Paul M. A. Baker,1 Senior Research Fellow, The Institute of Public Policy, George Mason University, Virginia
This paper presents preliminary results of a short-term (12 month) community based information infrastructure development project designed to demonstrate the role, function and benefits of an information infrastructure in community and commercial life--The Regional Information Infrastructure Policy Project (RIIPP). Initially organized around community needs for education, training and public information, this joint project focused on community and educational access in Fauquier County, Virginia, a rural ex-urban county of Washington D.C.; and more narrowly on educational Internet access in the suburban Alexandria, Virginia. The project was partially funded through the Telecommunications Information Infrastructure Assistance Program (TIIAP), of the National Technical Information Agency (NTIA), U.S. Department of Commerce.2
The demonstration project (in progress) links a variety of actors in the Washington D.C. metropolitan area, as a means of demonstrating the role of inter-organizational partnerships in promoting information infrastructure development. The primary participants include public schools systems in two geographically separated jurisdictions, an academic institution (George Mason University), acting as system integrator and policy consult ant; and the federal government.
Results of the project to date indicate success in implementation of the school based portion of the Internet service in both Fauquier and Alexandria, primarily because of the minimal need of the school based participants for technical assistance and training. In additional, the widespread Fauquier community interest in Internet service, especially World Wide Web service, has led to the initiation of three alternative ISPs within six months of the project start, which all owed the University to substitute a locally provided graphical ISP for the text based system the project initially started with.
The author wishes to thank the invaluable assistance and input provided by Arthur Melmed, J.W. Harrington, Kingsley Haynes, James Pfiffner and James Finkelstein.
This paper presents preliminary results of a community based information infrastructure development project, funded through the Telecommunications Information Infrastructure Assistance Program (TIIAP), of the National Technical Information Agency (NTIA), U.S. Department of Commerce. The demonstration project (in progress) links a variety of actors in the Washington D.C. metropolitan area, as a means of demonstrating the role of inter-organizational partnerships in promoting information infrastructure development. The primary participants include public schools systems in two geographically separated jurisdictions, an academic institution (George Mason University), acting as system integrator and policy consultant; and the federal government which provided initial funding support.
The geographic variable is particularly of interest here because there is considerable difference between objectives for two project subsystems. In the urban portion, in the City of Alexandria, a close in suburb of Washington D.C., the project objective was simply to provide Internet connection to two elementary schools, and demonstrate the educational advantages generated by this type of access. In Fauquier County, an ex-urban county part of the Washington D.C. SMSA, the project was more ambitious, in that there were two objectives, 1) provision of educational related Internet access to three public schools in Fauquier, and 2) provision of community access to the Internet as a means of stimulating community and economic development. Given the current arguments that information infrastructure can be used as a tool for economic and community development, the second objective was included to examine the effects (if any) of a pilot information infrastructure project in an under-served, rural area.
As of the date of this writing (July, 1995), results of the project indicate success in implementation of the school based portion of the Internet service in both Fauquier and Alexandria, primarily because of the minimal of the school based participants for technical assistance and training. With respect to the community related aspects of the project, the consistently high interest on the part of participants, and the continued willingness to work with a system that has gone through several major changes indicates that there is a demand for Internet access from community based users. Further, the arrival of three different ISP start-ups indicates that the market is beginning to respond to this demand.
The ability to provide information and services at a distance from their source has been made possible by advances in computers and information technologies. The interlinkages of computers and resources to create an information infrastructure has, for lack of more suitable terms, been discussed using various physical infrastructure analogues -- "the information superhighway -- being most common. The proposed names for this virtual highway -"Cyberia", "Infobahn", "I-Way", National Information Infrastructure (NII), or the ubiquitous "Internet", reflect the subtle differences in perception and visions for this information infrastructure.
By a combination of happenstance and design, the Internet has grown in a piecemeal fashion into an international matrix of users, computers, and information. This, coupled with what amounts to virtually free access for many users, has driven its exponential growth. From an initial architecture of only 200 host computers in 1983, the Internet has grown to at least 5 million computers in 94 countries; an expansion estimated to be at a rate of 9-12% a month.3 Industry estimates indicate that there are already in excess of 20 million users worldwide of the Internet in its current form.4
Many observers, however, have noted that the very success of the Internet is beginning to present a problem. The increasing size and volume of information flow threatens to break down the ad hoc nature of the system, leading some users to consider the unthinkable -- the need for a central authority or "manager" for the Internet. Under the current process, growth simply occurred as one system "hooked" onto another an d extended in this fashion. However, the increased density of information flow means that this pattern is increasingly unwieldy. Thus the recent calls for implementation of some sort of planning or oversight function to prevent the system from collapsing under the weight of conflicting uses, mismatched infrastructure and different policies for operation. The fact that the Internet has proven an extraordinary success in not in doubt, but fears that the system is already beyond the parameters originally envisioned, has led to proposed alternative schemes for operation and design of a "next generation" implementation.5
In the process of planning for a "next generation" Internet system, the "goals" or objectives become critical. These should include issues of access, and equity, especially for those areas that may be "bypassed" by the information infrastructure because of location or economic reasons. To some extent the information infrastructure is already being developed through efforts of the private sector; telephone companies, cable-TV, and even some utilities, are investing in physical plant. Alternatively, large nationally based information utilities such as America Online, CompuServe, and Prodigy, as well as smaller access providers, are expanding their reach. In many cases, especially in areas beyond large urban centers, access to these utilities requires the use of costly 800 numbers, isolating communities that might most benefit from the information content.
However these models are not the only (or even the best) approach to providing access to the "Infobahn", there remains a range non-commercial alternatives that can also meet community needs. Just as the Rural Electrification Administration was instrumental in promoting the extension of electrical infrastructure to under-served areas, the development of adequate information infrastructure will require a variety of policy initiatives, making use of private, public or a combination of both sectors. Regardless of the alternative chosen, there is still much that local communities can do to expedite these linkages.
While the information infrastructure (in whatever form it takes) can be expected to develop in areas of higher population densities, or where existing infrastructure exists, market mechanisms are more likely to favor the more affluent urban/suburban areas. It is in underserved or rural areas, or such areas as inner cities, where the low density of certain resources is analogous to a "rural" area, that a variety of alternative approaches to infrastructure development can be useful.
Several models exist for this ranging from a full scale public effort, such as the Sailor Network, a public Internet access provided by the Maryland library system, to public-private efforts such as The North Carolina Information Highway (NCIH) is a product of a public-private partnership between North Carolina and various telephone companies; Blacksburg Electronic Village, in Blacksburg, Virginia; to local non-profit or community based efforts such as Telluride in Colorado 6; the Cleveland Freenet, or the La Plaza TeleCommunity Foundation in Taos, NM. A variety of groups have formed on a national level to promote community based alternatives to the private telecommunication and information providers including: Center for Civic Networking (CCN), the National Public Telecommuting Network (NPTN), and the Morino Foundation.7
Getting the area "wired" however seems to be just the start of the endeavor. A number of authors have noted that the presence of a physical infrastructure (that is the literal hooks into the Internet architecture) implies that there is "something" to hook into the Internet. This means that organizations must acquire the computers and related peripherals, as well as supporting hardware and ancillary phone lines. But perhaps a greater barrier, though not as immediately obvious is the "soft-infrastructure" that is required to maintain the information flow.
This would encompass not only the technical skills required to operate the systems, but also the critical training to provide support to the individual endusers. On another, integrative, level this would include management of the reverse information flow if the organization is to be a provider as well as a consumer of information.8 Thus any meaningful development of community infrastructure must also take these issues into consideration. This is one distinct area where the activity of an intermediary, be it a community freenet, or other organizing group, can centralize and provide more efficient management of the infrastructure and linkages.
Finally, as is the case of any policy oriented activities, the tremendous efforts made to demonstrate project feasibility, leaves transmission of the "lessons learned" as a weary afterthought. The need to evaluate and assess the benefits and costs of projects , and the transmission of results, is essential to the effective development of new systems. This paper presents preliminary findings of the project, and reviews possible alternative options for the future.
Subsequent to funding received under the Telecommunications Information Infrastructure Assistance Program (TIIAP) of the National Technical Information Agency (NTIA), Department of Commerce; George Mason University, Fauquier County (VA) and the City of Alexandria (VA) initiated a 12-month policy demonstration of the role, function and benefits of an information and communication infrastructure in community and commercial life -- The Regional Information Infrastructure Policy Project (RIIPP), in October 1994.9 Initially organized around selected community needs for education, training and public information, this project focuses on community and educational access to the Internet in Fauquier County, and more narrowly on educational access in the suburban Alexandria jurisdiction.
In general, the RIIPP project was not intended to become a local ISP (Internet Service Provider), but was mainly concerned with "start-up" issues, such as barriers to entry. The initial overriding theme was to demonstrate the possibility of developing a public based or supported community network based in a technology-ready public school system. RIIPP’s successful demonstration of the Internet service in both Fauquier and Alexandria schools, as evidenced by the high participation rate of target teachers and the participant feedback, is due in part, to the minimal need for technical assistance, mostly for training. As regards the community access portion of the project, the school-site technology and typical school operating hours were, as a rule, inadequate to meet the service demand of an enthusiastic community in Fauquier. Specifically, the need for user support (user "hand-holding"), as well as the system support and maintenance generated by the extension of infrastructure design, was underestimated.
When the local (Fauquier County) government indicated only minor interest in providing a public based, or supported community network, the anticipated demand for community service was shifted to a private Internet Service Provider (ISP). It had initially been anticipated that community use could be provided for through a community freenet based wholly on a governmental, or non-profit based model. The absence of strong support for this approach by any single group, or the failure of a coalition to pull together to address this need, indicated that any near-term extension of the information infrastructure to this area would require an external agent (in this case the RIIPP project) to facilitate implementation of an information infrastructure.
In the more urban Alexandria (just across the river from Washington D.C.), by contrast, there were strong, but conflicting interests in developing a community network. At present, these actors have not come to closure on goals or objectives for the community, and there is no indication that this will be rapidly resolved, leaving most schools, and public and community agencies to shift for themselves. Several different groups have announced plans for providing community access, or provision for community related information and local government information. In addition, the City of Alexandria has initiated a trial, and tentative linkage to the Internet, by establishing an e-mail address through a local community network.10 In this case, development of a local ISP is not relevant in that a wide variety of possible access providers to the Internet already exist.
In Fauquier County, units participating in the demonstration project were selected on the basis or readiness, representativeness and a measure of geographic concentration. These included direct access by three schools in Fauquier County Public Schools (FCPS); the Fauquier County branch of the Virginia Cooperative Extension Service (VCES); and selected members of the Fauquier community, with approximately 250 users provided free dial-up Internet service for demonstration and planning purposes.
In the City of Alexandria Public Schools (CAPS), the two participating elementary schools were selected because of the significant number of disadvantaged students and the desire of district officials to more rapidly implement plans to network all schools. Aside from the primary participants described above, additional input was provided by other members of the community, including individuals, organizations and governments. This varied from participation in several focus groups held at several points during the project, for both design, and evaluation purposes, at one end; to actual end use by citizens in Fauquier, with the related feedback on system operation.
The Institute of Public Policy (TIPP), the University’s lead agent for the activity, maintained overall project management and policy evaluation, and additional technical assistance was provided by the Electronic Commerce Resource Center (ECRC) staff, an d from the School of Business. The University’s Institute for Educational Transformation (IET), under separate contract to Fauquier County Public Schools for the preparation of a school restructuring plan, coordinated teacher training in telecommunications and for the instructional application of telecommunication services.
Additional assistance in Fauquier, both in community support and citizen training was provided by Learning and Beyond, a community-based firm. The project team had initially intended to use George Mason University’s computer resources. Subsequent consultation (during September and October, 1994) with the computer support staff determined that existing University demand would prohibit additional external loads, until completion of improvements to internal infrastructure was finalized. By the end of October 1994, TIPP determined that an alternative access to the Internet would be necessary. This alternative -- access through an external service provider -- was determined to be appropriate for the duration of the project. Internet service (basic interconnection to the Internet - allowing e-mail, ftp, etc.) was provided by PSI, one of the largest Internet service providers in the United States. Subsequent adaptive changes, as well as the original network system architecture, were completed by PRC Inc., of McLean, Virginia, which prepared the ‘scaleable’ design linking the Fauquier County and Alexandria architectures.
At the end of the twelve-month demonstration phase, in Fauquier, the three target public schools, the VCES Office in Warrenton, and the individual community users, will have network access and be linked to the national information infrastructure for instructional and administrative purposes. In Alexandria, the more limited project goal was to provide in school access to the Internet.
It was anticipated that the successful implementation of this project in Fauquier would result in local access to the Internet either as a public service, with the county allocating revenue for ‘wholesale’ (block and/or quantity discount on cost of service) acquisition of Internet services from the State, or from a private provider. This in turn would generate a subsidy for economically disadvantaged households, should federal law require this in the future. Alternatively, successful implementation would demonstrate a demand exists for Internet access and private local service will be offered.
For Alexandria, successful implementation would result in significant access to the resources offered by the Internet, which did not exist before the project started. A successful implementation of Internet access would be evidenced by teachers in the target schools using information sources available on the Internet to augment their class presentations.
While use of traditional text based Internet services like e-mail, ftp, telnet, and gopher continue to grow, the character of this network of networks has become more commercially oriented. As such, there is increasing concern with considerations generally related to commerce such as privacy, security, access; and who pays, and how. The recent development of graphically based interfaces to the Internet has significantly increased both the number of users on the system, and users’ expectations, while concurrently increasing demand upon system throughput. The cost and value of digital community-wide communication and information services, however, remains largely untested and unknown. This policy demonstration project aims to help answer the questions of cost and value for Fauquier County and the City of Alexandria as well as for other similar areas of the nation.
Basic Range of Services: Basic services are defined here as e-mail, ftp, telnet and gopher, which can be used by teachers, students, State and local employees, and individuals after a modest amount of training, and dial-in modem access to an Internet server with a basic terminal device. We anticipated a demand for e-mail within and outside the jurisdictions; telnet to libraries outside the County, like the Library of Congress and the National Agricultural Library; and gopher to federal agencies in the Washington, D.C. area. Improved communication between school and home by e-mail was initially planned, with teachers responsive to inquiries from both students and their parents, system implementation problems limited this option.
Educationally-enhanced Internet Service: Educationally-enhanced service is defined here as pointers to databases and other sources on the Internet useful for instruction in subjects like English, geography and science education, which can be used by teachers (and students) after minimal additional training and assistance beyond basic Internet training. Additional training and assistance for teachers included lectures, and print and video materials explaining and demonstrating the instructionally effective use of telecommunications in the class room.
System Architecture: The system’s architecture employed some interesting solutions to the project requirements. The key goal was to establish an Internet connect in a rural area, about 50 miles from Washington D.C. This area, though on the extreme urban fringe of Washington, required a long distance phone call to reach an Internet service provider. The system design then essentially served to stretch the LATA (area in which local calls are "toll free") in order to eliminate the long distance barrier to Internet service.
The system design served to "extend" the Fauquier School network by connecting it to the Alexandria link to the Internet. A 56kbs Internet connection to Performance Systems Inc. (PSI) linked Polk Elementary (in Alexandria) to the Internet, and another leased line connected the other Alexandria school, Mt. Vernon Elementary, to Polk. From Polk, an additional dedicated line provided the Internet connection to the Virginia Cooperative Extension Service in Warrenton, about 50 miles distant. The original design called for the Extension office to be connected to a hub at Liberty High School in Bealton (about 5 miles distant from Warrenton), which in turn was connected to two other local schools, all via wireless bridges. Dial-in for the community was to be provided through two access points, one Liberty High School in Bealton, and an additional access point provided in Warrenton. Access to the servers was provided via two Shiva LanRovers (routers) configured to act as terminal servers.
The system architecture required several major modification in order to be functional. Due to unanticipated geographic considerations, the wireless bridge between the VCES in Warrenton and Liberty H.S. in Bealton was replaced by a leased line. A second major problem arose with the dial in access point. While Internet access was successfully achieved from on-site, the dial-in capabilities were never reliably implemented. Ultimately access was provided for the approximately 250 community users through an arrangement with an alternative local service provider, that had started partially in response to the demand demonstrated by user response to the project.
Fauquier County: Fauquier County, Virginia is located approximately 40 miles southwest of the District of Columbia, and 95 miles northwest of Richmond. Nearly 95 percent of the county’s land area is farmland and woodland. The county’s 1990 population of about 49,000 represents a 36 percent increase over the level in 1980. Even with this rapid growth, Fauquier County remains one of the most sparsely populated localities in Virginia with fewer than 75 persons per square mile, nearly one hundred times less densely populated than Alexandria.
This scenario, typical of many urban fringe counties in the U.S., results in certain economic and demographic characteristics that made Fauquier County especially suited for this project. The agricultural sector of the U.S. economy is in rapid transition, experiencing large substitutions of capital and technology for labor, resulting in a decline in the number of farms and jobs in the sector as a whole. Moreover, the productivity of farms in the southeastern U.S. is not competitive with that of farms in the midwest and western states, creating a regional competitive disadvantage for communities like Fauquier County. Thus, in order to keep public revenues up in the face of a relative decline in the local economy, land is shifted to non-farm residential uses. More often than not, new residents in urban fringe counties do not work where they live, but commute back to a closer-in suburb or the urban core.
This is the current situation in Fauquier County, where more than three times as many people commute out of the county to work as commute in. Traditional job skills are no longer in demand, and young people assume they must leave the community to find meaningful, well paying job opportunities. The ability to retain young residents is particularly important for Fauquier, with more than three persons 19 or under for every person 65 or over, one of the highest ratios in the state. Moreover, Fauquier ranks llth state wide in the percentage of married-couple households with children, at 36.4 percent. Counties like Fauquier face economic and geographic constraints that increased abilities to use and exploit information technologies can offset.
The schools selected for this demonstration project tended to have higher than average numbers of students from low income households, and close to or greater than the county average proportion of African American students compared to whites. Thus these students stand to gain from increased access to the educational resources and opportunities available through the Internet. As of the end of April 1994, the roster of students in Fauquier County Public Schools included 8,446 students: 7,289 white; 12 American Indian; 1,013 African American; 56 Asian; and 76 Hispanic. White students made up 86 percent of the student body, and African Americans accounted for 12 percent. All other combined made up the remaining two percent, and any other individual racial group accounted for less than one percent of the total. Hence, it makes sense to examine racial diversity in terms of the percentages of whites and African Americans.
Alexandria: By comparison, Alexandria is a highly urbanized area, characterized by extreme racial, ethnic, and economic diversity. The city is the most densely populated locality in Virginia, and the 11th most densely populated city in the United States, with 7,281 people per square mile. Twenty-two percent of Alexandria’s population is African American, compared to 2.6 percent for the entire state of Virginia. The city’s 9.7 percent Hispanic population is second highest in the state. This diversity is clearly r eflected in the composition of the Alexandria City Public Schools (ACPS), and particularly within the two elementary schools selected to participate in this study.
Of the approximately 9,700 students in the ACPS system, 2,720 or 28 percent are white; 4,670 or 48 percent are African American; 1,650 or 17 percent are Hispanic; and 620 or 6.5 percent are Asian. The schools participating in this demonstration project r eflect in some ways the ends of the spectrum in terms of racial, ethnic and economic. The student body of Mt. Vernon elementary school are 16.2 percent white; 54.5 percent African American; 26.5 percent Hispanic; and 2.7 percent Asian. In contrast, stud ents at James K. Polk elementary are 36.7 percent white; 36.9 percent African American; 19.6 percent Hispanic; and 5.8 percent Asian. Not only do they provide an excellent frame of reference to the rural situation in Fauquier County, but they offer intere sting contrasts within the Alexandria system itself.
The project can be divided conceptually and practically into six major tasks: establish management and organization; completion of network design, and equipment acquisition and installation; training; initiation of service; and planning for future implem entation. Each of these is described briefly below.
Establish Management and Organization: Team members were selected for the project management team and five implementation teams: network design, site management, service development and maintenance, policy, and evaluation. This included participants from both communities, as well as from the school systems, the Virginia Cooperative Extension Service and George Mason University, and PRC. Project planning took place during October, 1994, and the actual system implementation began during November, 1994
Equipment Acquisition and Installation: The flexible and responsive nature of this project required that network design, acquisition and installation of all necessary equipment, continue throughout the project duration. The various elements of the project came on line sequentially, with the t wo Linux (Unix-like) servers operational in April, 1995, and partial email capability available in May, 1995. The continuous problems with the dial-up capabilities in Fauquier were not successfully resolved in a manner which provided satisfactory service given the system architecture (detailed above).
In order to provide the participants with reliable and predictable service, GMU contracted with a local service provider (which had not existed at the time of the project start-up) to provide expanded Inte rnet access to community participants, which began the second week of July, 1995. The compressed transition period (due to certain cost requirements), required that training on a new system be initiated quickly. After some initial technical and coordina tion problems were solved, the transition to the new system went smoothly, and was well received by the participants. Internet access in Alexandria was limited primarily to an in-school basis, with partial in-school Internet access (including WWW) available at both Polk and Mt. Vernon completed by mid-May, 1995. Email access for individual teachers became fully available by June 1, 1995.
Training Internet services are not in the appliance category at this time, and even though software clients such as Web-browsers (Mosaic, Netscape, etc.) have greatly simplified usage. Thus training activities, especially for teachers, were emphasized in order to allow a ‘true use’ picture to emerge. Training for individuals in Fauquier was conducted by a local training and public relations firm. The initial enthusiasm for "getting on the Internet" generated a rush for training which overwhelmed the instructors , and left many of the users feeling that they had been neglected.. Subsequent training session required reservations, the number of instructors was increased, and training materials were revised to be more explicit. In the second phase, a "training of trainers" approach was utilized in order to build a self-sustaining capacity within the communities. This took the form of establishing a group of 11 "group leaders" with advanced training or expertise (preexisting in most cases) who provided user support and served as "guides" to the Internet.
Initiation of Service: Initial service was brought up May 1, 1995, in the schools, delayed by a design upgrade that allowed the use of graphical Internet web-browsers, which had not been available at the time of the initial project design. Community service, was brought on-lin e by May 8, though consistent and reliable service was not established using the original system architecture. As of the date of the report, community based users have been transferred to a locally based ISP, and initial results and use rate have not be en assessed. It was determined that due to certain, unresolved policy issues related to content, that Usenet news group access would not be directly provided.
Evaluation: Evaluation was ongoing throughout the project and designed to provide Fauquier, the City of Alexandria and like regions of the nation with planning and design data, and cost and value information for implementation of community and national digital info rmation and communication services. Initial system design elements were generated by input provided from a series of electronically conducted Internet surveys of both Usenet newsgroups, and local Bulletin Board operators; and through qualitative group res earch (focus groups conducted in Alexandria and Fauquier). Experience level and levels of training deemed necessary for community participants were derived from the initial application form submitted when applying for a user account in the demonstration project.
Ongoing monitoring both of use rate, and of user feedback was conducted on an informal basis. Final project evaluation will include a detailed survey of user experience, and in-depth interviews with selected users and participants in the process.
Planning for Future Implementation: Specific implementation plans, or transitional recommendations for project completion are in progress, for Fauquier County and the City of Alexandria including network design and financial planning as appropriate.
The initial assessment of this pilot project (which is scheduled to run through September 30, 1995) has produced some instructive results. A significant amount of project resources and staff time was required to accommodate the Fauquier County community users. This points out revealing differences from the needs and requirements of the different sets of users. For example, community residents reveal a much wider range of computer skills than teachers, placing greater demands on external training and help services than the in-school participants (i.e., teachers), who are typically able to assist one another after a period of initial training. Community residents use a wide variety of hardware platforms and communication software products, placing gre ater demands on the Internet communications infrastructure than the relatively well integrated school networks. And with widely varying levels of computer knowledge and skills, and varying interests and motivations, community residents make greater deman ds on any type Internet service providers.
Further, the results from the Alexandria phase differ from the more ambitions Fauquier phase. While schools in both Fauquier and Alexandria were technology-ready, neither would be likely to have implemented Internet access in the absence of a catalytic a ssistance organization (in this case, TIPP). As the technology for Internet service is not (yet) off-the-shelf, and because of the necessary training, all most all of the participants required training and assistance in acquiring and operating the technol ogy. Given this experience, it is not unreasonable to expect that this will be the case elsewhere, and that other small public and community agencies may need some kind of assistance in acquiring and in effectively using Internet service.
While the proliferation of ISPs is working to keep the price of Internet service down, ISPs do not see themselves in the role of assistance organizations. This was the case in Fauquier where, by mid-summer 1995, there were at least three local ISPs, onl y one indicated interest in providing the assistance that 80% of the community users required, as well as the access. However, the tremendous energy and willingness of the group leaders served to fill a critical gap. It can be assumed that grass-roots co mmunity networking efforts that aim at fulfilling an assistance role may more rapidly succeed in bringing Internet service to the largest number, as opposed to efforts aimed at aggregating demand in the interest of low service cost.
The availability of community access accounts was formally announced in March 1995 by means of press releases in two local Fauquier newspapers and through notices to organizations like the Fauquier Chamber of Commerce. An application form with Terms of Service Agreement (TOSA) developed by project staff was widely distributed in the community. The application form elicited information from applicants on their interest in using the Internet, on their computer equipment and modem speed, on their level of expertise with computers and Internet service, and on their ideas for use of Internet service on behalf of the community.
The TOSA advised on the time-limited nature of the RIIPP project, and secured the agreement of applicants to the specified rights and responsibilities of the involved parties and limitation of liability. The 200 community access accounts available were quickly subscribed. (An additional number was set aside for interested teachers and county officials, numbering under 100.) Over 50% of the applicants considered themselves advanced computer users, whil e less than 10% considered themselves novices. Eight in ten described themselves as Internet novices.
Five three-hour Internet training sessions were provided (on different dates, in order to accommodate the far flung residents of Fauquier County) by an experienced local firm, Learning And Beyond, Inc., three basic and two advanced. The well attended, ha nds-on sessions were conducted at Liberty High School, and revealed substantial disparity in computer knowledge and skills among participants. In consequence, experienced computer users were recruited and organized to serve as a self-help group for others.
The smooth course followed by the hands-on training process did not prefigure operational problems, the result of interaction effects, which soon reduced the quality and reliability of dial-in Internet service for all but the most knowledgeable and dete rmined community users. We describe the contributing factors, starting with a rational, i.e., cost-effective, design decision by PRC Inc., the project’s systems integrator. Dial-in lines were provided at two sites in Fauquier County (in order to limit the access cost to a local telephone call for participants throughout the county) connecting to a Shiva LanRover. (The same design was implemented in Alexandria for more limi ted dial-in service for school teachers only.) An assured effective and cost-effective solution when all users are using specified communication software products, this design solution was selected (in place of the more expensive terminal server), when a new Shiva software release prom ised satisfactory service for any mix of communication software products.
In practice, this solution failed as users with a wide variety of communication software products casually logged on and off, without necessarily following specified protocol, and the Shiva device failed to reset lines automatically that were not reset o perationally by users. One-by-one, dial-in lines were effectively removed from service, until the Shiva device could be manually reset. Liberty High School and the office of the Virginia Cooperative Extension Service, the two sites for dial-in lines, are not staffed for 24-hour operation. In late afternoon, evening and night hours, and during weekends and holidays, when community users are most likely to be active, no one is available to reset the LANrover, often resulting in a steady deterioration in the quality and reliability of service. (The same technical solution did not create the same practical problem in Alexandria, with the m uch smaller number of more careful teacher users, using a specified communication software product.)
Finally, advanced community users wanted World Wide Web service, which is widely viewed by the public today as synonymous with Internet service, but which was not promised in the project proposal ... and for which adequate system capability was therefore not provided. The interaction among these three factors combined to create frustrating operating conditions for community users. In terms of statistics, in Fauquier, while virtually all of the project teachers had made almost daily use of the Internet, the difficult community access can be seen in the use rates. A survey of the 230 Community users found that:11 63 (27%) have nev er logged in 28 (12%) have not used the system since May, 1995 140 (61%) used the system in June, 1995 Looking at usage rates for June: 45 (19% of total users) last logged in between June 1-15, 87 (38% of total users) logged in between June 16-30. And so 41% of users have accessed the Internet server since June 16.
While this is less than would have been desired, the intermittent service problems may have served to discourage users. Initial (and ongoing technical problems) generated increasing user frustration, and drop off in attempted access. This was due to two different causes. First, inexperience with some of the complexities involved in combing multiple and incompatible technologies (several different kinds Apple computers, Windows systems, Novell Networks, Linux, and Unix based systems, coupled by unproven wireless bridges) made it difficult to isolate internal system problems. Several start up problems were related impleme nting email access for the twin Internet servers. These problems were successfully resolved for internal (school Lan systems) users by the third week of May, 1995. As noted previously, consistent and reliable community dial-in access, however was never achieved.
The inability of the system designers to stabilize and bring up reliable dial-in connects led to increased (and vociferous) complaints from users. It became apparent notwithstanding the materials distributed to the community users, that the 40 hours a week of support that was guaranteed as part of participation in the project was insufficient to ensure user satisfaction. Many of the problems occur red at peak use, which was "after hours: relative to planned support. Maintaining system function necessitated extensive unanticipated use of staff resources on weekends and late into the morning hours.
What lessons can be extracted from the Fauquier portion of the project? First, training and development of "group leaders" needs to be undertaken as one of the preliminary tasks, optimally, in conjunction with technical system implementation, rather than as a "downstream" activity. Greater use can be made of community resources in maintaining the operating systems, taking advantage of the significant reservoir of technical expertise. As an extension of this, it makes sense to develop a volunteer organi zation, and bring them "on-line" before the project is significantly in place, preferably as part of a community planning/visioning effort.
The results of the more limited Alexandria portion of the project can be distilled as follows: Focus on the project in Alexandria was limited to providing in-school access to the Internet for teachers and selected additional users. The reduced complexity of the project reduced, but did not eliminate implementation problems. While no formal acces s was provided for community users, some "unofficial" dial-in capability was available. More limited in scope, the teachers indicated that they were happy with Internet content, but were not prepared for the slow system response at peak loading periods. Testing and investigation revealed that the problem was due to overall Internet use rather than design related system problems. In terms of types of usage, Alexandria teachers made extensive use of the graphical browsers (Netscape) to visit educational and resource related sites. Further, email was used to have classes "write" to each other, and to other classes in geographically distant areas. The teachers noted that the graphics and sound capabilities (provided by the Apple systems) served as a tremendous draw even given the limitations of system speed. The teachers in each of the school (Polk and Mount Vernon) also collaborat ed in the development of school oriented "home pages" linked to the City of Alexandria’s home page.
For the project’s first nine months, technological developments have so far outpaced political creativity in Fauquier. Much has been accomplished; much still remains to be done. While demand has been demonstrated for Internet access, the cost and benefits of a public information infrastructure in general are difficult to determine. With an operational communications infrastructure, trained teachers, and curriculum projects under development, county schools have achieved their goal for educational connectivity. Absent financial guarantees, and under heavy budgetary pressure, the school administration appears, at present, unwilling to commit itself to meeting the new costs of serving as the nexus of a community information and communications infrastructure.
The public library system has been successful in getting itself funded for Internet connectivity, but prefers to go its own independent way. Steps are being taken at this writing to transfer responsibility for community Internet service from the school communications infrastructure to a private Internet service provider for the remainder of the project term. In the absence of assured further federal project support needed to develop and implement a community information infrastructure, this action may help ensure the economic viability of at least one local Internet service provider for the county for a short period of time before the market decides. Alternatively, interested community residents will be forced to add the cost of long-distance dialing to the cost of Internet service, a clear disincentive. Political leaders appear to have a wait-and-see attitude. No unit of government appears to be a natural locus of planning and leadership.
In Alexandria, the project’s focus was limited to providing in-school access to the Internet for teachers and selected additional users. The less complex design of the project reduced, but did not eliminate implementation problems. While no formal ac cess was provided for community users, some "unofficial" dial-in capability was available. More limited in scope, the teachers indicated that they were happy with Internet content, but were not prepared for the slow system response at peak loading periods . In terms of the large issue of community access in Alexandria, the promise of an institutional network provided at low cost by the chartered cable-TV operator appears to satisfy some (though not all) organized community pressure. However, it appears that without an outside agent to assist (TIPP/GMU in this case) local development should be expected to advance only slowly, as the various actors resolve their differences.
1. While small city and county governments may not believe themselves ready or equipped to deal with the demands and costs of developing an "on-ramp" to the information superhighway, many of the necessary skills, and much of the physical resources, are al ready in place in the public schools systems. It may be necessary to change the community’s perceptions and thinking about how to leverage the available resources. This can be accomplished through an educational process that presents some of the possibl e options that are available, and that can be readily implemented through creative use of existing resources.
2. Setting concerns about technology aside, it seems reasonable to conclude that local government is not ‘up to speed’ on the meaning or implications of information infrastructure for economic and community development. This is not a condemnation of local government, or even unexpected; technology changes so rapidly that even experts in the field have a hard time staying current on the latest wrinkles in information technologies. In this case the use of an objective outside consultant can help, and serve as an organizing focal point that may sidestep problematic local stakeholder interests.
This applies to all policy as well an technical areas. From a technical standpoint networked systems multiply the opportunities as well as complexities involved in the delivery of information. While the underlying technologies may be understood by netwo rk vendors, it is critical make sure that they either have had actual experience in delivering access to the Internet in a networked environment, and utilizing the specific type of equipment that is extant. Barring this, the next best option in to insure that they have the technical expertise as well as a commitment to completing the project, in a satisfactory manner.
3. Determine the specific goals that are to be achieved. "Access to the Internet" is not a very clear goal, and is analogous to saying that "a library should have books." The community should have clear ideas about what constitutes an appropriate mission statement. For instance: "citizens should have access to several text based services including, email, remote searches of databases at the library, and updates on local governmental activities through a local dial-up connection." This can be accomplished through the use of Visioning or focus groups, or other polling or survey type methodology. This as the added advantage of generating the community motivation necessary to develop a viable community information infrastructure.
4. Schools with the necessary on-site infrastructure can act as the technology engine for a local community, at least initially, in the form of a pilot or demonstration project, and subsequently, generate community support and interest. This has additio nal policy implications particularly in times of budget shortfalls. Leveraging community resources, be they infrastructure, economic or human resources, shows how the entire community can benefit from investment in the public school system, not just th ose individuals with school aged children.
5. While "physical infrastructure" (e.g., hardware and software) are generally thought to be the critical entry barrier in implementing a computer network, this did not prove to be the case in the subject project. Thus, while the infrastructure may be in place in the school system, systems implementation, is a non-trivial additional task that requires human and technical resources for the selected school system. This is not an insurmountable problem, but it does require careful planning and objective set ting.
6. School based infrastructure offers a convenient locus for adding local community information infrastructure to superhighway services; through the use of centralized existing sources. On the other hand, alternative dedicated services providers such as cable-TV operators might be more suited for combining TV and Internet services in a single delivery system.
7. Other potential players may suddenly appear when sufficient community interest is generated. These include: locally franchised cable-TV operators, telephone and or wireless communication companies, utilities (there are networks that have been proposed by water, electric and gas companies), alternative Internet service providers, or even other public or public/private sector interests. Communities should consider these groups and organizations an asset to be used in designing a community system. Comp etition for users may generate offers of technical support and or physical resources that can be used to supplement limited community resources.
8. Finally, while a variety of successful models exist for developing community based information infrastructure, in general, they have several key elements in common. Typically there is partnership involving several different community groups. The acto rs/participants with diverse interests ensures that there is a broad base of support, and that the participants will perceive that the system is "theirs." This brings another key component into focus. A well planned project also looks to the future. As ide from the range of groups involved what elements meet the needs of both the citizens as well as the public sectors stakeholders. Flexible planning is also critical. A project "fixed in stone" will find that the rapidly changing world of telecommunic ations may leave the best planned project "high and dry" when the tides of technology shift.
1 Mr. Baker is a member of the research faculty, and a doctoral candidate in public policy atGeorge Mason University.
2 NTIA Grant 51-40-9404
3 Mann, Charles. 1995. "Regulating Cyberspace." Science Vol. 268 5 May 1995. p 628.
4 For an excellent, in-depth overview of the Internet see: The Economist "A Survey of the Internet" Vol. 336, Number 7921 (July 1, 1995).
5 Brody, Herb. 1995. "Internet@Crossroads.$$$" Technology Review Vol. 98 (4)24-31.
6 Carlton, Jim. 1995. "Mounta in Climbers" Wall Street Journal p R30, June 19, 1995.
7 Cisler, Steve, 1994. "Community Networks: Past and Present Thoughts." Paper presented at "Ties That Bind: Building Community Networks" Apple Conference Center Cupertino, CA, May 4-6, 1994
8 See Agenda for Access, Public Access to Federal Information for Sustainability through the Information Superhighway. 1995. Washington D.C.: The Bauman Foundation.
9 NTIA Grant 51-40-9404
10 It is interesting to note that rather more extensive services are provided by the adjacent Fairfax and Arlington Counties.
11 The survey was conducted during the second week of July, 1995
For more information, please contact:
Mr. Paul Baker at pbaker@gmu.edu