The Law of Opposite Numbers:
Standards Dynamics and the Global Logic of Software

Philip E. Agre
Department of Information Studies
University of California, Los Angeles
Los Angeles, California 90095-1520
USA

pagre@ucla.edu
http://polaris.gseis.ucla.edu/pagre/

Submitted to Research Policy.

This is a draft. Please do not quote from it.
Version of 7 August 1999.
4800 words.

 

Abstract

The emergence of the Internet illustrates a broader principle: the "platform cycle" whereby new service layers are abstracted from a variety of existing bundles of functionality. The new service layers exhibit substantial economies of scale that enable new applications to emerge while condemning the existing bundles to eventual obsolescence. Inasmuch as the resulting standards set rules for future developers and users, the dynamics of technical standards in the platform cycle resemble the coalition- and consensus-building processes of a democracy more than they resemble an idealized market. Customers hoping to shape standards thus face collective action problems.

 

INTRODUCTION

Although it has certainly not reached its final form, the Internet has nonetheless achieved the critical mass that it needs to support a great diversity of applications and community-building experiments. Competitors such as OSI have been decisively routed, and the Internet Society's slogan, "IP on everything, everything on IP", now seems entirely plausible. The Internet has won.

Yet opinions differ about what this might mean. Networked information technology, by which we now increasingly mean applications running over IP, has been identified both as the guarantor of a thoroughly decentralized marketplace and as one motor of an extraordinary era of industrial concentration. The Internet has likewise been characterized as the enemy of intermediaries, even as many of the most successful new businesses are precisely intermediaries (Halper, 1998). Disproportionate media attention to Web-based consumer commerce may have contributed to a widespread sense that the Internet leads to completely novel alternatives to existing institutional systems. Yet something like the opposite is the case.

It is still hard to demonstrate that we are facing anything like a computer revolution, any more than when that revolution was first announced (Kling, 1991). It is, however, reasonably clear that the Internet is participating in one change of potentially massive consequences: the reconstruction of the interfaces among a large proportion of social institutions. One example of this trend is the difficult yet vast development of IP-based protocols for supply chain integration (Frook, 1998). Ordinary e-mail opens up new channels between professionals and clients, elected representatives and their constituents, and the tops and bottoms of hierarchies (Sproull and Kiesler, 1991).

Although it is tempting to generalize about the consequences of these reconstructed boundaries, it would nonetheless be seriously premature. What is possible, however, is to articulate some of the counterintuitive economic principles that interact with the complex particulars of each case to produce what broad patterns we do see. This analysis does not dictate the chaotic sense of urgency and panic that one senses in much popular discourse on the place of new information technologies in society. It does, however, recommend that virtually every organization that depends on information technology, for reasons both of self-defense and social responsibility, take up a new and unfamiliar relationship to its environment. Motivating and analyzing that new relationship will be my principal purpose here.

WHAT HAS BEEN LEARNED

In order to understand the deeper significance of the Internet for system development and organizational IT strategy, it will be useful to review some of what was learned about these topics in the days before the Internet achieved critical mass. The most fundamental result is that information technology cannot be separated from its institutional context (Kling, 1996). For one thing, every computer system embodies an elaborate set of ideas about the institutional environment in which it is used. A technology-driven approach that ignores the larger environment is therefore unlikely to succeed -- not because the resulting technology is agnostic about institutional structures, processes, and relationships, but because its tacit ideas about these topics are wrong -- connected randomly at best to the reality.

One well-known example of this phenomenon is found in Orlikowski's (1993) study of a consulting firm's attempt to adopt Lotus Notes. The purpose of adopting Notes was to encourage the firm's consultants to share information, and yet the firm's competitive promotion criteria created strong incentives to keep information to oneself. The institutional ideas presupposed by the software were not aligned with those presupposed by the organization.

For this reason and others, sound systems development practice requires that clear definitions of organizational goals practically saturate the development process. Worked-out organizational methodologies are available for doing this (e.g., Weill and Broadbent, 1998). Yet at the same time, empirical research shows that systems development in real life organizations is often, if not always, a highly political affair, with different organizational coalitions seeking to shape the system architecture to their advantage (Danziger, Kling, and Dutton, 1983; Knights and Murray, 1994).

As time goes by, these politics can become ever more complicated. In the early days of computing, it was possible to speak of something called "computerization", in which an organization without computers was turned into an organization with computers. Few organizations face that trajectory any more. Instead, system acquisition and development is an ongoing process that takes place in an environment of multiple, often incompatible legacy systems (Kling, 1989: 506). Part of the inherited incompatibility is purely technical, as organizations seek to integrate legacy systems that embody the proprietary architectures of different vendors. Another part is institutional, for example in settled policies, practices, and distributions of decision-making authority and skills (Kling and Iacono, 1989). And yet another part is semantic: different organizational functions -- that is, different professions -- have their own language and use words in their own frequently incompatible ways (Davenport, 1997, Schmidt and Bannon, 1992). The manufacturing systems of one company, for example, may well have more in common with the manufacturing systems of another company than with the financial systems of the same company.

As a result of these institutional dynamics of computing, it is little wonder that clear-cut productivity increases from organizational computing have been hard to identify (Landauer, 1995; Madrick, 1998; Stiroh, 998; cf. Brynjolfsson and Hitt, 1998). Part of the problem is that many development and acquisition projects fail outright. Another part is that changes in technology lead to changes in organizational strategy that make before-and-after comparison difficult. And another part is most likely that the institutional learning curve required to take full advantage of new information technologies is steep, and the learning itself diffuses only with some pain (King, 1996).

STANDARDS DYNAMICS

These phenomena of the construction and adoption of information technology, for all of their complexity, reflect an earlier and simpler age. They occur within the boundaries of a single organization, and they concern relatively isolated and heroic development projects. For many reasons, from the unbundling of IBM's systems (Flamm, 1988) to the rise of personal computers and the Internet, organizational IT strategy has changed through the political economy of standards (Hawkins, Mansell, and Skea, 1995; Kahin and Abbate, 1995). A standard here is any written-down specification of how something will work, regardless of whether it was written down by a vendor, a private alliance of vendors, a national or global standards organization, or the government. The most important standards for present purposes are compatibility standards: written-down specifications of how things will work together -- specifications, that is, of technical interfaces (Bar, Borrus, and Steinberg, 1995; Lehr, 1995).

The organizational world, through its experience in buying and using standardized IT products, has developed a great deal of street knowledge about standards, and economists have lately formalized this knowledge in more abstract terms. One prominent feature of standards is lock-in: having bought equipment that conforms to a particular standard, future acquisitions are constrained to be backward-compatible (Arthur, 1989; David, 1985, 1993; cf. Liebowitz and Margolis, 1990). This creates an incentive for dominant vendors to minimize compatibility with their competitors' equipment, and an incentive for buyers to shop for equipment conforming to open standards, or to insist on discounts as a hedge against the expectation that future purchases will be artificially expensive (Shapiro and Varian, 1998). Given the considerable market advantages in control over product architectures (Morris and Ferguson, 1993), the competitive dynamics of these phenomena have by now been developed to a high art (Gabel, 1987; Grindley, 1995; Pesce, 1997). They can be understood as a negotiation by which vendors and customers, severally and in shifting alliances with one another, struggle to divide among themselves the value that is created in society by the establishment of a standard.

In cases of effective competition among technology vendors, perfect appropriability of standardization investments, and perfect information among technology consumers, these dynamics simply reflect an efficient market in action, albeit in a complicated way. It is a matter of great controversy, however, whether competition functions correctly in a market organized by compatibility standards (Katz and Shapiro, 1994; Liebowitz and Margolis, 1994). An analogy is frequently drawn to the telephone system. A prospective telephone customer, faced with two equally expensive but incompatible networks, one with a million customers and another with ten million, is likely to prefer the network that provides more parties to call. Thus a network that gains an advantage in subscribers, for whatever reason, is likely to have that advantage reinforced (Rohlfs, 1974; Rosenberg, 1994). The resulting positive feedback loop, it is said, makes telephone service a natural monopoly (Trebing, 1994).

Similar arguments apply to a wide range of so-called network goods, including networking protocols and operating systems -- goods whose value to a customer depends in large part on the number of other people who use them. At a minimum, competition in network markets requires the brutal combination of speed, luck, and the capital that is necessary to subsidize early subscribers until a critical mass is obtained (Katz and Shapiro, 1986). In the case of the Internet, the necessary subsidies were provided by universities nd certain parts of the US government, which was lucky enough to have little serious competition -- except, that is, for the exceptionally expensive-to-subsidize OSI.

CHANGING DYNAMICS

Strange and contested though they are, these observations about the dynamics of networked computing standards by now border on the commonplace. Yet they also have numerous deep consequences that are perhaps not so widely appreciated. For one thing, it is important to appreciate the several senses in which information technology, despite its revolutionary reputation, is in fact highly conservative (Agre, 1998b; cf. Callon 1993; Latour 1991). The demands of backward compatibility mean that legacy systems impose a conservative pull on future systems. Network competition, being prone to positive feedback effects, is a game for the lucky or the well-capitalized (Arthur, 1996). And standardized computer systems, like the custom-developed systems whose politics I mentioned above, are quite capable of inscribing misguided, backward, or biased ideas about the institutional environments in which they are used (Friedman and Nissenbaum, 1996; Hawkins, 1996; Mansell, 1996; Phillips, 1997). Standards dynamics do sometimes give the impression of revolution, on those occasions when a network standard reaches a sufficient critical mass to make it a sure bet for all of the other players who have been waiting on the sidelines (Economides and Himmelberg, 1995). But these seeming discontinuities must be understood in the context of the conditions under which such critical masses are obtained.

The Internet provides deeper clues about the emergent dynamics of a standards-driven world (Lemley, 1996). If the computing systems of an earlier day were driven largely by the internal dynamics of the organizations that acquired or developed them, in a standards-driven world the action must be understood in interorganizational terms. Although standards have always existed and always had consequences, it is now no exaggeration to say that computers are made out of standards, and that the construction and adoption of standards are collective decisions that are played out on a complex global stage, and not just on the level of the internal decision-making processes of individual organizations.

If it is the whole institutional world that adopts standards and not just particular organizations, then what kinds of standards does the world want to adopt? The Internet suggests one possible generalization. Among its numerous virtues, the Internet is functionally parsimonious. It is simple. It moves packets from point A to point B and does not even guarantee that the packets will arrive. The Internet's end-to-end principles encourage network architects to move complexity out of the Internet protocols and into Internet hosts -- or, put another way, into service layers qualitatively distinct from IP and whatever transport services it employs (Saltzer, Reed, and Clark, 1984; Tennenhouse, Lampson, Gillett, and Steiner, 1996; cf. Odlyzko, 1998).

The general principle here is parsimonious layering, and one intuition behind the principle is that a simple, general-purpose network protocol is equally compatible with a wide range of applications. Contrast this with the many special-purpose networking protocols, such as EDI (Brousseau, 1994), that the Internet is replacing, or that, like for example teletex (Jeppesen and Poulsen, 1994), never got off the ground. These special purpose networks can be understood by analogy to business terminology as stovepipe systems: systems that bundle together functionality on several levels that could have been taken apart into distinct and much more generally useful service layers. These stovepipe systems may have an early advantage because they solve a real organizational problem in a single package. But they are at great risk in the long run because their development and maintenance costs can only be amortized across those relatively few organizations that face that particular problem, and not across the much wider range of organizations that require some subset of the bundled functionality for some other purpose. Morris and Ferguson (1993: 92) describe this pattern with their dictum that "[g]eneral-purpose architectures absorb special-purpose solutions".

Here, then, is one general lesson that we might try to extract from the precedent of the Internet. Call it the platform cycle: large, organization- or sector-specific stovepipe applications are developed at great expense, and implemented with great difficulty, to solve organizational problems. Then a more general standard comes along that abstracts out some basic functionality in a way that is incompatible with the legacy system but is nonetheless broadly useful to other organizations. A great conflict results, being resolved at potentially great length by the construction of new, less-bundled applications on the now-standardized platform. The widespread adoption of that platform encourages the construction of many other large applications on top of it, and the cycle begins again. The result is a rising tide of layered standards, both for hardware and software, for data formats and protocols. Once established, each platform defines a (potentially quite uneven) competitive playing field for applications that might be built on top of it (cf. Meyer and Lehnerd, 1997), and in particular the newly modular platform architecture leads to modular competition and a modular organization of markets and vendor organizations (Baldwin and Clark, 1997; cf. Clark, 1985; Langlois and Robertson, 1992). Neat though it is in the eventual by and by, the platform cycle is immensely painful in the meantime. Organizations that delay adopting standards must build and maintain costly stovepipe systems, and organizations that too-hurriedly adopt standards before they achieve critical mass risk being catastrophically stranded if the market's winner-take-all dynamics end up settling on some other, incompatible platform standard.

And the problem is even worse. To the extent that information systems are bound up with organizational practices, interorganizational dynamics that exert a standardizing pressure on technology will also exert a like pressure on practices. This is obvious enough for those practices that are intimately tied to the workings of the machine, such as maintenance and data entry. But it is also true more basically for the ontological assumptions that a system makes -- assumptions, for example, about the nature of work processes, relationships and agreements, and the work materials that the system represents. As the platform cycle starts producing standardized service layers that support and constrain the actual substance of the work -- as for example in so-called enterprise systems (Gilbert, 1992) -- the progress of technology standardization becomes linked to progress in the standardization of everything else (cf. Quinn, 1992). This tension, too, holds the potential for significant conflict.

THE LAW OF OPPOSITE NUMBERS

To manage these conflicts, organizations must grasp their radical nature. The key is that powerful and qualitatively complicated forces that affect an organization's very structure and practices are being played out on a global stage, and prudent and responsible organizations, rather than simply ignoring or reacting to the forces, will seek to become players on that stage. To comprehend the kind of global participation that is required, it helps to observe that the dynamics of standards, even when organized wholly by the market, resemble the classical picture of the marketplace less than they resemble the democratic political process (Agre, 1998c; Schoechle, 1995; Weiss, 1993). To the extent that they embody and tend to reinforce substantive assumptions about work practices and relationships, and to the extent that they win out decisively against their competitors, compatibility standards resemble nothing so much as laws (Lessig, 1997; Reidenberg, 1998), and standards-driven markets resemble legislatures. These are legislatures, to be sure, in which campaign contributions count for a great deal, but that is nothing new.

More fundamentally, the legislative dynamics of standards reward interest group organizing among a standard's stakeholders, including both vendors and customers. Vendors that negotiate one-by-one with unorganized customers have a significant informational advantage, inasmuch as customers' bargaining power depends in significant part on what other customers are doing. In this light, the discourse of open systems (Cargill, 1994; Tassey, 1995) can be understood as a kind of social movement among information technology customers (cf. Davidow, 1986; Kling and Iacono, 1988). The point is not that these customers formed a cohesive political party, or (to use another metaphor) that they engaged in formal collective bargaining as a labor union might, but rather that they built a public consensus about the terms on which they wanted to purchase computers. The existence of this consensus created, if nothing else, the confidence among IT customers that they would not be stranded by purchasing open systems, because other customers would be doing the same thing. Expectations shape reality in a network marketplace, and expectations are created in large part through symbolic displays of unity among members of an interest group in the public sphere. When customers are not coordinated in this manner, vendors can create self-fulfilling prophesies through hype (Farrell and Saloner, 1986; Halliwell, 1993; Theoharakis, Wong, and Powell, 1998).

Of course, negotiations over the nature and meaning of the emerging norm of openness quickly became complicated as vendors and customers tussled over just what exactly the term "open" should encompass (Abbate, 1995). Nor is the ideological creation of an open systems movement sufficient by itself to ensure that open systems will actually take hold. Although supposedly detrimental to customers in a world where everything is equal, closed proprietary standards do enjoy the considerable advantage that their owners can, as I have mentioned, afford to subsidize early adopters, especially those influential early adopters to which other organizations look for self-fulfilling signs of leadership (Bank, 1997).

Observe that these vendor strategies work to the extent that customer organizations are dependent on others for guidance about market directions, as opposed to having organized among themselves to create a different self-fulfilling consensus. When uncoordinated customer organizations try to avoid the effort of coordinating among themselves by looking to a single influential leader for market guidance, they are effectively trying to free-ride on the research and learning efforts of the leader, and this suggests that the leader will experience suboptimal incentives to invest in research and learning. When customers free-ride in these ways, initiative for the shaping of standards once again returns to vendors, who can now rationally design their systems in a way that maximizes both the speed with which a critical mass can be assembled and the depth to which customer organizations are then locked in.

If the Internet helps create the conditions for this problem, it also, in its uncannily reflexive way, creates the conditions for a possible solution. For organizations looking to shape their future information systems to their own liking, rational strategy begins with what we might call the law of opposite numbers. Perhaps the most profound pattern in the use of the Internet is the construction or reinforcement of community among people who have things in common -- between people, in other words, and their opposite numbers. Parents in one family can compare notes with their opposite numbers in another family, and likewise reference librarians in universities, and principals in schools, and all manner of other geographically and organizationally dispersed groups of people who nonetheless occupy similar structural locations (Agre, 1998a).

The intuition here is mathematical: start with a sprawling structure -- all of the relationships among individuals in society -- and form a more compact structure by identifying all of the elements that are in some way equivalent -- so that all of society's truck drivers, for example, or ministers or cancer patients, end up being located in the same figurative "place". In each case we can ask how well each of these interest communities functions: how densely do the world's truck drivers, for example, know one another, communicate with one another, develop consensus opinions with one another, share information with one another, smooth out differences with one another, cultivate a shared identity with one another, and so on? Of course, one can ask these questions in a world without the Internet, but it is particularly interesting and pertinent to ask them in the world we have now, in which the Internet at least makes possible much higher levels of connectivity, community, and solidarity among those people who have something important in common.

In the context of organizational decision-making about information technology, these questions apply with special pertinence to each of the many types of people whose working lives are affected by the evolution of IT standards. How well networked are the systems maintainers in a given industry? How about the financial officers? The MIS managers? How able is each of these lateral communities to track the standards process, develop shared opinions about it, and intervene in it? How able are they to pass along lessons and warnings from their own experiences with emerging information technologies? And, ultimately, how able are they to act collectively to shape expectations and thus realities in standards-driven markets?

MANAGING THE STANDARDS ENVIRONMENT

The answer to these questions, of course, will be found somewhere between "none" and "a great deal". The evidence does not nearly exist to evaluate the matter empirically. My own purpose here, however, is not empirical but prescriptive. Based on the very tentative analysis that I have developed above, I want to draw out a framework that might guide organizations' efforts to manage their standards environments, both in order to shape their own destinies and to encourage positive values in design. If system development now effectively takes place on a global stage, it follows that customer organizations must participate on a global stage. And if the dynamics of that global stage resemble the dynamics of a democratic political process, then participation in those dynamics will resemble political action. In particular, they will resemble what public relations people have taken to calling issues management (Heath, 1988): not simply the reactive resolution of public relations crises, but the proactive tracking and shaping of issues in the public sphere that hold the potential to affect the organization's interests.

Organizations might exhibit several levels of participation in global standards dynamics, depending on the depth of their commitment to consensus-building within the law of opposite numbers. I will distinguish four levels of participation.

Level one consists simply of identifying and building professional relationships with those opposite numbers. As each individual reaches out to compare notes with his or her opposite numbers in other organizations, it becomes possible to engage in surveillance of the standards environment -- what standards are being adopted by whom? -- and to transfer skills -- what have others learned about the practicalities of adopting those standards, and what best practices are beginning to emerge as a result? To the extent that effective adoption of new technology entails institutional learning, networking with opposite numbers will promote the diffusion of that learning.

On level two, organizations use what they have learned from their opposite numbers in shaping their own information technology strategies. To the extent that the wisdom of adopting a standard depends in significant measure on one's opposite numbers adopting it as well, networking among opposite numbers can encourage more rational timing of adoption decisions, thereby lessening both the costs of outdated and incompatible legacy systems and the risk of stranding from premature commitments to new systems. More generally, an informed over-the-horizon sense of the direction and pace of industry-wide standards adoption can provide an organization with the lead time it will need in assessing whether technology trends are going to require more profound shifts in organizational practices and strategies, and to get started on any necessary shifts at less than panic speed.

The third level entails a more proactive involvement in the standards-setting process itself. Nonproprietary standards being public goods (Kindleberger, 1983; Steinmueller, 1995; Wagner, Cargill, and Slomovic, 1995), most standards organizations have experienced great difficulty in their attempts to involve users in their deliberations (Dankbaar and van Tulder, 1992; Foray, 1995; Jakobs, Procter, and Williams, 1996). Nonetheless, organizations' needs for advance warning of relevant standards, as well as their need for substantively congenial standards -- that is, standards that are not grossly incompatible with their preferred ways of doing business -- ought to motivate progressive user organizations to commit the resources and develop the expertise to at least somewhat counterbalance the generally overwhelming hand of vendors in shaping standards. When standards are proprietary, likewise, level 3 management of the standards environment will entail working with vendors, either providing input to the design process or at least testing alpha releases of the technology.

The fourth and highest level of management of the standards environment entails playing a leadership role in one's community of opposite numbers. This does not primarily consist of heavy unilateral investments in learning and evaluation. Instead, it implies organizing the community and building consensus around a mutual commitment to at least level-3 participation in the shaping of relevant standards. If information and learning are pooled then this participation need not be expensive, and the organization that chooses to play a leadership role will presumably still benefit from its proportionally deeper knowledge of the overall process and its dynamics. Building a consensus of this sort, of course, will be more difficult when one's opposite numbers are competitors, or when cooperation with them is frustrated by cultural or geographic barriers. But managing this kind of complex relationship is what success in an increasingly globalized economy and society requires (Brandenburger and Nalebuff, 1996).

CONCLUSION

In sketching this four-level hierarchy of commitments to managing an organization's standards environment, I have come a long way from the "cyberspace" understanding of the Internet as a revolutionary force that replaces or transcends an existing institutional order. The Internet, it turns out, is one manifestation of a much deeper phenomenon -- one that interacts with institutional orders in a profound and pervasive way. The outcome of these interactions cannot be read off of the workings of the technology. Quite the contrary, the technology itself is going to be shaped by thoroughly political processes whose dynamics will depend on the relative positioning of a great diversity of players, as well as upon the cultural and intellectual factors that enable the stakeholders in a given standard to unite around their preferred expectations for that standard, or else to disperse into their individual and isolated bargaining positions in the face of powerful and counterintuitive economic phenomena. The outcome of these technology-shaping processes will affect the fortunes of particular organizations, and that is important enough. But to the extent that social life more generally comes to be mediated by advanced information technologies, the outcome will also, in a profound way, set the rules of social life. Thus the urgency of the law of opposite numbers and its many practical consequences.

The consequences for technology policy are complex. The platform cycle offers qualitative predictions about the direction of standards evolution, and in particular it suggests that different strategies should apply at different points in the cycle. Leverage over the resulting architecture will be greatest toward the beginning of the cycle, in the process of facilitating a consensus about the nature of the common abstract functionality that a new service layer could embody. In the next phase, the priority is mediating the potential conflict between technologies that that are already locked in to old-generation stovepipe applications and new technologies that the platform can support. To the extent that platform standards arise through a collective negotiation between vendors and customers whose dynamics resemble a democratic political process, the "Constitution" of that process can be considered openly. Critiques of the democratic process should be useful, for example the collective action problems that benefit focused interest groups (e.g., Olson, 1965), as should some of the suggestions for systemic reform (e.g., Greider, 1992).

REFERENCES

Abbate, J., 1995. "Open systems" and the Internet, Society for Social Studies of Science / Society for History of Technology Conference, Charlottesville, October.

Agre, P. E., 1998a. Designing genres for new media. In: S. G. Jones (Editor), CyberSociety 2.0: Revisiting CMC and Community, Newbury Park, CA: Sage.

Agre, P. E., 1998b. Yesterday's tomorrow. Times Literary Supplement, 3 July, pages 3-4.

Agre, P. E., 1998c. The Internet and public discourse. First Monday 3(3).

Arthur, W. B., 1989. Competing technologies, increasing returns, and lock-in by historical events. The Economic Journal 99, pages 116-131.

Arthur, W. B., 1996. Increasing returns and the new world of business. Harvard Business Review 74, pages 100-109.

Baldwin, C. Y., Clark, K. B. 1997. Managing in an age of modularity. Harvard Business Review 75, 84-93.

Bank, D., 1997. How one sweet deal unraveled for Netscape after Microsoft called. Wall Street Journal, 13 November, page A1, A10.

Bar, F., Borrus, M., Steinberg, R., 1995. Islands in the bit-stream: Charting the NII interoperability debate, Working Paper 79 (Berkeley Roundtable on the International Economy).

Brandenburger, A. M., Nalebuff, B. J., 1996. Co-opetition, New York: Doubleday.

Brousseau, E., 1994. EDI and inter-firm relationships: Toward a standardization of coordination processes?. Information Economics and Policy 6, 319-347.

Brynjolfsson, E., Hitt, L. M., 1998. Beyond the productivity paradox: Computers are the catalyst for bigger changes. Communications of the ACM 41, 49-55.

Callon, M. 1993. Variety and irreversibility in networks of technique conception and adoption. In: D. Foray and C. Freeman (Editors), Technology and the Wealth of Nations: The Dynamics of Constructed Advantage, London: Pinter.

Cargill, C., 1994. Evolution and revolution in open systems. StandardView 2, 3-13.

Clark, K. B., 1985. The interaction of design hierarchies and market concepts in technological evolution. Research Policy 14, pages 235-251.

Dankbaar B., van Tulder, R., 1992. The influence of users in standardization: The case of MAP. In: M. Dierkes and U. Hoffman (Editors), New Technology at the Outset: Social Forces in the Shaping of Technological Innovations, Frankfurt: Campus Verlag.

Danziger, J. N., Kling, R., Dutton, W. H., 1983. Computers and Politics: High Technology in American Local Governments, Columbia University Press.

Davenport, T. H., 1997. Information Ecology: Mastering the Information and Knowledge Environment, Oxford University Press.

David, P. A., 1985, Clio and the economics of QWERTY. American Economic Review 72, 332-337.

David, P. A., 1993. Path-dependence and predictability in dynamic systems with local network externalities: A paradigm for historical economics. In: D. Foray and C. Freeman (Editors), Technology and the Wealth of Nations: The Dynamics of Constructed Advantage, London: Pinter.

Davidow, W. H., 1986. Marketing High Technology: An Insider's View, New York: Free Press.

Economides, N., Himmelberg, C. 1995. Critical mass and network evolution in telecommunications. In: G. W. Brock (Editor), Toward a Competitive Telecommunication Industry: Selected Papers from the 1994 Telecommunications Policy Research Conference, Mahwah, NJ: Erlbaum.

Farrell, J., Saloner, G., 1986. Installed base and compatibility: Innovation, product preannouncements, and predation. American Economic Review 76(5), 1986, pages 940-955.

Flamm, K., 1988. Creating the Computer: Government, Industry, and High Technology, Washington: Brookings Institution, 1988.

Foray, D., 1995. Coalitions and committees: How users get involved in information technology (IT) standards. In: R. Hawkins, R. Mansell, and J. Skea (Editors), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Edward Elgar.

Friedman, B., Nissenbaum, H., 1996. Bias in computer systems. ACM Transactions on Information Systems 14, 330-347.

Frook, J. E., 1998. Linking the supply chain with the cash register. InternetWeek 709, 6 April, pages S7-S8.

Gabel, H. L. (Editor), 1987. Product Standardization and Competitive Strategy, Amsterdam: North-Holland.

Gilbert, R. J., 1992. Symposium on compatibility: Incentives and market structure. Journal of Industrial Economics 40, 1-8.

Greider, W., 1992. Who Will Tell the People? The Betrayal of American Democracy, New York: Simon and Schuster.

Grindley, P., 1995. Standards Strategy and Policy: Cases and Stories, Oxford: Oxford University Press.

Halliwell, C., 1993. Camp development: The art of building a market through standards. IEEE Micro 13, 10-18.

Halper, M., 1998. Middlemania. Business 2.0 3, 45-60.

Hawkins, R., Mansell, R., Skea, J. (Editors), 1995. Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Edward Elgar.

Hawkins, R., 1996. Standards for communication technologies: Negotiating institutional biases in network design. In: R. Mansell and R. Silverstone (Editors), Communication by Design: The Politics of Information and Communication Technologies, Oxford: Oxford University Press.

Heath, R. L. (Editor), 1988. Strategic Issues Management: How Organizations Influence and Respond to Public Interests and Policies, San Francisco: Jossey-Bass, 1988.

Jakobs, K., Procter, R., Williams, R., 1996. Users and standardization: Worlds apart? The example of electronic mail. StandardView 4, 183-191.

Jeppesen, S. E., Poulsen, K. B., 1994. The text communications battlefield: Installed base, externalities and the fall of the teletex system. Telecommunications Policy 18, 66-77.

Kahin, B., Abbate, J. (Editors), 1995. Standards Policy for Information Infrastructure, Cambridge: MIT Press, 1995.

Katz, M. L., Shapiro, C., 1986. Technology adoption in the presence of network externalities. Journal of Political Economy 94, 822-841.

Katz, M. L., Shapiro, C., 1994. Systems competition and network effects. Journal of Economic Perspectives 8, 93-115.

Kindleberger, C. P., 1983. Standards as public, collective and private goods. Kyklos 36, 377-396.

King, J. L., 1996. Where is the payoff from computing?. In: R. Kling (Editor), Computerization and Controversy: Value Conflicts and Social Choices (2nd ed.), Academic Press.

Kling, R., Iacono, S. 1988. The mobilization of support for computerization: The role of computerization movements. Social Problems 35, 226-243.

Kling, R., Iacono, S. 1989. The institutional character of computerized information systems. Office: Technology and People 5, 7-28.

Kling, R., 1989. Theoretical perspectives in social analysis of computerization. In: Z. W. Pylyshyn and L. J. Bannon (Editors), Perspectives on the Computer Revolution, second edition, Norwood, NJ: Ablex.

Kling, R., 1991. Computerization and social transformation. Science, Technology, and Human Values 16, 342-367.

Kling, R. (Editor), 1996. Computerization and Controversy: Value Conflicts and Social Choices, second edition, Academic Press.

Knights, D., Murray, F. 1994. Managers Divided: Organisation Politics and Information Technology Management, Chichester, UK: Wiley.

Landauer, T. K., 1995. The Trouble with Computers: Usefulness, Usability, and Productivity, Cambridge: MIT Press.

Langlois, R. N., Robertson, P. L., 1992. Networks and innovation in a modular system: Lessons from the microcomputer and stereo component industries. Research Policy 21, 297-313.

Latour, B., 1991. Technology is society made durable. In: J. Law (Editor), A Sociology of Monsters: Essays on Power, Technology and Domination, London: Routledge.

Lehr, W., 1995. Compatibility standards and interoperability: Lessons from the Internet. In: B. Kahin and J. Abbate (Editors), Standards Policy for Information Infrastructure, Cambridge: MIT Press.

Lemley, M. A., 1996. Antitrust and the Internet standardization problem. Connecticut Law Review 28, pages 1041-1094.

Lessig, L., 1997. What things regulate speech, paper presented at the 1997 Telecommunications Policy Research Conference, Alexandria, VA, available online at .

Liebowitz, S. J., Margolis, S. E., 1990. The fable of the keys. Journal of Law and Economics 33, pages 1-25.

Liebowitz, S. J., Margolis, S. E., 1994. Network externality: An uncommon tragedy. Journal of Economic Perspectives 8, 113-150.

Madrick, J., 1998. Computers: Waiting for the revolution. New York Review of Books, 26 March, pages 29-33.

Mansell, R., 1996. Designing electronic commerce. In: R. Mansell and R. Silverstone (Editors), Communication by Design: The Politics of Information and Communication Technologies, Oxford: Oxford University Press.

Meyer, M. H., Lehnerd, A. P., 1997. The Power of Product Platforms: Building Value and Cost Leadership, New York: Free Press.

Morris, C. R., Ferguson, C. H., 1993. How architecture wins technology wars. Harvard Business Review 71, 86-96.

Odlyzko, A. M., 1998. Smart and stupid networks: Why the Internet is like Microsoft. ACM netWorker 2, 38-46.

Olson, M., Jr., 1965. The Logic of Collective Action: Public Goods and the Theory of Groups, Cambridge: Harvard University Press.

Orlikowski, W. J., 1993. Learning from Notes: Organizational issues in groupware implementation. The Information Society 9, 237-250.

Pesce, M., 1997. The great leap downward, Feed, February.

Phillips, D. J., 1997. Cryptography, secrets, and the structuring of trust. In: P. E. Agre and M. Rotenberg (Editors), Technology and Privacy: The New Landscape, Cambridge: MIT Press, 1997.

Quinn, J. B., 1992. Intelligent Enterprise: A Knowledge and Service Based Paradigm for Industry, New York: Free Press.

Reidenberg, J., 1998. Lex Informatica: The formulation of information policy rules through technology. Texas Law Review 96, 1998, pages 553-593.

Rohlfs, J., 1974. A theory of interdependent demand for a communications service. Bell Journal of Economics 5, 16-37.

Rosenberg, N., 1994. Telecommunications: Complex, uncertain, and path dependent. In: Exploring the Black Box: Technology, Economics, and History, Cambridge: Cambridge University Press.

Saltzer, J. W., Reed, D. P., and Clark, D. D., 1984. End-to-end arguments in system design. ACM Transactions in Computer Systems 2, 277-288.

Schmidt, K., Bannon, L. 1992. Taking CSCW seriously: Supporting articulation work. Computer Supported Cooperative Work 1, 7-40.

Schoechle, T., 1995. The emerging role of standards bodies in the formation of public policy. IEEE Standards Bearer 9, 1, 10.

Shapiro, C., Varian, H., 1998. Information Rules: A Strategic Guide to the Network Economy, Boston: Harvard Business School Press.

Sproull, L., Kiesler, S., 1991. Connections: New Ways of Working in the Networked Organization, Cambridge: MIT Press

Steinmueller, W. E., 1995. The political economy of data communication standards. In: R. Hawkins, R. Mansell, and J. Skea (Editors), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Edward Elgar.

Stiroh, K. J., 1998. Computers, productivity, and input substitution. Economic Inquiry 36, 175-191.

Tassey, G., 1995. The roles of standards as technology infrastructure. In: R. Hawkins, R. Mansell, and J. Skea (Editors), Standards, Innovation and Competitiveness: The Politics and Economics of Standards in Natural and Technical Environments, Edward Elgar.

Tennenhouse, D., Lampson, B., Gillett, S. E., Klein, J. S., 1996. Virtual infrastructure: Putting information infrastructure on the technology curve. Computer Networks and ISDN Systems 28, 1769-1790.

Theoharakis, V., Wong, V., Powell, P., 1998. Strategic implications of hype on the adoption of innovative technologies: The case of local area networks. In: T. J. Larsen and E. McGuire (Editors), Information Systems Innovation and Diffusion: Issues and Directions, Hershey, PA: Idea Group.

Trebing, H. M., 1994. The networks as infrastructure: The reestablishment of market power. Journal of Economic Issues 28, 379-389.

Wagner, C. S., Cargill, C. F., Slomovic, A. 1995. Open systems standards in manufacturing: Implications for the National Information Infrastructure. In: B. Kahin and J. Abbate (Editors), Standards Policy for Information Infrastructure, Cambridge: MIT Press.

Weill, P., Broadbent, M. 1998. Leveraging the New Infrastructure: How Market Leaders Capitalize on Information Technology, Boston: Harvard Business School Press.

Weiss, M. B. H., 1993. The standards development process: A view from political theory. StandardView 1, 35-41.