Home | Publications | White Papers | Case for New Economic Models

The Case for New Economic Models
To Support Standardization Efforts

by Clifford A. Lynch (clifford@cni.org)


This revised and expanded version of Cliff Lynch's prize winning paper in the 1998 Standards Engineering Society/World Standards Day paper contest presents a new model for supporting national standards development. Cliff Lynch is the Executive Director of the Coalition for Networked Information, a past member of the NISO Board of Directors, and a member of the NISO Standards Development Committee.

Introduction

We are beginning to recognize that standards documents are really a form of "public good," reflecting and recording achievements of intellectual consensus among a broad community of developers, rather than representing acts of individual creative authorship. They serve as a social construct and intellectual record. Standards draw their value from wide availability and adoption. Some standards organizations have attempted to capture and control standards under copyright law in order to establish a revenue source; but this reduces and limits the value of standards for both developers and users. As social documents and artifacts, standards need to be freely available to the public, particularly in electronic form, where the duplication and distribution costs are negligible. Electronic access provides the clarity and the luxury to cast access to these documents idealistically.

I strongly support public access on a pragmatic as well as a philosophical basis. In this paper I will focus on pragmatics rather than philosophy: I believe that the benefits of broad availability of standards - to the standards development organizations (SDOs) as well as to the public - are not sufficiently understood and are widely underestimated. It's not simply the right thing for standards development organizations to do; it's an important strategy for their long-term vitality. The case for new economic models to support standardization efforts while ensuring broad access is, I think, compelling; I'll sketch some of the benefits that broad and ready public accessibility to standards documents are likely to create, and argue that the standards developments organizations should be supporting this goal.

My experience as a standards developer, as an implementor, and as a teacher is primarily with information technology and networking standards. I have drawn my examples from those areas, not only because they are familiar to me, but because there is a considerable body of experience with alternative economic and access models in the work of the Internet Engineering Task Force (IETF), and because I believe that these areas are harbingers of broader change. The communities concerned with information technology and networking standards have enjoyed relatively ubiquitous access to the networked information environment longer than any other community, and have made reliance on these tools part of their culture for over a decade. As other communities concerned with standards assimilate access to these tools and capabilities, we can expect their behavior to change in similar ways.

In this paper I discuss - often critically - the "traditional" standards development organizations and their behavior. I recognize that there is tremendous variety among these organizations, both within the US and internationally. Many are making an honest attempt, through diverse approaches, to grapple with the issues raised in this paper. I fear that many of these responses are too modest and too late. I hope the characterization here will help illuminate the growing fault lines between these SDOs and their user communities, both traditional and emerging, and the contrasts between these SDOs and the other organizations that have developed both to complement and compete with them. My goal is to provide insight into the broad challenges facing the existing SDOs and to highlight strategies that may help to address these challenges, rather then to critique the policies and strategies of any specific SDO.

The Breakdown of Current Economic Models for Standards Development

The development of standards is not free. A standard captures a tremendous amount of work and represents a real store of value. The intellectual labor required to develop a standard and the organizational processes needed to manage and structure this intellectual labor are so great that the actual cost per page of the "first copy" of a written standard - the tangible product of a standards-making process - is enormously high. The question is how to underwrite these costs.

Historically, these costs have been met through a combination of revenue sources. There are membership dues for most standards organizations; sometimes government or industry subsidies help underwrite the operating costs of the SDOs. The time (and often the travel expenses) of the experts who actually develop and review the standards is contributed by their employers. The value of this contribution of expertise is immense, and participation costs for a large organization actively engaged in standards development dwarf any membership fees that the organization pays. Standards documents are copyrighted and sold, usually at costs that seem expensive, perhaps excessive, in comparison to other publications. It is not uncommon to see standards documents selling for a dollar a page or more. (As a benchmark, compare the per page costs of standards documents to the costs of scientific journals that are currently placing such stress on library budgets.) But royalties from published standards are a measurable - and for some SDOs, significant - revenue stream.

Over the past decade, there has been growing resistance to this traditional economic model for funding standards. Probably the greatest resistance has focused on the high prices and restrictive distribution practices surrounding published standards. Not only are they expensive, but, even worse, they cannot be obtained quickly, easily, or in convenient electronic formats. Part of the issue is that the networked information environment makes it possible to distribute standards, like so many other materials, widely, instantly, and at virtually no per-copy cost. This new environment is raising expectations and creating changes that go beyond standards to the broader economics of all kinds of publishing. Further, by rethinking the structure and organization of standards as electronic documents rather than simply images of printed documents in electronic form, it is possible to add value by making standards more usable. The shift to electronic distribution is perceived by most SDOs as a threat to their revenue base (despite considerable evidence from other sectors of the publishing world that electronic dissemination does not necessarily cannibalize print sales revenue). As a result, they have either ignored the Internet as a distribution medium or have awkwardly attempted to extend print publication practices to the networked environment.

Only recently have some standards developers harnessed the electronic commerce capabilities of the Internet to sell on-demand access to standards. While this does make it easier to obtain copies of standards quickly and conveniently, very high cost remains a major barrier to access.

The difficulties of obtaining access to standards have become acute as standardization efforts have begun to produce large, evolving suites of documents that are interrelated in very complex ways. Standards users cannot easily determine which documents they really need. Ordering one group of documents, only to realize that another group will also be needed in order to understand or to implement a product, is infuriating and impractical in a rapidly moving business environment. Each order costs hundreds of dollars and involves weeks or months of delay. Online purchasing addresses the delay, but not the costs.

Large corporations and other organizations have a history of participation in the standards development process and membership in the standards development organizations. They are increasingly frustrated by the costs and inconveniences they face in purchasing products that they have already underwritten through SDO membership and committee participation. And they are turning to standard development arenas such as the IETF or industry consortia that are not burdened with these costs and inconveniences. Traditional SDOs are finally realizing the amount of ill will they have created among their traditional constituencies and are searching for ways to address this problem. Yet they still myopically regard their member organizations as the only legitimate stakeholders in standards development and continue to overlook the broader picture. This includes the role of standards as works of collective authorship and collective expertise, and the expectation of the contributors (as individuals, as well as institutional representatives where relevant) to use, reuse, and publicize their own work freely; there is an interesting parallel here with recent arguments between scholarly publishers and faculty authors who want to be free to reuse their own work in teaching, post it on web sites, and the like.

The revolt of the large corporations and organizations is perhaps most evident in the areas of information technology and networking standards, where the Internet Engineering Task Force (IETF) - unquestionably the dominant standards force in most areas of computer networking - has consistently, aggressively, and proudly repudiated the traditional economic model for standards-making. A number of other, younger organizations such as the World Wide Web Consortium (W3C) and a vast array of more narrowly focused, ad-hoc consortia have followed the IETF's lead and have garnered major support in many areas of information technology and networking outside of the IETF's scope. They are now the major fora in standards development.

The IETF exploits networked information technology both to disseminate its standards and as a key tool in standards development. The IETF has historically relied on government and industry support for its secretariat functions and on volunteer labor (in management of the standards process and in participation in standards development efforts) to a greater level than some other standards organizations. It makes all of its work products, both in draft and final form, available to the public via the Internet. The IETF follows different procedures regarding consensus, adoption, and validation for standards development than those used by the traditional national and international standards bodies, and believes they produce "better" standards that are more precisely documented, more pragmatic, more timely, and of superior technical quality. An analysis of these claims goes beyond the scope of this paper, but the IETF standards are clearly dominant in the markets they address.

The question of access to standards is not fundamentally a legal one: It's a business model issue constructed by the SDOs using the legal leverage that copyright offers, and it could only potentially become a truly legal question within the political context that recognizes and negotiates the existence and prerogatives of standards organizations. SDOs can, and probably need to, claim copyright to standards; the issue is the ends to which they use that copyright.

Access and use are very separate questions from who holds copyright on standards. Copyright is important for many reasons. When held by an SDO, it can be used to ensure the integrity of standards documents and even their continuing accessibility. These purposes are valid even if the SDO intends to give away public access to its products. Certainly, copyright is also used by many SDOs as a means of controlling and limiting distribution and of ensuring that they are paid for copies of standards that are distributed; here the SDO follows the classic practice of all publishers in employing copyright as a tool to generate revenue. The only potential legal issue is whether the SDO really has clear copyright claim. (Many SDOs have recently begun requiring committee members to sign copyright release agreements to reduce this ambiguity.) So, copyright should not be used as an excuse for limiting access; there are good reasons for SDOs to establish and defend copyrights, but this need not translate into an exercise in limiting access and alienating authors, sponsors, and users.

Copyright can be employed as a powerful tool to support accessibility. The IETF experience here is instructive. For a long time, IETF standards (i.e., Requests for Comments) did not normally carry a copyright statement. Only within the last year or two has a copyright statement been added. This statement is carefully crafted to ensure broad and continued accessibility, even over the objections of the original authors, to the documents (including the ability of libraries or other organizations to archive them) and, while the documents may be widely used and repackaged, preservation of their integrity is ensured. By and large, I believe that the IETF community, which has long been articulately vocal about the importance of the accessibility of standards, is comfortable with this copyright statement and the objectives that it supports.

It is interesting to cast this issue within a theoretical legal context. One could easily argue that standards, as records of a process of achieving social and intellectual consensus among a broad community of developers and users, represent a form of community authorship and hence should not be subject to the moral rights constraints that are normally attributed to works of individual authorship. They belong to the community, no matter which individual acted as editor or scribe or contributor.

New Uses and New User Communities for Standards

I believe that any reasonable standards organization should view its mission today as facilitating the development of responsive, high-quality standards and the subsequent broad understanding, use, and implementation of these standards. Unfortunately and for too long, standards development organizations have viewed the community that will understand, implement, and use a standard as essentially identical to the community that developed it. This group of interested parties has been considered relatively static and populated mostly by large organizations. But I believe that it is also the responsibility of the SDO - in part as a service to its member organizations - to educate individuals about key standards, the standards process, and to help phase these individuals into the community as participants.

Standards in the information technology and networking areas have become very complex and pervasive, and now form an essential part of the engineering knowledge base for many disciplines. Often, one standard will make reference to a large number of other, related standards. An engineer, product developer, student, or researcher working on a problem may need to consult dozens of standards documents. Often these standards will be unfamiliar. Troubleshooting a problem may require reference to an extensive portfolio of standards documents.

Consider networking standards as an example. Every computer or information science undergraduate, and certainly every graduate student, must be familiar with many of the key standards documents in computer networking. They will know a few in great depth, and many more at a very superficial level, concentrating on the overarching architectural principles and interrelationships among an evolving body of standards, rather than the details of individual standards. Engineers and product developers need to make reference to these documents in developing new products. System integrators use a large range of standards documents when knitting multiple products together, or when evaluating whether such linkages are even possible and what technical strategies might be most fruitful. There are a huge number of people who read popular or professional accounts of new developments in computer networking technologies and who may wish to delve more deeply by consulting relevant draft or final standards documents.

The comprehensive output of the IETF is available to every one of these individuals for the cost of nothing more than a few clicks through a web browser. These standards have become a basic part of the publicly accessible educational literature - the community knowledge base, if you will - for any network engineer or network software programmer, or anyone aspiring to become one. They are readily available to specification developers or purchasers concerned with conformance and interoperability. Every small, one- or two-person startup company in its mythical garage - and these companies are an important engine of innovation in networking and information technology - can consult them quickly and easily. The ideas embodied in the standards developed by the IETF are a basic part of the intellectual foundation of computer networking. Indeed, it's common to see students, implementors, or others actively working with IETF standards to download copies of relevant materials and to create a local "personal digital library" of standards documents which can be searched to answer questions. In addition, many standards now contain pseudo-code, code fragments, and other materials that can literally be cut and pasted into program development tools. One of the most striking realizations of this new environment can be seen at the IETF meetings, where participants attending meetings employ packet radio modems to download relevant standards onto their laptop computers for consultation during committee meetings.

When seen in this light, it is clear why the competing Open Systems Interconnection work from the traditional SDOs, even if it had not been profoundly flawed from an engineering viewpoint, would have ultimately failed in the marketplace. If you are a student learning about networking, are you going to use an absolutely up-to-date and comprehensive digital library that is readily available for free from your computer? Or are you going to try to reference a vast series of printed documents that cost many thousands of dollars, are never quite current, and take months to obtain, or perhaps are only accessible through the institutional library, maybe under a restrictive license agreement that prevents wholesale downloading onto your personal laptop? As a faculty member preparing a list of readings and reference materials for a course, it should not be surprising that you will prefer to direct students to comprehensive, timely and authoritative materials that are readily available via the network at no cost.

From active engagement with standards as part of the educational process, students not only learn what current standards say, but they also learn how to write, analyze, and use them. These skills are important to their future professional careers. This process promises a much broader base of future contributors to the standardization efforts and of knowledgeable users of the existing product of standardization work.

Thus, the next generation of systems coming out of the research and development world will not be based on the inaccessible OSI documents, but on the familiar, readily available IETF knowledge base. And the next generation of products being developed by recent graduates from the universities, coming out of startup companies, or out of organizations that did not already have a strong investment in and commitment to OSI, will not be based on the unfamiliar, inaccessible OSI standards. As new standards are needed, they will be developed within the well-known IETF framework. This is how public access to standards - particularly within the higher education community - can translate into a very real, long-term competitive advantage for the SDO.

A traditional SDO struggling with this situation and concerned about revenue preservation might argue that a university library should take responsibility for acquiring their standards and making them available to the students and faculty - and for underwriting the work of the SDO. The library's acquisitions budget ensures that important material is available to the university community based on its intrinsic value and merit, rather than simply on considerations of cost and accessibility. At large research universities, it is possible that the library may have subscribed to the traditional networking standards in print, or perhaps on a CD-ROM that one can use at the library (since CD-ROM databases are rarely as easy to use over the network as web- or FTP-based resources). But the time lag and the inconvenience is still sufficient to make most students and faculty less interested in the "formal" standards than those readily accessible on the net; it is very hard to compete with a direct, non-intermediated strategy that offers users network information - particularly when this networked information is linked to the literature on one side and to public access code resources on the other.

Of course, there is a vicious cycle, a death spiral, at work here. As funds get tighter, libraries are questioning the value of spending thousands of dollars a year on arcane standards that seem increasingly irrelevant to current teaching and research, that don't appear on faculty reading lists, and that students and researchers aren't requesting. Every time a research library decides that these standards are not worth the money is further evidence of their irrelevance. If they are unavailable to faculty and students because the library didn't purchase them, then they certainly won't be used.

University libraries should help to underwrite the work of the SDOs through their acquisition budgets. But the SDOs will have to take a dangerous high road of soliciting contributions rather than of holding access hostage. They will need to make the case to university librarians - both directly and through the voices of students and faculty - that their standards are important and widely used, and thus worth some underwriting, rather than demanding underwriting in advance within a framework that minimizes interest in the standards products.

During the OSI debacle, had the SDOs offered the university libraries a comprehensive, affordable, timely, and usable subscription to the relevant standards, they may have slightly reduced the general preference for IETF-based standards efforts. But they failed to do so, and even if they had offered such a subscription, it would not have realistically addressed the problem of access beyond the university student and faculty communities. Further, there was no strategy to move beyond access to participation and engagement in the standards process by the academic community.

There is a lesson here that any SDO today ignores at its own peril. Standards, and particularly broad-based suites of standards informed and contextualized by an architectural model, can fail not only in the immediate implementor marketplace, but in failing to gain mindshare among the next generation of standards developers and users who are today's students. OSI, of course, had the distinction of failing on both counts, and as such serves as an almost unique case in point.

This is a indeed properly characterized as a "mindshare" issue (though an economist might instead speak of "network externalities," of the idea that wider use and adoption creates value). Current economic models that market standards as expensive printed works alienate today's standards users - the very people who have invested large amounts of money to develop them. And at least as importantly, they equally disenfranchise tomorrow's potential users - students learning the craft of engineering and the role that standards play in that base of knowledge. It is a competitive environment. Just as we are seeing scientific and scholarly journals becoming readily available on the network and taking mindshare away from hard-to-obtain printed journals, we are seeing inaccessible, expensive, print-only standards fade into irrelevance in the minds of the next generation of standards users and developers.

There is also the issue of participation. At least in the networking and information technology areas, standards documents often represent a focus for the best ideas emerging from the research world into broader commercial implementation. Draft documents that are publicly available in electronic form, coupled with an effective forum for comments and discussion such as an electronic-mail-based list, let students and faculty worldwide enter the discussion, even though they do not have the financial resources to participate in face-to-face standards committee meetings. While this can create a difficult-to-manage, anarchic process, I believe that the benefits of the broader community participation outweigh the convenience of a relatively closed standards process. Placing standards development into the networked information context makes this broad participation possible. It is interesting to note that the open, network-based IETF standards development work engages not only working professionals, but also some of the best minds among the graduate students in computer science and allied fields. The closed processes of the traditional SDOs, which emphasize face-to-face meetings, have largely failed to connect with this community. What is needed is a direct bridge from the implementor community both to the research world and to the community of students who will become the next generation of both standards developers and commercial implementors. The IETF has put such a bridge in place which the traditional SDOs have largely failed to do. And the building of this bridge begins with broad access to both the existing base of completed standards and the drafts of standards under development, along with a viable means of commenting on these drafts.

Conclusion: New Economic Models for the Standards Process

It is clear that the failure of traditional standards development organizations to embrace convenient electronic distribution is reducing their products' value for many current participants in the standards development and implementation process. These organizational participants are beginning to vote with their feet. The move from traditional standards processes to nontraditional venues like the IETF, the World Wide Web Consortium, and the endless series of ad-hoc industry consortia is clear. This move is not motivated simply by the slow speed and lack of agility that characterizes traditional standards development. It is also the result of an end product that fails to meet marketplace needs and is priced and distributed in a way that alienates the very organizations that played a key role in its creation.

More easily overlooked, however, are the changing user communities for standards and ways in which these user communities employ standards documents. The traditional economic model of limiting access to generate income through the sales of standards documents is now actively counterproductive to the broad understanding, implementation, and use of standards. This is particularly relevant in areas such as information technology and networking, where few standards exist in a vacuum. We are seeing large sets of standards that fit within broad intellectual and architectural frameworks which interrelate in complex, continually evolving ways, and that are of the core knowledge base of the discipline. And, as discussed, restricting access has perhaps its greatest, and most detrimental, impact on the new generation of engineers and scholars just entering the workforce rather than on the current generation of established engineers already committed to the standards process and the use of the products of that process.

So, what are the characteristics of a viable economic model for the development and dissemination of standards in the networked information age? Charging for access to drafts or to final standards documents via the network is demonstrably a disastrous error. This does not mean that there is no place for a print publication strategy. There is compelling evidence that at least some people will still pay for a nicely printed, bound, and reasonably priced copy of a standard or other document, even though most or all of the equivalent content may be available for free downloading via the Internet. The publication strategy must compete with the personal laser printer, however, rather than offering the proposition that the print publication is the only way to obtain the document. To meet this challenge, prices need to move more into line with other types of published works. In so doing, compendiums of multiple related standards in printed form (perhaps in conjunction with tutorial materials) are likely to enjoy a more robust marketplace as reference handbooks rather than as a series of highly priced, standalone standards documents. It is still less expensive to produce and distribute relatively large books through traditional channels than to have users print them on demand. And there are many opportunities to add value through tutorial or interpretive material that might be added to the standards proper, perhaps producing something that is closer to a textbook than a standards document. But the point here is to position products that compete with, and in fact complement, freely downloadable documents that can be printed on demand at the end user's expense or used directly in their electronic format. In order to be effective in this role, SDOs will have to behave much more like "real" publishers, and to provide incentives to authors who prepare the tutorial material.

SDOs will need to make their case for support to academic libraries. This is best done perhaps through effective service to the broader academic community, rather than only directly. And it's a hard argument, based on why libraries should contribute to their support, rather than why they must pay because it's the only way to get the materials that their user community needs. There are difficult choices about priorities and commitments here, and libraries too must be open to thinking about the new models. For example, while the Association of Research Libraries, for example, has been generous in underwriting the National Information Standards Organization over the years through membership (probably more to support NISO's work than to gain access to its products), few individual libraries are NISO members, though many purchase at least some of their printed standards. More generally, little has been done to move the broader engagement between higher education and libraries on one side, and standards development organizations on the other, forward. How high a priority will broad access to standards be in the competition for acquisitions dollars, particularly when standards are competing for these dollars on a more abstract basis of contribution to the overall public good than the pure commercial transaction that characterizes scholarly publishing?

The overhead of the secretariat and editorial activities of the SDOs in reviewing standards, editing them, and placing them on the Internet will still need to be underwritten either by SDO membership fees or by outright subsidy from government, higher education, and/or industry. The size of these overhead costs, as well as opportunities to open up, streamline, and reduce the expenses of the standards development process through information technology and the full use of the net, will undergo new scrutiny. The case will have to be made for the importance of standards to the narrow immediate implementor community and to the much broader community that includes students, educators, researchers, and the general public. The standards development organization will need to demonstrate that it has redesigned its processes and dissemination strategies to be responsive to these broader communities. All of these factors will need to be addressed as SDOs make the case for support from academia, government, and industry.

Finally, I believe that there is hope for marshaling increased government and higher education support for standards as part of a rethinking of economic models. For example, the National Science Foundation in the US has a group working on the conceptualization of a digital library to support undergraduate science, mathematics, engineering, and technology education. In my view, standards will be an essential part of the corpus of materials that will comprise such a digital library. Here is a new opportunity to link standards and education. This linkage is real, but poorly recognized, and it is important.

Standards today, particularly in information technology and networking, are part of the public knowledge base of education, research, and industrial development. It is in everyone's interest that they be broadly available, and that standards development work continue to be supported. We need new economic models that continue to insure financial underwriting of the legitimately costly standards process, but not at the expense of access or broad participation, particularly at a time when the new networked information environment offers so much opportunity to expand and leverage the value of standards work.

 

Clifford A. Lynch, “The Case for New Economic Models to Support Standardization Efforts,” Information Standards Quarterly Vol. 11, No. 2 (April 1999), pp. 5-10.