The technologies involved in these transformations — networking and the Internet, decision support systems, electronic commerce, knowledge discovery tools, geographic information systems — are not necessarily the most advanced tools available. Public sector innovations tend to result more often from the infusion of well-developed technologies into the complexities of governmental programs and processes, in an environment with many stakeholders and competing values. In this section, we highlight some of the technologies that support this transformation. We also point out special issues that affect how government takes advantage of these technologies and identify areas where additional research is needed.

In a relatively brief span of time, the Internet has led to significant changes in how public agencies disseminate information, how government staff interact with each other and with people outside government, and how government delivers services. The most obvious examples are government WWW sites. Agencies ranging in size from large Federal departments to small towns and villages have public Web sites, designed to meet their high-priority objectives — economic development, tourism, information about government services, purchasing, statistical data, or selected service transactions. An electronic mail address on the Web site makes elected officials accessible to their constituents and allows citizens to communicate directly with public employees.

In addition to person-to-person communication, the Internet is increasingly being used to exchange data between organizations. This includes record-oriented transactions, real-time querying of remote databases, as well as larger exchanges of complete GIS data sets or other databases integrated into data warehouses. Processing of workers compensation claims, for example, may involve real-time transactions against databases from a half-dozen organizations to verify eligibility.

The use of the Internet for public purposes inevitably raises the issue of equal access. In 1997, about 17 percent of private households in the U.S. had direct access to the Internet and these were concentrated in middle- and upper-income areas of cities and suburbs. Often those most in need of government services are those least likely to have access to the Web. Rural areas, with less likelihood of having high-performance technologies such as ISDN or cable modems, are at a disadvantage, as are lower-income households, grass- roots community organizations, and small businesses that often lack direct access. While Internet access through schools, libraries, or other public places is increasingly available, people without direct access remain at a disadvantage compared to the connected minority. Given this uneven distribution of access to the Web, traditional service delivery through telephone, mail, and face-to-face interactions will be needed for many years to come.

Communication tools support and nurture linkages and relationships that were not possible through more formal means of communication. More and more often, we expect people, including government staff, to have and use electronic mail. Discussion listservs and shared Web sites routinely connect distributed organizations and virtual communities, fostering increased discussion and cooperation among those who share a common interest. Easy public availability of such information as government contracts or grant programs fosters greater equity and efficiency of government purchasing and the distribution of public resources.

Videoconferencing is another technology that is making its way into standard government practice. For example, video technology is used today to interview crime victims who would otherwise have to travel long distances to a police precinct or court house. At present, however, the use of videoconferencing typically consists of dedicated facilities linked by telephone lines. As a result, the technology tends to be used in a localized and specialized fashion. As Internet videoconferencing technology matures, it is likely that many more such interactions will take place.

These relatively ubiquitous capabilities are being augmented by more advanced collaborative tools in such areas as distance learning, just-in-time training, and anytime-anyplace meetingware. Use of the Internet and video conferencing techniques to deliver entire curricula from remote sites extend higher education and lifelong learning to many who would otherwise not be able to attend classes. Distance courses in specialized topics enable elementary and high school students to pursue studies unavailable to them in their home districts. Thanks to networked collaboration, these students can even conduct joint science experiments with their counterparts around the world. Although the pedagogic effectiveness of alternate modes of study and instruction are still being evaluated, it is clear that network-supported learning will play an increasingly important role in the future of American education.

The Internet also has the capability to extend expertise across physical distances. In medicine, for example, a specialist can expand his sphere of effectiveness, without traveling, by remotely reviewing diagnostic tests. This technological capability has not yet been extensively used in the U.S., though, in part because it requires changes in insurance rules as well as changes in the culture and traditions of medical practice.

Data visualization, knowledge extraction, data integration, and digital library technologies have put the power of distributed information to useful social, scientific, and individual purposes. Data mining tools aid in identifying fraud and abuse in government programs. Data warehouses gather and integrate data from disparate sources, and data management and knowledge discovery tools are used to conduct planning and program evaluation in areas ranging from capital construction, to economic forecasting, to the performance of schools. Technologies such as data intensive computing environments facilitate the use of information from disparate heterogeneous databases. Digital library technologies are emerging to help users find and use information and records regardless of physical format or location.

Today use of these advanced analysis tools varies considerably across agencies and levels of government, and it is too early to tell which applications will be most useful and adaptable. Applications of these technologies are limited today by at least three important considerations: poor or variable data quality, the willingness and ability of organizations to share information across their boundaries, and, when applications involve information about people, threats to personal privacy.

Government, particularly at the Federal level, is already an active partner in the research needed to develop and employ this next generation of data management applications through such projects as the Next Generation Internet, the Partnership for Advanced Computational Infrastructure, and the Digital Libraries initiative.

The exchange of information through a network is not a new phenomenon in government. In the past, telecommunication was accomplished using dedicated point-to-point connections between pairs of agencies, or through secure value-added networks. TCP/IP networks are now replacing these facilities. The use of these Internet protocols facilitates communication between partners because only a single connection need be maintained to communicate with all partners on the network. However, the communication channel must retain properties that duplicate those found in earlier modes of communication: secure and private communication, authentication of messengers, integrity of messages, and stability of the network.

One way to achieve this goal is to create a separate network, closed to all but trusted communicators. This model works for certain types of transactions, but since government agencies often work closely with many other organizations, a more affordable and open solution is needed. At present, there are no commonly implemented models of security architecture that provide a trusted basis for electronic interactions. The array of issues, and the limited choices of technologies and strategies has led to very slow progress in deploying these architectures. In such an environment, it is not surprising that issues of security dominate much of the discussion in government about networking.

An increasing number of important government records are now stored exclusively in electronic media. Many of these records contain multi-media formats, and they are often associated with automated workflow and electronic document repositories. Depending on the circumstances, informal information such as electronic mail messages may be part of an official government record. Few guidelines exist for effectively managing digital public records, yet their numbers grow dramatically every month.

Preservation of electronic records is a particular challenge, as the media, software, and hardware used to create records and maintain them for active use are replaced with new generations every few years. Ironically, while the records of the 17th and 18th centuries remain readable today, our own generation of records is rapidly disappearing due to technological advances. At the same time, government archives are increasingly trying to accommodate the digitization of historical records in order to make these holdings more widely accessible to more users.

Finally, with the increasing availability of information in electronic form, it is becoming easier to use information for purposes beyond the original reason for its collection. Yet most government records systems are created without regard to the needs or preferences of secondary users, whether they are in different units of the same agency, in other organizations, or are future users whose interests come into play long after the records have served their primary purpose. More extensive research into archives and records management theory and practice are needed to resolve these issues.

The standard user interface and the World Wide Web browser, itself a product of NSF-sponsored research, have done much to extend useful computing to every area of our society. The standard interface, commonly based on Microsoft Windows, flattens the learning curve needed for each new application. The Web browser's ease of use and widespread public acceptance have led many agencies to use this technology in direct public contact.

One attractive feature of the WWW is its ability to integrate information and services from separate organizations into a single user presentation. This technique has been used to develop Web sites that serve as a portal to all that a government unit has to offer. Today, most of these sites are limited to a single level of government, and do not represent true integration of services. Instead, they typically provide an extensive table of contents of many agency programs and services. However, many government agencies have begun re-orienting their Web services around the needs of users rather than around their organizational structures.

Further advances in user interfaces are likely to focus both on simplicity and increased power. Digital library technologies, for example, will put the power of multiple databases to more effective uses. Data visualization technologies allow users to manipulate large data sets to get a better understanding of the information they contain. Research into the interaction between people and machines, including speech recognition and 3D modeling, will likely lead to innovations in the way people perceive and use the information environment.

The models and processes for designing and developing large, complex systems have advanced much less than the specific technologies they might employ. While all organizations face this issue, the development of large government information systems face special challenges that lead to an especially risk- prone environment. Typically, a significant number of participants and organizations have a stake in the system. This may be due to the innate complexity of the underlying program or existing systems, to legislative mandates, or because a large number of organizations play a role in the system development process. In addition, because of government funding rules, multi-year projects must usually be developed with a series of single-year budgets. Because they are developed with taxpayer dollars in a public setting, these projects are subject to a high level of external criticism and public scrutiny. In such an environment, it is very difficult to maintain consistent approaches to architecture, data definitions, data collection, information quality, data integration, and overall system functionality.

These complications add time, cost, and complexity to the development life cycle. As a consequence, design and implementation may take years, conflicting directly with the rapidity of technological change. By the time they are completed, the best technologies for the job may well have changed. For example, the recent redesign of the air traffic control system by the Federal Aviation Administration was begun before the widespread commercialization of global positioning systems. Such major technology shifts can cause wholesale changes to system design in the middle of the development process.

Existing software development models such as the waterfall and spiral models do not deal explicitly with these kinds of changes. Prototyping, while very useful in some projects, seems to have less utility in dealing with the complexities of these large systems with their enormous interoperability issues, and long development times.