The Global Diffusion of Computing: Issues in Development and Policy
Seymour E. Goodman
The Global Diffusion of Computing
Twenty years ago, computing was very much an American domain. A handful of American companies manufactured most of the computers in the world, and annual machine production of the biggest companies measured in the tens of thousands. Most of this equipment stayed in the United States, and little of the rest moved outside of industrial countries, most of which were American allies. Foreign manufacturers (mainly in Western Europe, Japan, and the former Warsaw Pact nations) paled in comparison. The ARPANET had about a hundred hosts, with minimal connectivity with England and Norway, and no comparable network existed outside the United States. Computers were sparsely scattered outside the First and Second Word nations.
Today's computing world is profoundly changed. Global production of microprocessors and computers amounts to tens of millions of dollars annually. Capable manufacturers exist in many places around the world, including several nations that were technologically backward in the early days of computing. The United States itself imports a lot of computing equipment, mostly from East Asia and from the foreign manufacturing sites of American companies. International sales of many important American companies amount to about 50 percent of total sales. The ARPANET has evolved into the Internet, which has about 10 million hosts. The non-U.S. share of these hosts exceeded 50 percent for the first time in 1995 (the foreign share of computers in use passed that milestone about 1990). More than 170 countries have some worldwide network connectivity, including Third and Fourth World countries such as Bangladesh, Mongolia, Mozambique, Peru, and Romania. Today, there are about half as many computers on the planet as there are cars, trucks, and buses.
More generally, information technology (IT) has become closely associated with modernization, with being part of the global economy, and—in one form or another—is considered necessary or highly desirable almost everywhere. According to some, IT will help bring the world together, paving the way for economic development and even sweeping aside the traditional borders (and governments) that divide people. (We might also note that the global diffusion of IT is hardly uniform and that, in some ways, IT is causing increased stratification.)
NOTE: L. T. Greenberg also contributed to the content of this presentation.
Players on a Much Expanded Field
The United States has had an enormous role in stimulating and enabling this change. American contributions include cheap and powerful hardware; great quantities of (often "free") software; networking; American buying and importing habits (which recently include a trade deficit in IT products); the "Mecca" of the U.S. technical education community; and the fruits of military and space research and competition (most notably, from a global standpoint, the Internet and global positioning system). The United States still comprises about 40 percent of the world's computing market.
Although its relative position has declined, the United States still has almost as much computing as the rest of the world together. This country is now the most important player in a much bigger game. As such, the United States is more prominent, accomplished, and extensively involved in computing, and more people in the rest of the world have direct contact with the United States than ever before. Thus, more foreigners than ever before care about what is done by the United States.
As important as the United States is in global IT, there are also many other players. Some are already quite substantial and others are becoming so. These players are not just governments of countries that now have a serious presence in the global IT scene and are not just foreign IT companies. Rather, the spread and application of IT has been such that many nongovernmental, non-IT industry stakeholders have emerged or have been strengthened. Their involvement results from their interests in the applications of IT (including its "application" as their exports), and they are empowered because of their wealth, constituencies, and newfound voices (often enabled by IT). The interests and the numbers of these actors also create new, or exacerbate old, conflicts and issues. These stakeholders cover a much broader spectrum of business, academic, nongovernmental organization, and private citizen actors than ever. There is even a larger set of stakeholders within governments. Furthermore, the global diffusion of IT has weakened some older stakeholders, and they have not all taken their diminution quietly.
To illustrate how international changes in value conflicts and stakeholders have affected a major technological policy issue over the past 10 years, consider U.S. national security and foreign policy export control. This is arguably the earliest, longest-lived, computer-related U.S. public policy issue, and one with an enormous footprint across the U.S. government. Export control was the subject of one of the first major Computer Science and Telecommunications Board (CSTB) studies,1 as well as one of its most recently completed major studies. We limit this discussion to export controls on computers, particularly high-performance computers, but a similar short analysis might also be made with respect to the control of encryption.
First and foremost, what has changed in the world is that with the demise of the Soviet Union, the peer threat pursuing a full spectrum of advanced military systems has all but disappeared. This is not to say that there are no significant threats to the United States from foreign sources who may benefit from enhanced computing capabilities, but these threats are not in the same league as the earlier possibilities for global superpower conflict. Taken together with the realities of technological progress and diffusion, this means that advocates of controls find it harder to justify the continuation of controls on the basis of specific applications that potential enemies might be able to pursue using computer products that might still be effectively controlled.
To appreciate how difficult effective control has become, we note that in early 1993, a raw computing performance threshold of 195 MTOPS (millions of theoretical operations per second; since appropriately revised upward twice by the U.S. government) was used to define a control regime for "supercomputers." Today, millions of microprocessors are produced each year with performance levels of about half to four times that value. There is now so much made-in-the-U.S.A. computing in so many parts of the world that it is unfortunately inevitable, but essentially unpreventable, that some of it will be used against U.S. national security interests.
Furthermore, there is now a greater division of value judgments within the national security group of stakeholders. For example, some Department of Defense beneficiaries of COTS (commercial, off-the-shelf) high-performance computing (HPC) see more benefit to what they do and to U.S. national security from a strong, technologically vigorous, domestic industry than from continued export controls.
The extent to which foreign governments, even those of our closest and former Coordinating Committee for Multilateral Export Controls (COCOM) allies, consider the value of export controls covering what are increasingly COTS dual-use technologies is not obvious. None gives it the same level of visible attention. When computing was much more exclusively in the U.S. domain and these allies shared a common threat to survival, it was easier for all to agree to controls. In today's world, it is unimaginable how an effective unilateral control regime might be constructed.
Turning now to stakeholders in business and industry, we can find several factors that expand the set of participants. One is that technological progress has been such that HPC, at least by the early 1993 export control definition above, is no longer limited to the extreme high end of the total industry product line. It now covers a substantial fraction of the computers in the world, and thereby many of the computer makers, a significant fraction of which are not on U.S. soil. Furthermore, about half of the new computer sales in the world are not made on U.S. soil.
In many countries, the promotion of business interests is a basic part of national economic policy. It is even fashionable for many countries to be formulating explicit national informatics policies. In the United States, which does not have a comprehensive national industrial policy, there are, nevertheless, many government players promoting business interests at the national, state, and local levels. With many government stakeholders and their constituents, the concern for jobs and exports carries more weight these days than claims of computer-based military threats (although computer-based security problems have been attracting a much greater level of concern in the business and privacy arenas, partly as a result of global diffusion).
Today's world of computer sellers and buyers—including huge multinational corporations (MNCs), many technologically capable small companies, and an immense consumer community ranging from governments to other large MNCs down to millions of small organizations and individuals—constitutes a large and growing, push-pull combination that is under pressure from intensifying global competition. As in the drug trade, there is a huge traffic among willing and resourceful buyers and sellers, and it is difficult for third parties with conflicting interests to interfere with this, for example, through export controls, tariffs, or import prohibitions. Some vast and wealthy industries (most notably the entertainment industry) depend increasingly on IT as a vehicle to do their business worldwide, adding a powerful "pull" element to the "push" interests. As IT diffuses more widely and becomes increasingly pervasive in human activities everywhere, this buyer-seller community gains in importance and influence.
There are other parties with stakes in the international diffusion of computing. Most are recent entrants, and almost all favor the diffusion of computing and related technology transfers. These include various players in developing countries who see computing as necessary for development and as a way to join the rest of the world. Others see IT, in the words of Ithiel de Sola Pool, as the "technologies of freedom," supporting openness, civil and human rights, and democracy; encouraging American values; and making it harder for tyrants to hide abuses. The media also ride on this track, promoting freedom-of-the-press values (and for other reasons, for example, to sell entertainment and to provide better facilities for their reporting).
People in these categories might argue that instead of controlling the infusion of computing into Eastern Europe and the Soviet Union in the late 1980s, the United States should have tried to pour it in. They might further argue that the infusion of IT did more to bring about the decline of overall Soviet power than export controls at that time did to limit its military growth.
The academic community tends to favor transferring more IT. This viewpoint arises in a large number of contexts, including support for the reasons discussed above and for other reasons such as promoting academic freedom, encouraging privacy and security through encryption, participating in efforts to bring useful technologies to colleagues in other countries, and forming virtual scientific communities. American academic stakeholders are also increasingly dependent on foreign student enrollments to keep their departments viable. Foreign shares of graduate student enrollments in science and engineering departments in U.S. universities amount to 30 percent or higher. Simple demographics would indicate that faculty shares may follow suit. Such training constitutes one of the most effective forms of technology transfer, and it is being done on a large scale. For example, it is estimated that on the order of 30,000 to 40,000 Chinese students are studying in the United States. How does one justify a
major regime for controlling COTS hardware products under such circumstances? It seems like locking doors in a building when the walls are falling down.
Export control is but one example of a major technology issue that has been dramatically affected by the global diffusion of computing and the attendant expansion of stakeholders. The spread of global networking in various forms is creating or greatly exacerbating a number of issues that are attracting, or will attract, further attention. For example, many things from American political correctness to law enforcement practices to intellectual property rights to Islamic religious tenets may be seen as local sensitivities that are increasingly to be battered in a global multimedia free-for-all. Many forms of conflict, both civil and military, are finding fertile international or transnational breeding grounds in the IT-based media.
Many will not like the redistribution of—or compromise of, or assault on—moral or legal sensitivities, jobs, markets, wealth, influence, military power, and governmental authority via the international use of IT. Many of these will be influential stakeholders who will seek help in various ways from their governments or international organizations. There are notable cases in which this has already happened. IT may also enable them, as well as those they oppose, to band together to seek help or assert influence independent of their governments.
CSTB studies, and those of the National Research Council more generally, essentially have tried to bring science- and technology-related trends and issues to the attention of interested constituencies, and the attendant analyses have tried to inform or guide choices for policy formulation. To a large extent, the latter is a matter of balancing or choosing between value conflicts among these constituencies. The primary funders, and the principal targets as consumers of these studies, have been parts of the U.S. government.
The activity levels and pressures on national policymakers in these matters are likely to increase, as will some frustration levels. It appears possible that the relative power of national governments in these matters will diminish, but certainly not disappear, and that governments will have to contend with and work with other players as never before. The activity and frustration levels of advisers to governments, such as the CSTB, may also increase substantially.
The main purpose of this short presentation has been to draw attention to the greater internationalization of both the issues and the affected stakeholders that is accompanying the global diffusion of IT. This is not a short-term or transient phenomenon; it is going to be forever.