Technological Reflections: The Absorption of Data Communications in the Soviet Union

A dissertation by Joel M Snyder, in partial fulfillment of the requirements for a Doctor of Philosophy in the Department of Management. Presented June 1, 1993. This entire document is copyright Joel M Snyder and may not be copied or otherwise stored without written permission. Brief quotations from this thesis are allowable without special permission, provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part must be obtained from the author.

Abstract

The breakup of the Soviet Union into fifteen autonomous republics marked the end of an era of atomic superpowers born in the first half of the twentieth century. As the Communist Party relinquished its hold on the reigns of power, the Soviet Union changed in profound ways, economically, politically, and socially. Strongly isolationist policies which kept the U.S.S.R. separate from its neighbors in Western Europe and North America loosened significantly. Those isolationist policies encouraged a Soviet technological and industrial economy based almost entirely on locally developed materials and expertise---an economy which Western analysts found inferior in technological development, manufacturing capabilities, and absorption of information technologies in comparison to other industrialized nations.

Networks can be a metric to measure technological capabilities and absorption. Networks cannot be a priority project of a single ministry: they depend on hardware, software, training, and telecommunications infrastructure throughout the country. Thus, they act as an indicator of the capability of the economy to develop, distribute, and absorb new technologies. The absorption of networks indicates the capability of an economy to absorb similar new and recently-developed technologies. Networks are valuable tools for inter-organizational and international information transfer. How the Soviets use networks both internally and in external communications can indicate the amount of change, both in attitude and implementation. This study examines the development, manufacture, dissemination, and absorption of computer network technologies in two environments: the pre-1990 Soviet Union and the post-1990 former Soviet republics.

This study relies on detailed technical examination of the manufacturing technology, equipment choices and capabilities, and observed installation and use. In situ visits, reviews of open literature, interviews with Soviets, and, above all, networks themselves, are woven together to form a technological picture of how networks were, are, and can be used.

Using a model for the use and absorption of computer networks, this study presents extensive evidence showing the status of the former Soviet republics. It is concluded that:

• Changes in the post-U.S.S.R. economy have been to the detriment of Soviet network development and manufacturing capabilities;
• Absorption of computer networks is largely restricted to a few cities and republics; and
• Growth of computer networks has been explosive, although the total scale of absorption remains very small.

Introduction

Having written The End'' to this story of my life,
I find it prudent to scamper back here to before the
beginning, to my front door, so to speak, and to make
this apology to arriving guests: I promised you an
autobiography, but something went wrong in the kitchen.''
-- Bluebeard (Kurt Vonnegut, 1987)

Background for this Research

In 1937, International Business Machines Corporation was approached by a professor at Harvard University, Howard Aiken. He wanted to build what had not been successfully built before: a stored program digital computer. The Automatic Sequence Controlled Calculator, later known as the Harvard Mark I, built at IBM's Endicott Development Laboratories, became operational in August, 1944. [Randell73]

On August 14, 1945, the final formal hostilities of World War II were ended when Japan surrendered to the Allied nations. The political and economic structure left in place after six years of European war carved the world up into super powers, pivoting around the United States and the NATO nations, and the Soviet Union and the Warsaw Pact nations. (Footnote: NATO and Warsaw Pact, as formal treaty organizations, were established much later. However, the core players in the bilateral division after World War II were established almost immediately.) [Stein66]

Sometime between 1937 and 1944, the age of the digital computer began.

Over the next twenty years, Western nations, particularly the United States, embraced this new technology and set forth industrial and federal research programs which were aimed at understanding, developing, and capitalizing on information technologies. While the Soviet Union began with nearly the same fervor, other priorities soon drew attention from these projects. The centrally planned (command) economy they had developed was not working. The war had devastated the population, national coffers, industrial and agricultural output. Realistically, it was not until the mid--1950s that the Soviet Union tottered back from the brink of starvation. [Shmelev89]

A 1988 report prepared for the United States (US) National Research Council (NRC) drew a broad brush across the computing canvas of the Soviet Union from 1955 to 1985 and found it sorely lacking:

... It is clear that there are few, if any, substantial areas in computing in which the Soviets have achieved parity or are ahead of the west. ... The Soviet computing industry has failed to provide a microelectronics base sufficient to match Western developments ... The shortage of high-speed computing capabilities is particularly acute ... The Soviets also have far, far fewer personal computers than the United States ... [Goodman88, p. 213]
The eventual importance of information technologies could not have been predicted in 1945. In 1992, however, their significance can hardly be de-emphasized. The former Soviet Union had and the newly independent republics have substantial economic woes. (Footnote: The nomenclature here is particularly difficult. For the sake of convenience, I will refer to the geographic and political region which used to be known as the Soviet Union by that term when speaking of events and situations which occurred before 1992. For more current discussions, I will use the best name which represents them: former Soviet republics.''] (Figures~\ref{income-growth} and~\ref{industrial-growth} show this rather graphically.) Many factors have contributed to their low productivity of labor, stunted industrial growth, and lack of economic competitiveness. One of these is the failure to produce, disseminate, and absorb many different kinds of technology.

\begin{figure}[htbp]
\scaledpicture 7.25in by 5.03in (income scaled 750)

(Sources: \cite{Hewett88}; \cite{Economist91})
\caption{Income and Consumption Growth}
\label{income-growth}
\end{figure}

\begin{figure}[htbp]
\scaledpicture 7.29in by 4.14in (industry scaled 750)

(Sources: \cite{Hewett88}; \cite{Economist91})
\caption{Industrial Growth}
\label{industrial-growth}
\end{figure}


Structure and Research Question

This study is one of a series which examine the invention, manufacture, distribution, and use of information technologies in the Soviet Union.

In examining the absorption of the information technologies in the Soviet Union, both before and after the breakup into independent republics, I will focus study on a single area: computer networks. This area is particularly appropriate for several reasons:

• Networking technology is heavily dependent on infrastructure, more so than other information technologies. While the focus is on computer networking technology, the greater dependence of networking on other elements: computer systems, trained users, local and long distance telephone lines, etc., gives a greater confidence level that the area of computer networks is not isolated, but can actually provide representative conclusions for the absorption and dissemination of information technologies as a whole.
• Both absorption and dissemination occur, and are discernible, in computer network technology. In an environment such as the Soviet Union, widespread physical dissemination of a technology tells little about the actual absorption of that technology. In a capitalist economy, where organizations have to allocate scarce resources to acquire products, acquisition generally correlates with usage. In the Soviet economy, where products are delivered to organizations in a planned manner, the presence of information technology equipment has little correlation with the use of that equipment. However, because inter-organizational computer networks have obvious and significant usage characteristics, it is not difficult to measure the real absorption of networking technology.
• Computer networks are a common technology, and see widespread use in many environments in the West. Other technologies, such as supercomputers or artificial intelligence, can tell more about the state-of-the-art of Soviet research. Widespread technologies, on the other hand, tell more about the impact of information technologies.
• Computer networking technology will measure both vertical (innovative and inventive developments) and horizontal (production based on externally injected or solicited technology) technology transfers. Since the technology of computer networks is widely available, inexpensive, and well reported in the scientific literature, an opportunity for both types of technology transfer will occur. The additional benefit is that it is possible to see the mix of horizontal and vertical transfers in this technology area.
The goal of this study is to bring an in-depth understanding of the capabilities of the Soviet Union, both before and after the breakup into independent republics, in the manufacture, development, technology transfer, dissemination, and absorption of computer network technologies. Specifically, I hope to answer the following:

1. What were the capabilities of the Soviet Union (and what are the capabilities of the former Soviet republics) to develop indigenous network hardware and software products? What were the manufacturing capabilities? What quantities of products have actually been produced?
2. To what extent did the Soviet Union (and do the former Soviet republics) rely on foreign technologies, both proprietary and public, in distribution and manufacture of networking hardware and software?
3. How widespread were computer networks in the Soviet Union are they in the former Soviet republics? To what extent have these technologies been absorbed into enterprises and institutions? How has the absorption of these technologies changed over time, especially given recent events?
4. What prospects for manufacture, development, dissemination, and absorption of computer network technologies are there in the former Soviet republics? How does this differ from the prospects of other industrialized nations such as West Germany, Japan, France, and the United States?

Structure of the Study

This study is composed of eleven chapters. The first provides some background on why computer networks are an appropriate technology, and gives the research questions to be investigated.

Chapter 2 surveys prior research in this area, from both a technological and an economic point of view. It also presents the methodology to be used, and the analytical techniques which will make up the core of the study.

Chapters 3 and 4 discuss the absorption of computer networks and related data communications technology in the Soviet Union, starting from roughly 1985 and continuing through the end of 1989.

Chapter 5 presents a model for analysis of the state of networks in the former Soviet republics.

Chapters 6, 7, 8, 9, and 10 examine the state of computer networking technology in the former Soviet republics along the lines of the the model proposed in Chapter 3.

Chapter 11 presents conclusions of the study, and answers the research questions raised in this chapter.

Conclusions

Although written many years ago, Lady Chatterley's Lover has just been reissued by the Grove Press, and this pictorial account of the day-to-day life of an English gamekeeper is full of considerable interest to outdoor minded readers, as it contains many passages on pheasant-raising, the apprehending of poachers, ways to control vermin, and other chores and duties of the professional gamekeeper. Unfortunately, one is obliged to wade through many pages of extraneous material in order to discover and savour those sidelights on the management of a midland shooting estate, and in this reviewer's opinion the book cannot take the place of J. R. Miller's Practical Gamekeeping.'

Ed Zern, from Field and Stream,'' November, 1959

Absorption of Networks

Chapters 3 and 4 of this study discussed the use and absorption of networks in the non-military sectors of the Soviet Union prior to 1990. The broadest conclusion that can be drawn is that there was almost no absorption of computer networks and data communications in the Soviet Union prior to 1990. Figure~\ref{old-concs} provides some general conclusions based on the model from Chapter 5.

Chapters 6, 7, 8, 9, and 10 discussed the changes in network technology manufacturing, integration, development, and absorption from 1990 until early 1993. The situation in the former Soviet republics has changed dramatically. Networks of all sorts are becoming more available, starting with population and economic centers like Moscow, St. Petersburg, Kiev, and the Baltic capitals. Figures~\ref{new-concs-1} and~\ref{new-concs-2} summarize all these conclusions.

\begin{figure}
\ssp \centering \footnotesize

\begin{tabular}{ll}
\hline
{\bf Area} & {\bf Conclusions} \\
\hline
{\bf Infrastructure} & Infrastructure was in poor shape. \\
& There was no significant indigenous manufacture of \\
& network hardware or software.\\
& Telephone lines were scare, expensive, and of poor quality.\\
& National networks for building applications were practically\\
& nonexistent. \\
& \\
{\bf Environment} & Environment did not foster growth of networks. \\
& The command economy did not encourage enterprises to adopt\\
& networks.\\
& Communist security and political mindset made sharing of \\
& information among enterprises and organizations a liability\\
& rather than an asset. \\
& Computer and network hardware of all types in continuous \\
& extreme scarcity.  Allocation system sent already scarce hardware\\
& without regard for real enterprise interest/need.\\
& Little or no hard currency was available for foreign products.\\
& \\
{\bf Support} & Support for networks existed to some extent.\\
& Enterprise-level support was highly unusual.\\
& Support services from hardware vendors did not extend to cover\\
& integration technologies such as networks.\\
& Free software reduced economic impetus for support.\\
& \\
{\bf Applications} & Applications drove only a few networks.\\
& Traditionally networked applications did have networks: \\
& transport, oil and gas industries, military, banking. \\
& In general, very little demand push/pull for networks.\\
& \\
{\bf Technology} & Technology to implement networks was not widely \\
& available.\\
& Personal computers were not available, and were so expensive \\
& that possible productivity gains were not economical.\\
& Specialized hardware such as high-speed modems wasn't available\\
& indigenously, and was held back by export control and hard \\
& currency shortages.\\
& Accessibility of network software and hardware did not allow \\
& boutique'' style adoption; networks were only facilitated \\
& when allocated from above.\\
\hline
\end{tabular}

\raggedright \vspace{.5cm}

\caption{Absorption of Networks Prior to 1990: Conclusions}
\label{old-concs}
\end{figure}

\begin{figure}
\ssp \centering \footnotesize

\begin{tabular}{ll}
\hline
{\bf Area} & {\bf Conclusions} \\
\hline
{\bf Infrastructure} & Infrastructure is being improved.\\
& Manufacturing still quite weak, and much indigenous \\
& manufacturing has been shut down as non-competitive.\\
& Telephone physical plant abysmal, but being improved.\\
& Basic telephone service is being supplemented by cellular \\
& telephone service.\\
& Lead in satellite technology is promising, but has had a small\\
& impact so far, except in international access.\\
& X.25 network access is available from many different vendors.\\
& TCP/IP coverage and absorption is weak, but growing.\\
& Western funded or operated projects continue to make a \\
& significant impact.\\
& Former Party and military facilities are easing the shortage \\
& State-funded network projects are not doing well at all.\\
& UUCP (Relcom) is now the strongest network in the republics.\\
& \\
{\bf Environment} & Environment is weak and chaotic.\\
& Complete regulatory and policy chaos.  \\
& Russian Ministry of Communications policy not being taken \\
& seriously.\\
& Personal use of computer communications is exploding in an \\
& era of uncertainty. \\
& Networks are adding an element of controversy and\\
& excitement.  They fit right in with traditional Soviet \\
& conspiracy theory attitudes.\\
& International environment is improving with changes in export\\
& controls and additional foreign assistance.\\
& Economic collapse is still a very real danger, and colors all\\
& \\
{\bf Support} & High quality support is available in some areas.\\
& Support now available through creation of a Usenet news \\
& hierarchy in Russian/Cyrillic.\\
& Some international cross-pollination is present.\\
& credible service offerings, as are international suppliers such\\
& as Digital, IBM, ComputerLand, {\em etc.} \\
& Some larger institutes and organizations have faded away, but \\
& the biggest-of-the-big are still in business and working away.\\
& Coverage in major cities is not a problem.  Rural areas are \\
& being ignored.\\
\hline
\end{tabular}

\raggedright\vspace{.5cm}

\caption{Absorption of Networks Since 1990: Conclusions (1 of 2)}
\label{new-concs-1}
\end{figure}

\begin{figure}
\ssp \centering \footnotesize

\begin{tabular}{ll}
\hline
{\bf Area} & {\bf Conclusions} \\
\hline
{\bf Applications} & Applications are driving the creation of networks.\\
& Wide area network applications are showing up all over.\\
& Enterprises are realizing the economic benefits of networking.\\
& LAN installations are way up, likely due to economic benefits\\
& rather than for data sharing reasons.\\
& Cyrillic is going to be a long-term problem for use and \\
& acceptance of computers, but Russification is not.\\
& \\
{\bf Technology} & Technology is available to work with infrastructure.\\
& High-speed, error correcting modems are available and are being\\
& used with great success.  Without this technology, networking \\
& might never have come about in the former Soviet republics.\\
& Packet radio is seeing increasing use, especially as a bypass\\
& for international lines, and in rural areas.  It is not yet \\
& a significant contributor to networking. \\
\hline
\end{tabular}

\raggedright\vspace{.5cm}

\caption{Absorption of Networks Since 1990: Conclusions (2 of 2)}
\label{new-concs-2}
\end{figure}


In this section, the questions proposed in this study will be examined in light of this research.

What were the capabilities of the Soviet Union (and what are the capabilities of the former Soviet republics) to develop indigenous network hardware and software products? What were the manufacturing capabilities? What quantities of products have actually been produced?

The Soviet Union had a civilian program aimed at researching and developing network hardware and software. This program was not a priority within the context of the general informatics development programs. While multiple institutions and organizations were engaged in the parallel development and manufacture of computer systems, software, and peripherals, interest in networking was limited to a small number of institutes spread over the entire Eastern Bloc.

The products which were created were not competitive with other international suppliers. In many respects, the Soviets were wasting their time in manufacturing equipment which was far behind the state-of-the-art in Western countries. The indigenous software development was also of little utility. Soviet LANs existed to link the very limited computational capabilities which did exist. The inevitable replacement of PDP-8, PDP-11, IBM 360/370 architecture, and CP/M operating system computers with more smaller, more reliable, more powerful, and less expensive systems made much of this LAN research an exercise in futility.

In the former Soviet republics, much of the obviously wasteful and useless research and manufacturing has been eliminated. This implies that there is virtually no manufacturing of network hardware at all. The few exceptions to this are either joint ventures using foreign technology, or military conversion enterprises.

The research and development programs of the Soviet Union may be continuing. Certainly, many of the most important organizations and institutes from the Soviet Union are still in place and operating. These same organizations are using networks. The practice of wasteful and useless research has almost completely stopped. In some institutes, these programs are being replaced by more current, profitable, and useful research programs, almost always with the direct fiscal support of foreign partners.

Most software development is focused on creation of localized versions of foreign software or on software which will be most useful in the local market. Software development in networks is dominated by products which use Russian and Cyrillic; by products which support networks in an environment where telecommunications resources are of low quality and intermittently available; and by products which are designed to run on low-end MS-DOS systems.

To what extent did the Soviet Union (and do the former Soviet republics) rely on foreign technologies, both proprietary and public, in distribution and manufacture of networking hardware and software?

The Soviet Union had indigenous network software and hardware. While some networking was done with local functional equivalents of Digital's DECnet and IBM's SNA software, there were existing networks which were based entirely on either indigenous network hardware, network software, or both. It is unclear what the mix of indigenous/foreign networks was. Although I have anecdotal evidence of more DECnet and SNA networks than indigenous Soviet and CEMA networks, this evidence is not strong enough to suggest one style as dominant over the other.

In the area of network hardware, though, the situation is more clear cut. Networks in the Soviet Union were based almost exclusively on hardware produced in the CEMA countries. Even in networks which were running SNA software directly (i.e., not a functional equivalent), the CEMA-manufactured front end hardware was seen. Modems in organizational networks were almost entirely of local manufacture; modems on personal computers were more likely to be imported Western technology. The quantities in the latter case were minuscule.

The traditional Soviet mindset of total independence from foreign hardware technology meant that most networks were composed of indigenously designed and manufactured hardware. This same mindset also created a great shortage of network hardware, and helped to retard the dissemination of networks throughout the Soviet Union.

In the former Soviet republics, virtually all network hardware is of foreign manufacture, and networks are using software based on foreign non-proprietary standards. At the same time, indigenous networks have almost completely disappeared. The sales of non-Ethernet and non-Arcnet LAN adapters continue to drop, and there are no public reports of new indigenous network developments.

While many different networks are being created, these are built on a foreign hardware and software base. What contributions to network software are being made are generally add-ons and localizations, such as support for Cyrillic, low-end systems, poor telecommunications facilities, and Russian language addition.

Virtually no indigenous hardware is being manufactured. What is being created lives under the Soviet stigma:'' it is believed unreliable, of poor quality, slow, and below international standards. There is strong direct evidence that buyers in the former Soviet republics prefer hardware which is of foreign manufacture. Local equipment, usually bought for rubles, is being ignored or discounted heavily.

How widespread were computer networks in the Soviet Union? How widespread are they in the former Soviet republics? To what extent have these technologies been absorbed into enterprises and institutions? How has the absorption of these technologies changed over time, especially given recent events?

There was little or no absorption of computer networks into Soviet enterprises. McHenry found many Management Information Systems (MIS) but few instances of effective utilization: dissemination without absorption. In networks, this was true as well, but on an even smaller scale: few networks, and even fewer integrated into enterprises and organizations. State sponsored networks (such as Akademset' and PD-200) were a resounding failure. It was estimated that fewer than 5,000 LAN adapters existed in the Soviet Union prior to 1990. Networking development, dissemination, and absorption was retarded strongly. What networks did exist were almost exclusively intra-organizational.

In the intervening years, networks have become more and more common in enterprises. More importantly, because enterprises are now free of central planning constraints, the existence of a network in an organization or a network connection to an organization signifies that the organization feels that the network is of value to it. Network hardware is still expensive; network services such as Relcom and other WANs are expensive; personal computers are still expensive and scarce. When an enterprise allocates scarce resources to a network, this indicates that the network is of use to the enterprise.

The surge in database applications, bulletin-board exchanges, and electronic mail indicate that networks are being used for inter-organizational information exchange. With thousands of personal computer local area networks installed in organizations, it is clear that networks are being absorbed into enterprises and institutions.

However, networking activity is focused on a small set of cities. The spread of technology, expertise, and support is strongly centrist, emanating from Moscow and other similar centers in the Western quarter of the geographic area of the former Soviet Union.

It is also important to understand the total magnitude of networking in the former Soviet republics. An optimistic estimate of the total number of WAN nodes in all republics would be 20,000, combining Fido, Relcom, and other networks. At the University of Arizona, a school of 35,000 students, there are over 9,400 registered TCP/IP systems. Dissemination in Tucson, Arizona, is three orders of magnitude higher, per capita, than in the former Soviet republics.

What prospects for manufacture, development, dissemination, and absorption of computer network technologies are there in the former Soviet republics? How does this differ from the prospects of other industrialized nations such as West Germany, Japan, France, and the United States?

Dissemination and absorption of computer networks in the former Soviet republics is occurring at a rapid rate. The most significant barriers to networking in enterprises are economic ones. The former Soviet republics must slow the inflation rate and loss in value of the ruble against the dollar. Without economic stability, enterprises will be weak. As long as there is continuing real danger of collapse, the creation of inter-organizational computer networks will be held back.

Telecommunications infrastructure in the former Soviet republics is very weak. However, existing technologies such as error correction modems are able to work around much of this weakness for initial installations. High speed networks, depending on fiber optics and better quality telephone systems, will be supported in larger cities as the infrastructure is improved. The telecommunications infrastructure is also being supplemented by former Party and military technology which has opened up into the civilian/commercial sector. These resources offer breathing room to the telecommunications providers.

Government regulation of telecommunications could be a problem. During this period of transition from the centrally planned economy to a more open, market based economy, the government is in a chaotic and less powerful state. The Ministry of Communications has announced policies, plans, and regulations governing telecommunications in both the news media, at conferences and presentations. Many organizations are ignoring or working around these policies---or are simply ignorant of them. Unfortunately, without strong government support and financing for telecommunications infrastructure development, many rural areas of Russia and the Central Asian republics could remain information backwaters.

The other extreme: Soviet Union-style regulation of communications, could cause a chilling effect. The Ministry of Communications charged very high prices for telecommunications services during the pre-1991 era. During the first part of 1993, federal uncertainty is a way of life in Russia. It is unclear which way the pendulum will swing.

Support structures are growing in areas where there is a need. This reinforces the centrist orientation of networking. However, the Soviet Union's emphasis on science and technology in education is working to its advantage, as technicians and engineers trained on the CEMA clones begin to pick up the slack in rural areas. The Soviet tendency to distribute plants and institutes into small towns across the country also distributed trained and educated people who are moving to the new networks during this transition period.

Cyrillic will continue to be a problem in the absorption of networks, and in Soviet/European and Soviet/American international communications. While there may be strong adoption of Soviet standards for Cyrillic within the Russian-speaking world, it will take many more years for such standards to propagate to the rest of the world. The language barrier will be compounded by an encoding barrier.

Many new projects are being driven by foreign investment and joint ventures. This technology transfer will continue as long as there is money to be made in the Soviet market. Many of these joint ventures transfer not only products, but also manufacturing technology, as they use less-expensive Soviet labor to finish or modify products manufactured elsewhere. Although it may not be a goal of the former Soviet republics to sell labor at a discount, this disparity in labor values will move manufacturing operations into the republics. This move, though, will likely be focused on those republics which outside observers see as stable and ordered; it is likely that most areas of the former Soviet Union are considered too unstable and chaotic for serious business investment. Manufacturing enterprises are being brought on-line which exist only for export purposes. This devalues Soviet labor, but does bring a valuable transfer of technology. It is likely, though, that the Soviets will benefit more than some of the lesser-developed countries have from this arrangement.

For the short to medium term, though, it is likely that indigenous manufacturing and development---not part of any joint venture---will be low to insignificant. The entire computing research and development community has been through such a period of upheaval that it will take years before an industry will grow that operates at a competitive level with the rest of the world.

Most importantly, applications are appearing which will drive the formation of wide area networks. In an era of scarce resources, American and European companies have demonstrated that networks have tangible economic benefits. Many Soviet enterprises are beginning to realize this, and are building networks to give them a competitive edge in their new market-based economy. They are building and using networks to compensate for weaknesses in the business infrastructure as well.

Conclusions and Observations

This study has examined the absorption of computer networks in the former Soviet Union. Three significant conclusions can be drawn from the work presented.

First, the changes in the post-U.S.S.R. economy have done little to improve development and manufacturing capabilities. It is reasonable to state that the wide availability of Western goods, loosening of controls on foreign currency transactions, and a virtual convertibility of the ruble have driven network developers and technology manufacturers out of business. The networks which are being installed are based on foreign technology and use equipment of foreign manufacture. With a complete loss of all explicit and implicit economic protection, local development and manufacturing has almost completely disappeared. Eager investments by foreign companies in telecommunications facilities such as switching systems, fiber optics, cellular telephony, and bypass operations bring significant improvements in the infrastructure in some cities, but at a cost: little manufacturing is done in former Soviet republics; Soviet partners see little technology transfer; and what little profits are made leave the country with the foreign supplier. (Footnote: This trend is slowed, but not stopped, by the difficulty of converting rubles into hard currency.)

Second, the which have occurred are centered on a few cities and republics. There are fifteen former Soviet republics. This study deals almost exclusively with six of them: Russia, Ukraine, Byelorussia, and the three Baltics (Estonia, Latvia, and Lithuania). Virtually no developments have been reported in the other nine republics. Within these six republics, there is still a strong centrist flavor: Russia might well be composed of Moscow, St. Petersburg, and perhaps Novosibirsk, by the information found in this study. Just as all Aeroflot flights had to pass through Moscow, all network communications must do the same. Yet these Russian cities represent less than 10% of the population of the Russian republic. [Stevenson88] 96% of the network nodes serve 5% of the connected cities. The dispersal of network technology away from these centers is not keeping up with the growth in the central sites.

Third, the growth of networks is explosive. Networks in the former Soviet republics have increased in size by two orders of magnitude in less than five years. This is possible thanks to a series of factors which encourage such expansion. The telephone network, once thought to be an insurmountable barrier to computer networks, has become useful through the application of new modem technologies. The general complexity of building networks has been reduced through appropriate software, such as UUCP-- and Fido--compatible packages. The English language and Latin alphabet of Western systems have been overcome both through indigenous development and a general trend to internationalize software. The paradigm of a large computer as a network node hosting many users has changed with the introduction of personal workstations, a trend starting in the West and eagerly embraced by lesser developed countries. The critical shortage of equipment capable of building networks has been alleviated by inexpensive, powerful, and easily available systems based on Intel architectures. And many of the barriers created by the political and economic environment of the Soviet Union have been lifted in the past decade.

Economic Style

After the tremendous changes in the Soviet Union, a Western observer might be tempted to draw the mantle of the Cold Warrior close in making conclusions about the former Soviet republics. It is certainly true that the Soviet Union placed itself as a peer to the industrialized nations of Western Europe, North America, and the Far East. Observation of the Soviet progress in science and technology entirely contradicts that assertion. The Soviet Union's industrial output in the area of computer networks and data communications more closely resembled that of Latin than North America.

The reasons for the lack of Soviet progress are outside of the scope of this study and the expertise of this author. But the researchers cited in Chapter 2 drew conclusions which point directly at the command economy of the Soviet Union. Whatever additional factors may have been involved, central responsibility for the failure of the Soviet Union to innovate, diffuse, and absorb computer networks is the socialist economic discipline. In this respect, the Soviet Union was not a first-world country, but an LDC (lesser developed country).

Can the change from a command to a market economy account for the growth in absorption of computer network technologies in the former Soviet republics? Again, rigorous support for such a statement is outside of the scope of this study, but there is clear evidence that the economic change is at least partially responsible for the changes. The elements which were absent from the Soviet command economy---incentive for innovation and a strong profit motive---are significant forces in the development of the Relcom network, many of the database services, and many of the changes in telecommunications infrastructure. Other changes, such as the pervasive police state atmosphere, have also made contributions, most visible in the expansion of the Fido network and other Soviet bulletin board systems as well as in the willingness of Soviet network participants to engage in free and unrestricted discourse with their peers in the West.

The Difference between One and None

When the first Internet connection comes to an LDC, this fact is usually trumpeted throughout the network community as a major change. Yet these connections are usually low-bandwidth, initially unreliable, and available only to a small set of people in a country. Is there really any significance to the first connection in Turkmenistan?

A connection can bring a wealth of information to a small community of people. But such small populations are insignificant in the larger context of a developing nation-state. It might be argued that the right information in the hands of the right people can substantially change the course of a country's future. But will the right information'' come from the Internet? Are the political and economic pressures within a developing nation too complex to notice a contribution drawn from the worldwide network community? Or does Doyle's axiom that the little things are infinitely the most important'' apply? (Footnote: From Case of Identify'' in The Adventures of Sherlock Holmes, 1892.)

Leaving aside these unresolvable Machiavellian notes, one can turn to a purely economic argument. A single connection represents a substantial cost and a substantial capability. Even if the cost is not fiscal, it represents a level of knowledge and expertise, of commitment to a connection, and the solution of a basic set of problems. Once a single connection is in place, the marginal cost for a second connection is very small.

For these reasons, a single connection is significant. Even if the significance is not immediate or measurable, a single connection is the seed that can grow into a larger network. A single connection also provides a demonstration for skeptics. While the benefits of networking are easy to consider in the abstract, a connection which has real access to real people and information is a much better sales tool than any number of lectures or papers.

In the context of the former Soviet republics, it is clear that the single network connections which have grown into small network communities are important. If the former Soviet republics hope to build a nation-wide network which mimics the scope and coverage of the Internet in the United States, these small seeds of networking are an important part of the process.

Foreign Comparison

All of these changes have released a flood of networking technology on the Soviet business and academic communities, which brings up a very important final question: what does it matter? In less colloquial terms, what is the real impact of these technologies on the people who live in the former Soviet republics? Have their lives been enriched? Has this made business opportunities where none existed? Has all of this networking made a difference? Is this technology a net gain, or a net loss, for the economy as a whole?

To draw conclusions about the former Soviet republics' use of inter-organizational networks, it is helpful to seek parallels between use of networks in the United States/Western Europe and the former Soviet republics.

In the United States, it can be conservatively estimated that between five and ten million people have access to some form of inter-organizational computer network. [Snyder910610; Snyder9302] This electronic connection serves many uses, along a continuum of possibilities.

At one end are person-to-person communications. The networks serve to bind together people over long distances with simple and inexpensive communications. Through electronic mail, very quick communication is possible, almost completely ignoring any geographic limitations. Electronic mail is (typically) as quick from Tucson to Tulsa as it is from Tucson to Tokyo. For many information workers, particularly those in academic environments or at high-tech corporations, electronic mail is more convenient, less expensive, and more reliable than the traditional postal service or facsimile transmissions.

Person-to-person are not limited to individuals who have prior knowledge of each other. One popular way in which these electronic networks are used is as electronic merchandise exchanges. These may be one-on-one exchanges, similar to an electronic swap meet, or they may be one-to-many exchanges, where a merchant advertises and sells his wares electronically.

Electronic networks also bring people in contact with electronic resources, such as databases and information depositories. These range from for-pay database services, such as those provided by CompuServe in their CompuServe Information Service, to without-fee depository services, such as the anonymous ftp areas maintained by many academic and commercial organizations.

At the other end of the spectrum are electronic groups, often called communities of interest.'' These communities are supported through a variety of mechanisms, such as the USENET news service, BITNET's LISTSERVs, and smaller bulletin boards such as Sovset'. These mechanisms bring together small and large groups who have a similar interest in some topic. These communities range from purely recreational interests to very serious electronic research groups. In many ways, interest groups make research more accessible to outsiders by making researchers themselves more accessible. Many of these electronic communities are public, and an inquiring aspirant can lurk,'' reading the contributions in a more informal setting than a journal article or conference presentation.

These communities of interest also form with a smaller critical mass than more traditional professional associations. In the field of computing, for example, the difficulty of creating a new Special Interest Group within the ACM or IEEE makes such an enterprise only appropriate for the most dedicated proponents in the most promising interest areas. However, a mailing list, bulletin board, or news group can be created with little or no investment of capital or time, building an electronic water cooler around which a new topic or area of interest can be discussed.

The advantages to the participants of their inclusion in computer networks are great, and easily perceived. Using a basic economic argument, one can simply state that one million people would not pay CompuServe \\$9.00 each month for access to CompuServe's products if they did not believe that they were getting a valuable service for their money. Universities and businesses would not pay thousands of dollars a month for access to the Internet, if it was not obvious to the organization that the access provided benefits.

To enumerate these benefits is useful for later analysis:

1. Person-to-person electronic mail;
2. Semi-anonymous communications for transactions such as buying and selling of merchandise;
4. Creation of communities of interest.
Thus, we can conclude that, in general, users of computer communications networks in the United States perceive that the use is to their benefit, and where there are costs, perceive that the benefit is greater than the total cost.

Parallels and Predictions

What parallels can be drawn between use of computer networks in the United States/Western Europe and in the former Soviet republics?

It is clear from the evidence presented in this study that at least the first three, and possibly all four, of the benefits of communications networks are being enjoyed by users in the former Soviet republics:

\begin{tabular}{ll}
\hline
{\bf EMAIL} & heavy use in evidence\\
{\bf Buy/Sell} & moderate to heavy use in evidence\\
{\bf Database} & some services available, low use in evidence\\
{\bf Communities} & very small amount of use in evidence\\
\hline
\end{tabular}
`
A major transition of the Internet occurred sometime during the mid-1980s, when network costs dropped dramatically, and network access rose explosively. At that time, the Internet began its transition from a small, tightly knit community of computer professionals at well-funded organizations and institutions to a looser network-of-networks. An important change at that time was in the characteristics of the user population. The Internet was no longer the sole province of a technological elite within certain departments of an organization, but a general purpose communications network available to a broad spectrum of communities. Significantly, this new user community had no prerequisite experience in using computer systems as programmers or managers. For many, the only reason to use the computer was network access.

The scope of usage in the former Soviet republics is dramatically smaller than in the United States. If there are five to ten million users in the United States, it is likely that there are ten to twenty thousand users in the former Soviet republics---a difference of three orders of magnitude. The mix of academic and commercial users in the former Soviet republics is also dramatically different than in the United States. The Soviet networks are built almost entirely by commercial entities for business subscribers.

These two differences have contributed to the creation of a computer nomenklatura(Footnote:Nomenklatura is the Russian word used to describe the Communist Party members who held offices at certain levels which included as perquisites access to goods and services far beyond the normal Soviet citizen.) in the former Soviet republics, whether intentional or not. Information is valuable to those who have it, and access to information conveys a commercial advantage to the companies who have access to communications networks. Networks convey an even stronger advantage in the former Soviet republics than in the United States (for example), because they may form the only mechanism for communication between certain commercial communities of interest. Alternative communications mechanisms common in the United States, such newsletters and industry publications, are not typically available in the former Soviet republics.

Networks in the former Soviet republics have helped to create a new division, where a small number of people have access to resources which are denied to the populace as a whole. In this case, the dividing element isn't the Party, but technology. Such a division usually has a strong self interest in restricting membership---an elite with too many members dilutes power and isn't much of an elite. However, there are no signs that this is happening in the former Soviet republics. The discriminating factor is purely economic. If you can afford it, you can join the network.

It can be posited that network access and resources have been valuable to the small number of people and organizations connected. Again, the economic argument is compelling: one would not connect to the network and commit scarce resources if there were no perception of value received.

However, the population of network-connected Soviets is still far too small to see the strong effects which are observed in the United States. Indeed, it can be reasonably stated that the trend is not in the same direction. Because the network resources are, for the most part, restricted to businesses, rather than private individuals or academic/research organizations, much of the network activity has a distinctly commercial flavor. Nevertheless, it is far too early to make a prediction in this area.

One problem which may prove to be insurmountable is the language barrier. Soviets have long been welcome to join the existing Internet communities of interest. Nevertheless, few have taken advantage of the opportunity, except passively. The difficulty of participating in fora in a foreign language is substantial, so much so that it may be reliably predicted that Soviet interaction will be minimal at best.(Footnote: The same is true for Western European nations, even ones where English is encouraged as a second or third language---participation exists, but almost never at the same level as native English speakers.) Although the Internet now connects New York to Novosibirsk, the momentum of an English-centric network is unstoppable.

This implies that the best possibilities for use of networks are internal to the former Soviet republics, rather than externally with foreign countries. This is especially unfortunate. Networks have been promoted as tools for international cooperation and understanding, but the trend is for a small number of highly educated and dedicated individuals to use them, rather than as a true vox populi.

Although networks have made significant impact on individuals, there is no evidence that their impact on the economy as a whole, on the Soviet standard of living, and on international relations, is other than negligible.

Last updated: December 3, 1994.

jms@opus1.com