Last update at http://inet.nttam.com : Wed May 10 18:03:47 1995 Measuring and Comparing the Return on Investment on Network- Mediated Empowerment: A Developing Nations Perspective Stephen R. Ruth RUTH@GMU.EDU Abstract Especially in developing nations, defining the benefits from unit investments in telecommunications is a challenge at national, regional and organizational levels. This paper addresses the problem in terms of measurable criteria that can be used to examine ex ante the feasibility of investments in networking services. The methodology involves development of realistic dependent and independent variables based on varying assumptions about the depth and quality of services being considered. Specific data from implementations now going on developing nations in four continents is included. Introduction This paper offers an overview of several approaches to discovering variables that make it easier to predict the success of networking in developing nations. Since the purpose is not to offer a complete methodology, but an array of possibilities, the emphasis will necessarily be on breadth rather than depth. Increased availability of electronic networking skills and services can have an important impact on the mechanisms by which scientific interchange is effected. Networking technology speeds communications and can support frequent interaction, but it does not generate useful services spontaneously. The growth rate of electronic networking participation will depend largely on the behavior of individuals who understand the professional opportunities that the technology offers them. It is useful to recall that electronic networking first emerged to meet the needs of a computer-literate scientific community. (9,11) As networking evolved, it has gradually been extended to new user communities working in disciplines and regions that were initially less familiar with computer technology. In research centers in most developing nations the current methods of communication are predominantly telephone and fax, plus mail and parcel service. These methods have a high unit cost and preclude the kind of daily scientific interchange that gives the sense of collaboration so vital to scientific progress. As an example of the costs involved, a one page fax from Nairobi to the US costs about $6. By using e-mail that cost would be about ten cents. The advantages of extensive use of international networks go far beyond drastic reductions in unit cost. While connecting scientists to each other using Electronic Mail is an obvious gain, there are many benefits that are far more important. For example, it is now possible to be connected with most of the great libraries of the world using networks; major global data bases can be tapped; hundreds of thousands of new manuscripts and research findings are available to be shared through file transfer protocols that routinely download a new scientific article at the University of Minnesota, say, to a research site in the Philippines; large groups of users with similar interests can be kept informed of new developments, opinions, seminars, etc., through bulletin boards and other services that are beamed to over one hundred fifty countries in the world. But this potential gain in productive capacity is not achieved routinely, by simply buying equipment. The major challenge is in adapting the existing conditions in a particular region to the solutions that are possible. Some of the key variables associated with successful and unsuccessful use of electronic networks are psychological and managerial as much as technical. Sometimes perceived status, social context cues, and other behavioral issues can have a major effect on networking success. (2,18) A different implementation scheme is necessary for each site and situation. In Sub-Saharan Africa, and many Asian countries, where research institutions are severely hampered by political and infrastructure problems, in addition to funding and capacity building challenges. (3,12,21) Public financial resources have not been widely available for promoting the emergence of local electronic networking communities (South Africa is a notable exception). Where local efforts do exist, they rely largely on volunteer efforts, in many cases led by a loose consortium of motivated individuals and groups who are trying to succeed despite the harsh conditions. (20) Dependent Variables--Simple Indicators of Success While there are many complex indicators of successful network use (6,7) the most valuable indicators would seem to be those that are related directly with how much the network is used and how many people use it. There are three statistics that are easy to gather at most levels of the organization: number of active users, number of messages in and out and the gross volume of message traffic for a unit of time, say, a month. In 1992, using BITNET statistics for Hungary, the Czech and Slovac Federated Republic and Poland, gross monthly statistics of these three variables were successful proxies for networking success, showing at the time that CSFR was far ahead of the other two nations. (16) Since Internet volumes are more complicated to measure the national statistics are more difficult to obtain now than when the BITNET data were collected at monthly intervals by the European Academic Research Network (EARN). Another important indicator of success is the trends that are manifested in these three variables. If a research group, or a nation, or a university, shows significant increases in the number of users and messages over a year, these statistics are certainly an important indication of successful implementation of networking. Quarterman and Mitchell's statistics published in Matrix Map Quarterly (10) as well as those published by the Internet Society (8,17) give these results regularly. Some of the trends can be deceiving, because a country that barely shows volume one month can indicate prodigious percentage improvements even with a small amount of gain. But these problems of evaluating data are part of any broad-based analysis and are easily explained and rationalized. Independent Variables--Trying to Predict Success There are a myriad of potential independent variables-- characteristics that can be causes of success in networking in international networking. Literally hundreds of variables have been hypothesized, from the availability of powerful network backbones (4) to the dominant religious affiliation of the country involved (5), to the specialties of individual users (2). Table 1 (2) is an example of some of the possible predictor variables for five countries from four continents using demographic and other variables along with predictions for network success based on results in mid 1993. Such summaries are not capable of developing a predictive theory--they simply allow a small window on the challenge of predicting which variables are candidates for contributing to network success. Table 1 Selected Variables for Examining Network Use Characteristics of Five Developing Countries in 1993 Pakistan Tunisia CSFR Chile Romania Gross Domestic Product ($Bil) 40.2 10.2 123 25 78.8 Population (Millions) 113 8.2 15.6 13.2 23.2 Area (Thous. Square Miles) 310 63 49 292 91.5 Population Density (Persons/ Square Mile) 335 125 317 44 252 GDP Per Capita ($ per person) 370 1253 7870 1979 3445 UNDP Development Ranking 132 93 26 36 77 Predominant Cultur- al/Religious Tradition Mos- Mos- Christ- Christ- Orth- lem lem ian ian odox Year of First Major Internation- al Connectivity 1991 1990 1990 1987 1993 Pace of Early Connectivity Slow Slow Moderate Slow Rapid Current users (thousands) .2 .5 5 5 1.5 Current network Activity (Giga- <1 <1 Mid- Mid- Ten Bytes per Month) tens tens Current Network Pace of Use N/A Slow Rapid Moderate Very Rapid Predicted Pace of Use Very Rapid Rapid Rapid Rapid Rapid Source: Ruth, S. The networking revolution: a cui bono perspective, Higher Education Policy, 6 (4) 1994, p. 41 As another example, there is a close relationship between the United Nations' Human Development Index (19) and sophistication of network capabilities. Table 2 shows the obvious connection : that becoming a powerful nation is accompanied by elaboration of many technologies--including networking. Table 2 : Quality of National Connectivity Compared with National Development based on Internet Statistics and UNDP Development Index (1993 data) Quality of Connectivity Average Development Index Outstanding 28.3 Fair to Good 63.6 Poor to Non Existent 110.4 Sources: L. Landweber, International Connectivity (Version 7) Internet Society News, 2 (1), Spring 1993, p. 40, and UNDP Human Development Index 1993, pp 135-136. Note: The UN's figures are listed in order of development; that is, a low development index means high development. The Real World: Experience-Based Independent Variables The author has spent several years in working with over twenty Internet connectivity projects in developing nations. In the process of these implementations it has become apparent that some variables continue to appear and reappear as possible determinants of success or failure. One obvious choice for predicting success is the availability of hardware and software technology that facilitates connectivity. Yet having all the machines and the programs may not be very useful if another major technology determinant, data communications capacity, is absent. country, or a research center with good equipment but very poor phone lines, low bandwidth in existing telecommunications capacity, will have low probability of successful networking results. There are many examples of very low cost FIDO net stations in a country having much more success than more complex networks. (13) Another recurring experience is the impact of national and regional priorities. In a country like Costa Rica where national priorities in the early 1990's were directly linked to international networking connectivity for universities and a role for the nation as regional center for networking support, it was easy to predict high volumes of users, traffic and network services. This result did in fact occur. Similarly, in other nations of Central America, as in many African countries, the national priorities often could not include networking when more pressing and urgent matters demanded attention. So El Salvador, Nicaragua, Cameroon, Malawi and Burkina Faso, among many others, have been unable to place networking at a high level of national budget planning. A third variable that, in the experience of the writer, often can lead to success or failure, is the infrastructure of expertise and training that inheres in the organization. The success of the Zambia network appears as much due to a progressive accumulation of expertise as to any rich endowment of machines or particularly enlightened government policy. (20) In contrast, it appears that major investments in training do not always lead to broad diffusion, as in the case of Saudi Arabia. Associated with this variable is the degree to which networking is a grass roots activity, a self starting process. In Chile and Lebanon this has been the case, where a very small number of individuals associated with universities led the way, sometimes under governmental or PTT constraints. The author has personal experience in both these cases, observing Chile become the first BITNET node in South America in 1987 and assisting the American University in Beirut in 1994 to develop a student network. In both cases there was an government environment that was complicated and constraining, indicating that lack of strong government support (variable one above) was unable to stop networking diffusion. These three experience-based variables could be augmented by dozens more. They could be integrated with others and they could be elaborated with far more precision. But they represent a starting point for establishing predictors based on actual conditions in developing nations. An Example-Six "Pacesetter" Organizations A realistic example of this approach is now discussed. The author is responsible for arranging to establish improved connectivity for about twenty research centers in developing countries to partner institutions in the United States. Examples of these centers are: Reproductive Biology Institute Nacional del la Nutricion Salvador Zubiran, Mexico City, Mexico National Institute of Immunology, New Delhi, India Mahidol University, Thailand Institute of Primate Research (IPR), Nairobi, Kenya ICMR, Santiago, Chile Demography ICDDR-B, Bangladesh University of Cape Coast, Ghana IFORD, Cameroun Demographic Institute, Faculty of Economics, University of Hanoi, Vietnam Institute of Sociology, Hanoi, Vietnam Pakistan Institute of Development Economics Population Research Institute, Yunnan University, Kunming, China University of Zimbabwe, Harare, Zimbabwe Obafemi Awolowo University, Ile-Ife, Nigeria University of Ibadan, Nigeria Institute for Sociology, National Center for the Social Sciences, Indonesia, Jakarta, Indonesia Office of Population Studies, University of San Carlos, Cebu City, Philippines University of Blind, Algeria University of Zambia First, a series of pilot projects in Africa, South America and Asia was selected with centers that were representative of biology and population research. The countries are Mali, Kenya, Tanzania, Uganda, Chile and Thailand. These six centers, or "pacesetters", vary widely in the degree of infrastructure as well as other variables. Those selected were: - University of Dar Es Salaam, Demography Unit and Tanzanian Commission for Science and Technology (COSTECH), Dar Es Salaam, Tanzania. - Institute for Statistics and Applied Economics and Computer Services Center at Makerere University, Uganda. - CERPOD in Bamako, Mali, the population studies center of the Institut du Sahel (Sahel Institute or INSAH); a regional institution serving the nine Sahelian countries of the Permanent Interstates Committee for Drought Control (CILSS). - Institute for Population and Social Research, Mahidol University, Thailand - Institute of Primate Research at the National Museums of Kenya in Nairobi - Instituto Chileno de Medicina Reproductiva (ICMER) in Santiago, Chile CERPOD, The Centre for Applied Research on Population and Development, is a particularly important site for communications since it has linkages through the Sahel region. With headquarters in Bamako, Mali, CERPOD serves as a focal point for studies in drought control, population and other research and has active relationships with these countries: Burkina Faso, Cape Verde, Chad, Gambia, Guinea Bissau, Mali, Mauritania, Niger and Senegal. ICMER in Chile and IPSR in Thailand have differing needs and entering conditions. While both are affiliated with institutions that already have some Internet connectivity (Catholic University of Santiago and Mahidol University in Bangkok) ICMER has better training infrastructure. IPSR, with less training infrastructure apparently has fairly good hardware for data communications. Each location requires different approaches to achieve sustained connectivity. COSTECH, located at dar es Salaam, Tanzania, is primarily a population research center with linkages to Johns Hopkins and University of Pennsylvania population centers. Makerere University in Kampala, Uganda, has a center that supports demographic studies. Table 3 gives a summary of the pacesetters. The six institutions needed to be connected to sufficient capacity to offer basic network services : E-mail, file transfer, lists, bulletin boards and data bases. Table 3 Summary of the Six Pacesetter Institutions and Estimates of Needs for Achieving Sustainable Connectivity Institution CERPOD ICMER IPSR IPR COSTECH ISAE Location Mali Chile Thailand Kenya Tanzania Uganda Specialty Pop. Biol Pop. Biol. Pop. Pop. U.S. JH UNC/ UW/UNC UNC/ PSU/JH JH/UP Partner(s) JH/UP PSU/ Baylor UP/Brown Brown Potential Internet Traffic (MB/Month) 500 100 100 100 200 300 Availability of Local Support Fair Fair Poor Fair Poor Good Institutional Training Exten- Exten- Exten- Exten- Exten- Exten- Required sive sive sive sive sive sive Hardware and Software Exten- Exten- Exten- Needed Some Some Some sive sive sive Followup Exten- Exten- Exten- Exten- Exten- Exten- Required sive sive sive sive sive sive Estimated High Low Moder- High Moder- Moder- Total Cost ate ate ate * JH-Johns Hopkins University; PSU-Pennsylvania State University; UNC-University of North Carolina; UCD-University of California, Davis; UP-University of Pennsylvania; UW- University of Washington. Goals of Connectivity for Pacesetter Institutions Goals for these pacesetter institutions needed to be specific so as to assist local authorities (directors, managers, etc.) in determining the degree of success. In an opportunity loss context it was estimated that without specific efforts at improving connectivity the following outcomes would occur: Estimated Outcomes if Connectivity Efforts were Not Undertaken Relatively little connectivity, even in sites that have some capacity; only a few users, primarily those with some technical background; delay of up to three years in being able to use network services routinely; message volumes in the range of several per month; use of only e-mail instead of the broader capacity of network services. With successful connectivity efforts it was possible to attain the following outcomes: Estimated Outcomes if Connectivity Efforts Are Undertaken: Good connectivity capacity at all pacesetter sites; daily use by a wide range of researchers and students; no significant delay in leveraging the technology; message volumes of several hundred per month and traffic of 100-200 Megabytes per month; use of file transfer and data base search capabilities, not only E-mail. Selecting Independent Variables To predict the success of a process it is helpful to aggregate the factors that seem to have the ability to influence the outcome. In the case of the six pacesetters at least four general categories could be considered. Each of these will be discussed briefly below. It should be noted that there are many ways to define these categories and that the ones selected are somewhat arbitrary. This approach is recommended as a beginning, not as a verified, final analysis, of a very complicated task. First Proposed Predictor Variable--Governmental Milieu This variable has to do with the degree to which government policy in the country is helpful to the specific implementation. It includes the presence or absence of regulations affecting transborder data flow, the role of the PTT in pricing network services, the attitude of the government toward diffusion of networking services and the role of the government in facilitating or impeding the regular and continuing increase in networking services. A related factor in this variable is the degree to which a government is willing to allow the private sector to become a major force in the diffusion of network services. An "outstanding" rating on this variable would be a governmental milieu characterized by reasonable network tariffs, relatively stable networking diffusion policies, relative openness to allowing network providers to operate profitably and a demonstrated willingness to allow larger numbers of persons to be systems users. Second Proposed Predictor Variable--Infrastructure and Organizational Expertise This variable is related to the actual venue for the implementation, the six pacesetter centers. For a center to have a higher probability of a successful implementation it would be necessary to have a skilled cadre of persons on site who have the expertise, and the time, to facilitate connectivity. This cadre would need training from some legitimately certified entity, as the Internet Society's annual technical workshops. Equally important in this variable is the attitude of the managers and users of the technology toward its diffusion. In some organizations the diffusion process is limited to only one office, or in the context of a university, to a small number of academic disciplines. A rating of outstanding on this variable would be an institution with a strong cadre of persons with excellent networking skills and some who are responsible solely for network management. This institution would also be characterized by relatively broad diffusion of networking activity across specialties and offices. Third Proposed Predictor Variable--Technology Return on Investment This predictor variable does not measure the amount of hardware and software technology but the relative leverage that is gained from the technology. For example, while one organization may use FIDONET for networking, and another UUCP service or even full Internet connectivity, the FIDONET user may be far more successful, given the limited resources used and the results attained. Similarly, a relatively limited time-constrained bandwidth or even telephone line can be employed in a way that has a drastically lower unit cost than a more sophisticated, but poorly leveraged, networking system. An outstanding on this category would be characterized by an organization that is characterized by making skillful, cost-efficient use of existing technology at a unit cost that reflects high return on investment. Fourth Proposed Independent Variable: Outside Investment A basic force in technology transfer in developing countries is the idea that specific investments in training, machinery, infrastructure development and the like have a direct influence on some favorable outcome, like raising living standards. In the cases of the six pacesetters selected investments were made to improve the chances of connectivity. These investments were of at least four different types. The most typical investment was in computer equipment, usually a powerful 486 PC and modems for Fidonet operations. These investments are coded P in Table 3. Investments were also made in training (T), interim funding of line charges (L), facilitation of day to day maintenance and follow-on training (F). A major problem introduced by this variable is the vast differences in the cost of hardware and software. In Chile, for example, the line cost for ICPR is in the range of $50 per month for 5 megabytes, a reasonable fee. In Nigeria the cost is much higher. (1) In some cases, like CERPOD, it may be possible to share services with organizations like AID. Costs for PC's and modems are not as different as line costs . Proposed Independent Variable--Current Trends and Results This variable is an overall assessment of progress to date in three easily measurable dimensions: Trends and results in messages in and out. (Actual messages per month in and out and the percentage increase or decrease from the previous months) Trends and results in megabytes in and out. (Actual megabytes per month in and out and the percentage increase or decrease from the previous months) Trends and results in system users (Actual users each month and the percentage increase or decrease from the previous months) An outstanding connotes significant numbers in all dimensions and a trend that is stable and positive. Results for the Six Pacesetter Centers Table 4 summarizes the results with the three independent variables scored as Satisfactory, Excellent and Outstanding. (Rationale for this scoring is described above.) As of Mid-April 1995, there has been considerable progress at the six pacesetter institutions. ICMER in Chile has become the most advanced user of the technology, due in most part to Dr. Patricio Morales, a cell biologist from the Universidad Catolica, Santiago, who has been teaching his colleagues not only basic network services but full scale World Wide Web capability, when available bandwidth permits. Perhaps the most successful of the six is CERPOD, in Bamako, Mali, where the director of the institute, Dr.Dieudonne Ouderaogo, has made a specific management decision to provide the infrastructure, long term funding and results-oriented management leading to high volumes of message traffic and broad diffusion of Internet technology in the institute and among their partner institutes in the Sahel region of Africa. IPSR in Thailand is on the verge of making major improvements in its Internet capability, including user ID's for all researchers and ten-fold increases in traffic volumes. IPSR is also the only one of the pacesetters that has its own gopher. (22) This is due almost entirely to the interest shown by Dr. Apichat Chamratrithirong, IPSR's director and Dr. Charles Hirshman, of the University of Washington, a research colleague of Dr. Apichat's. Of the remaining three pacesetters, COSTECH in Tanzania has already increased its traffic volumes several fold, primarily due to the facilitation of Mr. William Swingili, a technology expert there. Similarly, in Uganda, due to the daily assistance of Mr. Charles Musisi, it has been possible to see dramatic increases in the volumes of incoming and outgoing traffic. Finally, at the Institute of Primate Research in Nairobi, Kenya, there has been a lesser degree of success. There is sufficient training and computer capacity there but some management decisions have not led so far to the growth potential that is possible. Table 4 Summary of Results for the Six Pacesetter Institutions as of May 1, 1995 CERPOD ICMER IPSR IPR COSTECH ISAE Mali Chile Thailand Kenya Tanzania Uganda Predictor (Independent) Variables Government S S S S S O Milieu Infrastructure/ O S S U S E Organization Technology E S S U E E Return on Investment Investment P,L,T F,L P,F,T P,F,T P,T P,T Outcome (Dependent) Variable Specific Trends and Results O E S U E E Note: Evaluations for first three predictor variables are Outstanding (O), Excellent (E), Satisfactory (S) and Unsatisfactory (U). For the fourth variable, Investment meanings are Personal Computers and other equipment (P), Line Costs (L), Training (T) and Facilitation by outside experts (F). Insights and Conclusions This paper has the advantage of concentrating on specific, ongoing examples of the realities of predicting networking success in developing nations. Its disadvantage is the necessary aggregation required to arrive at criteria. But in spite of this several ideas emerge clearly from the pacesetters' performance. First, the successful pacesetters, like CERPOD, Makerere University and ICMER, have two characteristics in common: a small group of enthusiastic users on site and specific management support in the form of providing administrative support for infrastructure development. Less successful implementations lacked this. IPCR at Mahidol University in Thailand is on a trajectory to a similar success story since their director has taken the lead in pressing for greater network utilization. Another insight is the relatively small role played by extensive investments in machinery if there is little user interest and expertise and only lukewarm support. All the sites had received about the same financial assistance for networking but this investment did not prove to be a predictor of success. The author's experience in nearly two dozen Internet implementations at research sites and universities in developing nations all over the world fully conforms to this model. That is, if there is management support, a trained cadre of enthusiastic users and relatively low levels of government intervention long term success can be achieved at unit costs that are very modest. This experience does not lead to a recommendation against large investments in better lines, more powerful equipment and long term use of all networking tools, especially the World Wide Web. Quite the opposite. If early networking efforts like those described in this paper meet with success as it has been defined here-a continuing trend toward more qualified users, more messages in and out each day and broader use of available tools--then the foundation is in place for much more generous and comprehensive investment that will yield continuing better results in government, education, business and health. References 1. Agbodike, C. Networking in Nigeria Internet Society News, 2 (3) 2. Dubrovsky, V., Kiesler, S. and Setha, B. The equalization phenomenon: status effects in computer mediated and face-to-face decision making groups. Human Computer Interaction, 6 (1991), 119-146. 3. French, H. Linked to the Internet, could Africa's voice be heard? New York Times, October 1, 1994, section A, p 2. 4. Goodman, S., Press, Ruth, Rutkowski. The global diffusion of the Internet: patterns and problems. Communication ACM, 37,8 (Aug. 1994), 27-31. 5. Goodman, S.E. and Green, J.D. Computing in the Middle East. Communication ACM 35,8 (Aug. 1992), 21-25. December 1994, 110-116. 6. Grantham, C.E. and Vaske, J.J. Predicting the usage of an advanced communication technology. Behavior and Information Technology, 4 (1989), 327-335. 7. Gurbaxani, V. Diffusion in computing networks: the Case of BITNET. Communications ACM, 33,12 (1990), 65-75. 8. Landweber, L. (1995). Connectivity table. The Internet Society, Reston, VA. 9. Lederberg, J. (1978). Digital communications in the conduct of science: the new literacy. Proceedings of the IEEE, 66(11), 1314-1319. 10. Matrix Map Quarterly, Matrix, Inc., Austin Texas 11. Newell, A. and Sproull, R. F. (1982) "Computer networks: prospects for scientists. Science, 215, 843-852. 12. On Line communities:Asians taste free speech on the Internet. Wall Street Journal, December 8, 1994, B1 13. Richard, J. and Funk, G. The International Fidonet- 22.067 bulletin boards with an attitude. Boardwatch (Aug. 1993), 66-68. 14. Ruth, S.R., "Using low cost, high yield academic networks in developing nations: a value added services perspective" International Information Systems Volume 1, No. 2, pp 87-99 1992 15. Ruth, S. The networking revolution: a cui bono perspective, Higher Education Policy, 6 (4) 1994 16. Ruth, S. and Goet, R. (1993), Must Invisible Colleges be Invisible? An Approach to Examining Large Communities of Users. Internet Research, 3 (1), 36-50 17. Rutkowski, A. Internet metrics, Internet Society News, 2 (1) pp 37-38 1993 18. Sproull, L. and Kiesler, S. "Reducing social context cues: electronic mail in organizational communications" Management Science 32, 1492-1512 1986 19. United Nations Development Program, Human Development Index, 1994, page 135 20. White, W. Speaking out on African networking. Internet News 2 (3), Autumn 1993, 10. 21. Williamson, W. Speaking out on African networking. Telephony, April 1, 1991, 4 22. Mahidol gopher can be reached at gopher.mahidol.ac.th Vital Services o Active assistance in details of establishing connectivity including some hardware and software augmentation, where needed o Assisting users with online technical support for modems and communications software and helping them to identify and properly use internet mail addresses. o Professional guidance in choosing and installing electronic networking technologies that can be readily maintained and supported in a particular developing country environment and are suited to a specific user's needs and budget. o White pages services (seeking individuals) o Yellow services (seeking organizations and special services) o Direct "hand-holding" links with users and prospective users, including planning and delivering on-site trainings and demonstrations as needed. o Operating a model electronic information system offering ready access to population-related resources and freely distributable software. o Researching and field testing new developments in electronic networking software and hardware to keep pace with technological advancements. Useful Services o Encouraging electronic connectivity and information resource exchange between Mellon-supported researchers and their partners in the region. o Maintaining a supportive liaison between the population research and other Mellon-supported research communities and efforts to build local and regional electronic networking capacity and skills. o Preparation and distribution of a population researcher's guide to e-mail addressing (listing key contacts by institution), electronic conferences, file libraries, and databases searchable via e-mail queries. o Providing assistance to groups interested in developing in-house electronic networking capabilities; especially in cases involving multi- country research projects coordinated by developing country research institutions (e.g., CERPOD migration studies). o Training in techniques for broadening the reach of electronic networking through integration with other information technologies, including desktop publishing, visual displays, and audio. o Distributing information on the status of electronic networking initiatives in the region and providing referrals to locally available skills and services. o Promoting cooperative electronic networking resource-sharing initiatives and small business startups in electronic networking services; o Support for timely electronic dissemination of newsletters, papers, workshop announcements, job notices, grants and scholarships, and other documents. o Publicizing Mellon-supported research in progress through contributions to conferences and online newsletters. o Reporting on successful uses of electronic networking tools by the developing countries population research community. o Strengthening complementarily and integration of electronic networking and related information technology initiatives funded by World Health Organization, World Bank, and others. o Planning and organizing regional seminars associated with networking activities of other organizations, like the Internet Society, RARE, Eureka, etc. REFERENCES