophony? Some specificity is offered, especially concerning the danger of the "vendor-specific" standard:

The use of Internet standards such as firewalls, virtual private networks (VPN), digital certificates and directory servers such as the Lightweight Directory Access Protocol (LDAP) all combine to provide a complete range of security and administration solutions that are much more powerful and cost-effective than any vendor-specific technology

12. Some sort of explanation of this difference constitutes the rest of this section, especially as it becomes a concern that crosses more than just competiting vendors, but includes lawyers, revolutionaries, and hackers in collectively reconstructing capitalism. But for Adrian, statements like these serve primarily to "educate" potential customers about the uses of the internet, and strategically leverage a widespread concern about standards in healthcare:

Standards, specifically those surrounding the Internet, have impacted all aspects of health care.

13. These statements, directed at hospital radiologists, purchasing agents and CIOs (and explained in more detail below {q.v. below}, are perfect examples of the manner in which strategic thinking has come to incorporate standards as a primary economic consideration. And whether increasing returns and network effects theorists become justification for such decisions or not, this indicates that standards decisions have penetrated to a level of strategic thinking that used to be occupied only by marketing, sales, returns and occasionally, regulation.

14. The commercial focus may be appropriate, a recognition that corporate executives never spoke of, or worried about, standards as a strategic issue until now. But of course, that doesn't mean that these battles were ever purely technical issues (q.v. section F), only that they were informally political, rather than programmable by managers and CEOs

15. From the standpoint of the increasing returns theorists, as well as managers who treat themselves to these lessons, standards can be achieved by explicit agreement, and often are, but the system is hierarchical, inefficient and riddled with rule by judgement and abuse of power. By analogy with Nature, the object lessons of increasing returns, lock-in, path dependency, especially when applied to technology, teach how markets use positive feedback to reinforce choices, and the principal problem then becomes: how to make the right choices at the right time to ensure that the 'right' standard is chosen. (as with the problem of VHS's inferiority, the issue of rightness is a technical one, which invokes a set of criteria no less governed by a problematic of convention and nature, and often the last instance metric of correctness in standards is reference to some convention of simplicity, elegance, robustness and completeness— Sean's admiration of the HTTP and HTML standards cites the "non-commercial" work of scientists collaborating around the globe.)

16. The uses to which this common sense is put are myriad and contradictory— but they recuperate a sense of freedom in the development of technology. Rather than always waiting for an authority to standardize a technology, companies and executives can aim any technology at a market and hope that consumers "make the right choice." CEO's may approach this set of theories with cupidity or cynicism, they may believe the economics has the certainty of science, or they may simply urge the theory towards towards truth by insisting it against lived experience of the economy (that companies conscionably release poor quality technology, and leverage marketing to make it standard, thus becomes a familiar denegration). The strange mix of normative and descriptive language the bedveils economic science, and especially economic popularization, leads to an equally strange mix of actions on its behalf. On the one hand, defeatism strikes those who miss the jackpot of lock-in, while a kind of unreconstructed genetic determinism often infects those who watch their products "become the standard." On the other hand, the Justice department has taken a very serious interest in these theories (which has in turn prompted more economic theorizing) in the case against Microsoft, where they could play a role in the first real test of how the government will figure itself in the 'new economy.' All of these issues concerning standards and the economics of standards are visible in the mediascape, which makes them more than models and less than gospel, but nontheless a figure in a set of decisions that are political, even when they are not labeled as such.

Last Modified 11-Sep-99 9:01 PM ckelty@mit.edu

Go Back to the Start