July 2001 Letter Report to the National Science Foundation
July 9, 2001
Lawrence E. Brandt, Ph.D.
Digital Government Program
Directorate for Computer & Information Science and Engineering
National Science Foundation
4201 Wilson Boulevard Arlington, Virginia 22230
Dear Dr. Brandt:
To obtain input that could help inform planning for future e-government innovation programs, you requested on June 16, 2001, that the Committee on Computing and Communications Research to Enable Better Use of Information Technology in Government provide an interim assessment of the potential role of computer science research in these efforts. The committee was established by the National Research Council’s Computer Science and Telecommunications Board (CSTB) to examine how information technology (IT) research can enable improved and new government services, operations, and interactions with citizens. The study is funded through the National Science Foundation’s (NSF’s) Digital Government program, which supports joint research programs between academic researchers and government agencies.
The first phase of the committee’s study featured workshops examining two illustrative application areas—crisis management and federal
statistics—and concluded with the publication of two summary reports.1 The second phase of the project, which is currently being completed, synthesizes what was learned in the two workshops, information gleaned from other published work examining IT research and e-government, and inputs gathered in the course of two data-gathering meetings and supplemental individual interviews.
This letter is based on the committee’s two previously published workshop reports, as well as other relevant CSTB publications on the evolution and impacts of computer science research and the application of IT in government.2 The committee’s findings based on this work are as follows:
Realizing the potential of e-government that has been demonstrated in early efforts will require addressing implementation issues, resolving shorter-term technology issues, and conducting research on longer-term challenges.
While government can in many cases build on technology developed for the commercial sector, targeted computer science research is needed where government leads demand or has special requirements.
Providing a sound foundation for e-government and other applications of information technology throughout society will depend on ensuring a continuing, broad federal computer science research program.
Challenging computer scientists to address real-world problems in the government sphere can stimulate interactions benefiting researchers, who need access to computer and information artifacts and realistic contexts, and government agencies, which gain expertise and insights that can inform and improve their IT acquisition and management and research results that can be applied in government and elsewhere.
These findings are developed below.
1. REALIZING THE POTENTIAL OF E-GOVERNMENT THAT HAS BEEN DEMONSTRATED IN EARLY EFFORTS WILL REQUIRE ADDRESSING IMPLEMENTATION ISSUES, RESOLVING SHORTER-TERM TECHNOLOGY ISSUES, AND CONDUCTING RESEARCH ON LONGER-TERM CHALLENGES.
The emergence of the Internet and other technologies for electronic commerce has given rise to the concept of “digital government” or “e-government”—the application of information technology (IT) and associated changes in practices to foster a more informed, engaged citizenry and more efficient, accountable government operations. Among the key features envisioned for e-government are increasing access to government information, facilitating transactions with government agencies, making access to information and transactions ubiquitous, better meeting the needs of specific groups of users, increasing people’s participation in government, and meeting expectations for advances in government-unique areas.3 Constituencies include citizens, businesses, nonprofit organizations, and the diverse agencies of federal, state, and local government.
Ideas from early e-government experiments have contributed both to technology development and to the improvement of government’s business practices. Among the most visible enhancements have been aggregated cross-agency portals. These Web sites provide users with access to information and services organized by broad topic and user constituency rather than by specific government departments or agencies, and often are task-oriented.4 Computer-based tax-filing and inquiry-response services provided by multiple agencies are other publicly visible illustrations of positive changes in the way government does its business. Also apparent—in news accounts from across the country— are difficulties experienced by government agencies seeking new capabilities.5
In its examination of IT for crisis management and the collection, analysis, and dissemination of federal statistics, the committee identified a number of challenges to the government’s effective exploitation of IT. These include ensuring the interoperation and integration of diverse systems used by different departments and agencies with multiple stakeholders and a significant legacy base; accommodating existing organizational structures that are difficult to rework to enable exploiting new opportunities afforded by IT; improving trustworthiness, including guarantees of information systems security as well as assurances regarding user privacy and system availability; bridging significant gaps between current practice and best available practices; and meeting specific technology needs related to government missions (see finding 2, below). Each challenge incorporates a mix of implementation, management, short-term technical challenges, and long-term research needs.
Technology transition is an important consideration when research is needed to achieve desired capabilities. The work of researchers may focus primarily on inventing a new capability, but implementation must address the additional challenges of enabling the translation of that invention into a reliable, working product or service.
2. WHILE GOVERNMENT CAN IN MANY CASES BUILD ON TECHNOLOGY DEVELOPED FOR THE COMMERCIAL SECTOR, TARGETED COMPUTER SCIENCE RESEARCH IS NEEDED WHERE GOVERNMENT LEADS DEMAND OR HAS SPECIAL REQUIREMENTS.
Although it can generally build on the technologies and services emerging in the commercial e-business marketplace, government leads demand in some areas. Research in these areas could help government agencies to better accomplish their missions. Moreover, properly managed research in areas of leading demand can stimulate commercial interest and development and thus enable government to move more rapidly away from acquisition of expensive custom systems, whose full life-cycle costs government must bear, to the use of off-the-shelf commercial capabilities. As a result of this process, government now can make use of commercial operating systems with process separation, multimedia databases, and packet-switched networking gear, among other capabilities.
Important Areas of Leading Demand
Ubiquity. Governments must provide services to all citizens—they cannot, in general, opt to serve only the easiest-to-reach customers or participate only in particular market segments. Nor can citizens choose which government they will deal with. Breadth of service encompasses the range of individuals’ physical, cognitive, and language abilities as
well as their education, income, and geographic location. Near-ubiquitous service is provided today primarily through in-person, telephone, and postal mail interactions, but e-mail, the Web, public kiosks, and other IT-based approaches are providing new opportunities to expand access to—and the accessibility of—services. Government, which must comprehensively address the needs of its citizens, has been a leader in exploring how to make technologies accessible, whereas industry has the option of focusing selectively to maximize returns. Research in areas such as human-computer interaction, information retrieval, language translation, and speech recognition and synthesis can help to increase the ubiquity of information and services. Achieving usability and accessibility requires that systems be used, and government applications may provide a range of contexts in which researchers can frame their efforts.6
Trustworthiness. Citizens expect government to provide assurances of security—which includes confidentiality (protection of personal and business information), integrity of information and systems, and availability of information and systems—that are generally stronger than those expected of the private sector. But what is desired may be beyond what the technology and practice can actually offer. Conventional business practice incorporates risk management, in which the costs of implementing security measures are balanced against the consequences of not having them—a calculation that certain levels of exposure can be tolerated for certain applications. Government agencies, however, are expected to adhere to a higher standard—no improper disclosure of personal information contained in statistical data, tax filings, social security records, and the like—even though government is also charged with releasing certain kinds of information, which may be derived from sensitive personal or corporate information that it collects, and making it uniformly available to all.
Trust in public systems is essential for public compliance with government mandates (e.g., paying taxes, completing census forms); equally critical is trust in the safety and reliability of systems on which lives may depend. In military applications, requirements for trustworthiness have led to efforts to promote “high-assurance”7 technologies for critical sys-
tems, and similar requirements apply to IT for transportation and health8 applications, areas in which government is often in partnership with the private sector. Addressing issues related to trustworthiness involves intermingled considerations of policy, organizational behavior and culture, and technology. Computer science research could contribute tools and approaches for facilitating dissemination of information without compromising confidentiality9 as well as for designing and developing systems that provide appropriate, comparatively high overall levels of trustworthiness.10
Coping with Structural and Legal Constraints in Government. Government, like business, has been working to create portals for information and transactions to meet the needs of particular groups of users. Frequently, this requires spanning organizational boundaries. Difficult in any setting, the associated modification of processes and organizational structure in government is often significantly constrained by legal and administrative strictures and may be further complicated by the involvement of state and local governments as well as federal agencies. Also to be considered are information-sharing barriers established legislatively to protect citizen privacy. IT capabilities such as trusted lightweight intermediaries that aggregate for the customer while dispatching components of the aggregate query or transaction to the various government entities involved in responding can help provide a usable interface to citizens despite the array of overlapping, partly interconnected agencies found in government.
Large-scale Systems. Making IT Better discussed a multitude of problems associated with large-scale systems, including delays, unexpected failures, and inflexibility in coping with changing needs, and it observed that notable examples of such systems and problems exist in government.11 Many of the publicly acknowledged failures have occurred in government systems, such as those of the Internal Revenue Service and the Federal Aviation Administration at the federal level and numerous systems at the state and local level. This situation reflects both the large scale of some of those systems and the shortages of IT expertise chronic in government. The problem is growing with expanding use of the
Internet, which has fostered proliferating interconnected systems. And the importance of the problem is also growing as people come to depend more on such systems. Making IT Better points to the need for research to address deep interactions among system components and intersystem dependencies, unintended and unanticipated consequences of system alteration, emergent behaviors in systems with large numbers of components and users, unstable behaviors, properties of federated systems, and other phenomena. In addition to these systems engineering issues, research must address operational engineering issues such as how errors are corrected, how security breaches are detected and corrected, and how backups or other robustness measures are executed. Making IT Better concludes that a research program including case studies of particular systems and methodology research on architecture, techniques, and tools is needed to address the difficult technical (and nontechnical) challenges posed in realizing these systems.
Examples of Government Activities Posing Technical Challenges Outside the Commercial Sector’s Normal Purview
Crisis Management.12 Crises include natural disasters such as hurricanes, earthquakes, floods, or fires and man-made disasters such as industrial accidents, infrastructure failures, or terrorist attacks. Crisis management encompasses crisis response—the actions taken immediately in the wake of a disaster—as well as consequence management, which encompasses the longer-term activities associated with addressing disasters past, present, and future, including planning, preparedness, mitigation, and recovery efforts.
Computer science research can address the critical need for timely, authoritative, and relevant information in crisis response and management efforts. Meeting this need requires (1) a robust, high-performance communications infrastructure; (2) the ability to quickly deploy temporary but robust infrastructure when extensive damage has occurred; (3) the ability to access and compose information and communications systems operated by government at all levels and by nongovernmental organizations, sometimes on an ad hoc basis; (4) support for effective decision making and coordination in the face of uncertainty and stress; (5) tools to help overcome language and other barriers to communicating with citizens; (6) enhanced means of warning populations at risk, especially by providing information targeted to local circumstances faced by individuals or neighborhoods; and (7) means for adapting e-commerce
technology and practices to accommodate handling the exceptions inherent in crisis situations.
Collection, Processing and Analysis, and Dissemination of Federal Statistics.13 Federal statistics play a key role in a wide range of policy, business, and individual decisions that are based on statistics about population characteristics, the economy, health, education, crime, and other factors. These decisions affect the allocation of federal funding to state and local governments; the apportionment of legislative districts; and adjustments to wages, retirement benefits, and other spending. The necessary information is collected from a large number of respondents and at a level of detail sometimes significantly greater than that queried for in private surveys. The federal statistical agencies are characterized not only by their mission of collecting statistical information but also by their independence and their commitment to a set of principles and practices aimed at ensuring the quality and credibility of the statistical information they provide.14
The tasks of protecting confidentiality and ensuring trustworthiness become more complex when information based on individual records is made public and information records are linked across multiple statistical data sets. Linking, or the use of information integration technologies, enables answering complex aggregate queries, a capability that can be important to domains ranging from crisis response to epidemiology. But linking also enables the use of multiple databases to obtain information that, by policy, should not be divulged because, for example, privacy rules are violated.
In addition to IT confidentiality protection, the federal statistical agencies also lead in demand for IT that can effectively support users with diverse interests and capabilities in retrieving, integrating, and interpreting information drawn from the diverse and heterogeneous sources of data that statistical agencies provide, and for technologies that can support the scope and complexity of data collection efforts associated with surveys and the decennial census. In addition, government has a strong interest in and may lead demand for relating various kinds of data to geographical information. Accordingly, the Digital Government program has supported work in geospatial information systems.
Management of National-Interest Electronic Collections in the Digital Age. The federal government is responsible for three major libraries that play
a leadership role within the national system of libraries and in the international context as well: the Library of Congress, the National Library of Medicine, and the National Agricultural Library. CSTB’s review of the Library of Congress’s information systems strategy15 pointed to significant unsolved research challenges related to the Library’s mission, including how to construct a massive-scale, decentralized, robust collection and how to deal with preservation of “born digital” and digitized materials whose format and contents are dynamic rather than fixed. The federal government faces substantially similar challenges in the task of archiving government records, an effort led by the National Archives and Records Administration, which has itself turned to a computer science research team for help in planning for its long-term digital archiving and preservation needs.16 These and other governmental bodies play a key role in preserving national history and heritage, and computer science research can help with such aspects as the development of suitable systems architectures, automatic indexing, and retrieval of information in multiple media. That opportunity is reinforced by the multiagency support for the Digital Libraries program coordinated by NSF. Related and equally challenging needs arise in the management of massive data sets used for scientific research in areas ranging from genome research to space science.
3. PROVIDING A SOUND FOUNDATION FOR E-GOVERNMENT AND OTHER APPLICATIONS OF INFORMATION TECHNOLOGY THROUGHOUT SOCIETY WILL DEPEND ON ENSURING A CONTINUING, BROAD FEDERAL COMPUTER SCIENCE RESEARCH PROGRAM.
CSTB’s Evolving the High Performance Computing and Communications Initiative to Support the Nation’s Information Infrastructure17 examined the payoff and key lessons learned from federal investment in computing research. The study concluded that the sustained, broad federal investment in IT research has profoundly affected the development of computer technology and ultimately led to many commercially successful
applications. Many IT research programs not only reached their intended mission or science goals but also yielded long-term benefits realized in commercial products, companies, and industries and also in the training of a cadre of capable researchers. Ideas were often transferred to the commercial sector through employment or entrepreneurship. The history of federally funded IT research shows that problems motivated by government needs, such as networking and parallel processing, when suitably framed in a carefully designed research program, proved to have wide commercial application (as evidenced by the Internet, distributed transaction processing, and data mining). Broad goals were often pursued in order to infuse new thinking into the technology supply chain of vendors and technology developers for a mission agency. Examples of the success of this approach include process separation for security in operating systems (DARPA in the 1970s and 1980s), computational science (NSF, DOE, and NASA in the 1980s), and custom very large scale integrated circuit chip design (DARPA in the 1970s and 1980s).
The 1990s saw the evolution of computer science research programs to embrace a broadening set of applications and users. CSTB’s 1997 project examining every-citizen interfaces to the nation’s information infrastructure underscored the opportunity and challenge of developing technology that could be used easily and effectively by all.18 Emphasis has also increased considerably on what CSTB’s Making IT Better termed “social applications.” That study committee observed that emerging demand for “more and better use of IT in ways that affect [people’s] lives more intimately and directly than the early systems did in scientific and back-office business applications” presents “issues with which the traditional IT research community has little experience. Successful work on the social applications of IT will require new computer science and engineering as well as research that is coupled more extensively and effectively to other perspectives—perspectives from other intellectual disciplines and from the people who use the end results, that is, the goods, services, and systems that are deployed.”19
4. CHALLENGING COMPUTER SCIENTISTS TO ADDRESS REAL-WORLD PROBLEMS IN THE GOVERNMENT SPHERE CAN STIMULATE INTERACTIONS BENEFITING RESEARCHERS, WHO NEED ACCESS TO COMPUTER AND INFORMATION ARTIFACTS AND REALISTIC CONTEXTS, AND GOVERNMENT AGENCIES, WHICH GAIN EXPERTISE AND INSIGHTS THAT CAN INFORM AND IMPROVE THEIR IT ACQUISITION AND MANAGEMENT AND RESEARCH RESULTS THAT CAN BE APPLIED IN GOVERNMENT AND ELSEWHERE.
Working on government IT problems offers researchers a number of potential benefits, an important one being access to the artifacts—computer systems, software, and data sets—that are needed in experimental computer science research.20 For example, addressing the challenges of large-scale systems can be facilitated by studying real examples of such systems found in government,21 and research depending on large, diverse information data sets can tap the wealth of public information resources generated or held by government agencies. A government setting gives researchers access to applications with a richness and texture typically lacking in the laboratory, and it diminishes the likelihood of constraints stemming from proprietary considerations that typify work in the private sector—ironically one of the reasons that the private sector can benefit from research on government problems, too, since the results will be publicly available. In Making IT Better, the study committee cited NSF’s Digital Government program as contributing to the kind of researcher-end-user interaction that has become increasingly important to progress in IT, and observed that the government setting also lends itself to the recommended project-based organization of research work.
For government agencies, benefits stemming from research can extend beyond the research results themselves. CSTB’s work on IT modernization efforts at the Department of Defense, Internal Revenue Service, Social Security Administration, and Library of Congress have all pointed to the critical need for technology expertise and leadership.22 While there is no substitute for in-house information technology talent,
those studies have all suggested interactions with researchers as a means of tapping additional technical expertise, especially top-caliber research talent that is unlikely to be obtainable in-house or through the usual contract mechanisms. Thus, e-government provides an unusual opportunity for mutual benefit.
This letter provides an overview of key issues and illustrations of how computer science research can contribute to e-government initiatives. Technical problems arising from government operations have the potential to inspire important computer science research, and a well-managed research program focused on these problems can have an impact not only in helping to achieve e-government capabilities, but also much more broadly. The broadening of the computer science research agenda to encompass user needs and social computing issues argues for researcher involvement with the sorts of real-world problems found in government. In bringing together computer scientists and government agencies to tackle problems of mutual interest, the NSF Digital Government program has begun to demonstrate that potential.
Looking forward, a key question is how to leverage the research-management expertise of organizations such as the NSF, as well as mission agency research organizations (such as DARPA and NASA Ames)— and the results of the research they support—to meet e-government requirements. Making IT Better noted particularly that the leadership role provided by the NSF is essential for preserving the emphasis on long-term research and impact that is needed to engage top computer scientists and meet future needs. Government agencies, like businesses, have real operating needs that often demand short-term fixes, and these are not generally an appropriate target for government-funded research. Such problems, addressed in typical systems integration projects in which the problems and risks are better understood and managed, differ critically from broader, deeper challenges that can be addressed most effectively through collaboration with computer science researchers. Needs of an unprecedented character are most effectively addressed through programs that are carefully managed to engage computer science researchers in advancing the state of the art and participating in the development of solutions that are both realistic and appropriately aggressive with respect to the likely trajectory of emerging technologies.
Moving rapidly to address these unprecedented needs requires a strategy that incorporates not only the development of requirements and the invention of new technologies but also technology transition, organizational culture, and acquisition processes. Carefully designed research efforts can often anticipate these process issues and develop new concepts that are more likely to succeed as they are brought into government systems through acquisition, development and integration, deploy-
ment, and evolution. Finally, finding the right mechanisms for transitioning technology from research to government IT systems is an important practical challenge that must be addressed to reap the full benefits of research. The committee’s final report will expand on these research management issues (and the other points discussed in this letter) in greater detail.
William L. Scherlis,
Committee on Computing and Communications Research to Enable Better Use of Information Technology in Government