Summary and Recommendations
In response to a request from the National Science Foundation (NSF) for advice on planning for e-government innovation programs, the Computer Science and Telecommunications Board (CSTB) convened the Committee on Computing and Communications Research to Enable Better Use of Information Technology in Government. The committee was charged with examining how information technology (IT) research can improve existing government services, operations, and interactions with citizens—as well as create new ones. 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 in 1999 and 2000. 1 The second phase of the project synthesized the results of the two workshops, information gleaned from other published work on IT research and e-government, and material obtained in the course of two data-gathering meetings and supplemental individual interviews. Preliminary results of the second phase were described in a letter report to the National Science Foundation
in 2001 (see Appendix B). In this “Summary and Recommendations” chapter, the committee presents the final results of its study and offers recommendations intended to foster increased and more effective collaboration between IT researchers and government agencies. Chapters 1 through 4 provide supporting discussion and analysis.
Government has done much to leverage IT to deploy e-government services, but much work remains before the vision of e-government can be fully realized (see Finding 1.1, below), whether through adoption of already-existing commercial technologies and practices (Finding 2.1) or through targeted research efforts directed at helping government and its suppliers address challenging new requirements (Finding 2.2). This report identifies research challenges related to e-government and looks at these challenges in the wider context of government IT practice and the transition of innovative information technologies from the laboratory to operational systems.
The conceptualization, design, development, testing, delivery, and support of operational government IT systems for agency end users involve an extensive “supply chain” that includes system integrators, vertical suppliers, major vendors, smaller technology companies, consultants, architects, and researchers. Government end users and IT researchers are, in some sense, at opposite ends of this supply chain and so may seem to be unlikely allies. But both in fact have a natural shared interest in innovation and in meeting future needs (Findings 3.1 and 3.2). Indeed, by working together, they can conceptualize new technology opportunities— and with lower overall risk than if they had been working independently of each other. In collaborative efforts, researchers can gain understanding of the real problems that users face, and so can reduce the risks associated with the process of selecting research problems to address. Working in a government setting also provides researchers with access to artifacts— computer systems, software, and data sets—to support experimentation and study. Government end users gain understanding of emerging technologies and of the feasibility of implementing new operational concepts. In addition, there is mutual reinforcement of the dual roles of government as a farsighted customer working with its suppliers to ensure that future mission needs can be met and as an investor in long-term research with broad socioeconomic impact (Findings 3.1 and 3.3).
The establishment of ties between IT researchers and government agencies does not mean, however, that researchers and end users should somehow short-circuit the supply chain in developing new systems, or that the government will incur risks of early adoption as a result of entering into this kind of collaboration. The consequence, rather, is less risk for the government—because it is better informed in doing its job of defining requirements for its interactions throughout its supply chain. Indeed,
success in this overall model requires care and thought regarding technology transition at all stages of the supply chain (Findings 2.3 and 4.2).
THE OPPORTUNITIES AND CHALLENGES OF E-GOVERNMENT
Finding 1.1. Early efforts have demonstrated the potential of e-government, though much work remains before that potential can be fully realized. Achieving that potential will require addressing a broad set of interrelated issues regarding organization, policy, technology development and transition, systems architecture, and engineering practice. In addressing these issues, the government will benefit from collaboration with the computer science research community.
The emergence of the Internet and other technologies for electronic commerce has led naturally to the development of “digital government” or “e-government” services—the application of information technology, combined with changes in agency practices, to develop more responsive, efficient, and accountable government operations. During the 1990s, federal legislative and executive branch initiatives and state and local government efforts added impetus, fostering experimentation and new programs. Early government Web sites were the pioneering creations of enterprising individuals and groups. Today government agency and organization Web sites get attention from the highest levels of government and the most senior officials. Underscoring the importance of government IT capability, in 2002 the federal government appointed a chief technology officer, whose mission includes coordination of efforts in e-government. E-government—which has diverse constituencies that include citizens and other individuals; businesses; nonprofit organizations; and the many federal, state, and local agencies—is envisioned as providing some of the following key benefits:
More accessible government information;
Faster, smoother transactions with government agencies;
Enhanced ubiquity of access to information and transactions;
Greater effectiveness in meeting the needs of specific groups of users;
Increased participation in government by all people, fostering a more informed and engaged citizenry;
Greater ability to meet expectations for advances in government-unique areas, including challenges in the newly emerging homeland security mission; and
More efficient internal government operations.
Ideas from early experiments in e-government have contributed both to technology development and to the improvement of government practices. A tremendous amount of experience already exists in creating and deploying e-government capabilities. This experience resides in state and local government as well as in a broad span of federal agencies. Increasingly, the best practices emerging from this experience are being adopted by many government agencies.
From the citizen’s perspective, among the most visible enhancements have been new means for the public to access government information, the development of task-oriented cross-agency portals (such as seniors. gov and students.gov), and the recently launched governmentwide portal firstgov.gov. Other visible illustrations include computer-based tax filing and inquiry-response services that are provided by multiple agencies, each accessed through a task-oriented Web portal.
The benefits to business have also been considerable, including broader access to government data, more rapid and efficient interaction with regulatory agencies (such as electronic filing of diverse required reports and comments on proposed regulations), and streamlined acquisition and procurement processes. In addition, administrative burdens associated with initiating and managing small and start-up businesses are being reduced through online resources.
Much work remains if government is to fully realize the ambition of broadening e-government services from information access to transaction support—services that enable citizens, businesses, and other government entities to submit information to, engage in financial transactions with, or otherwise interact with government organizations. Much also remains to be done in enhancing systems used for information management and collaboration among government officials and agencies. In these realms, issues such as confidentiality, data integrity, information management, and usability can present significant obstacles. These issues mix technology and policy, and they can be difficult to address without appropriate collaboration between the technology and policy communities.
Also apparent—from briefings and workshops organized by this committee and in news accounts from across the country—are difficulties experienced by government agencies seeking to deploy new capabilities. The committee identified a number of challenges to the government’s effective exploitation of IT. These challenges, which encompass a mix of research and implementation issues, include the following:
Ensuring the interoperation and integration of diverse systems used by different departments and agencies with multiple stakeholders and a significant legacy base;
Adapting organizational structures so as to maximize their effec-
tiveness in concert with IT-based innovation, which tends to be harder in government than in the private sector;
Improving trustworthiness, including guarantees of information-systems security as well as assurances regarding user privacy and system availability;
Bridging significant gaps between current practices and best-available practices; and
Meeting specific technology needs related to government missions.
Recommendation 1.1. Government should continue to improve its support for transactions with individuals, businesses, and organizations. In doing so, it should emulate, where possible, the commercial trend toward integration of services to improve usability for customers. This means, for example, that government should continue the transition from program- or agency-centered service offerings to user-centered services, which can imply aggregating services from multiple government agencies and potentially from private-sector third parties.
A number of major government online services display a mature realization of information-access technology. Future steps are likely to be much more technically challenging. For example, functionality will expand from emphasizing information access to providing comprehensive support for the multitude of transactions between government and users (citizens, businesses, and other organizations), which significantly lags information access today. Correspondingly, modalities will expand beyond people’s interacting with Web browsers to include routine online access to government information and services through common programming interfaces and access through devices other than desktop computers. As these services evolve, governments will have to figure out how best to leverage IT as part of overall customer service strategies. Part of the process has been adopting a customer service perspective on interactions with citizens, businesses, and others—a trend that began even before the widespread use of the Internet.
To better meet customer needs and serve particular customer groups, such as small businesses or students, businesses are participating in the rapid coevolution of e-business technology and commercial processes. Today, improvement in customer focus is often accomplished by reorganizing corporate structure to aggregate information and transaction services in ways that better serve specific market segments. Although the process can be organizationally difficult and technically challenging, business organizations have often been able to reorganize internally to match the procedural flows appropriate to the new “portal” technologies—for example, by restructuring to bypass intermediaries in relationships. Gov-
ernment has been taking some initial steps in this direction as well, as evidenced by multiple government-information portals (firstgov.gov, fedstats.gov, seniors.gov, and others) that aggregate information from individual agency Web sites and early efforts to aggregate transaction services from multiple agencies. In deciding whether to provide integrated capabilities, an organization must weigh the potential benefits against the technical challenges. Integration is hard and the results tend to be fragile (a change in one of the systems being integrated may cause cascading effects).
As alluded to above, one complication special to the government setting is that integration is fraught with legal and structural challenges. Legal constraints can significantly inhibit rapid, wholesale structural changes in government structures (at both inter- and intraagency levels). For example, legal and regulatory barriers to sharing information have been established to protect citizen privacy. At the same time, the stovepiped appropriations process—in which agencies compete for resources and appropriations that are determined by separate congressional committees—makes cross-agency collaboration to build more integrated systems very difficult. Explicit authorization and resources to undertake cross-agency efforts is an important element of a solution. Also, as noted above, governments differ from businesses in their inherently comprehensive scope. Finally, the desired span of aggregation can include state and local governments as well as federal agencies.
One option for building integrated user-oriented services is the use of intermediary software that draws on data from multiple sources to formulate a set of smaller queries that can be separately dispatched to databases operated by different government entities. For example, an augmentation to fedstats.gov could be devised to support an increasing range of aggregate queries that combine statistical data held by individual agencies.
Intermediaries might be operated commercially as well as by government. Indeed, a competitive market of commercial services that conduct transactions with multiple separate agencies could result in improved access to government services, suggesting that government should explore this opportunity. For example, a commercial service might interact with government transaction services to undertake all of the actions required to register a new small business. An interesting challenge with respect to commercial services is how to provide the expected level of confidentiality and integrity of information, especially as the point of aggregation shifts from government to third parties. Meeting this challenge will require codevelopment of both policy and technical safeguards. For example, some of the data that government collects from individuals
is subject to legislative or regulatory safeguards designed to keep anyone but the individual from combining the information.
At present, much of the thinking about e-government focuses on what can be delivered with today’s technology; this is reflected in the foregoing finding and recommendation. But it is also essential that, in looking ahead, planners contemplate how both technology and user expectations will evolve. This is another element of the rationale for direct collaboration and interaction between technology innovators and end users.
GOVERNMENT INFORMATION TECHNOLOGY PRACTICE
The delivery of new digital government services depends on access to advanced information technology, the right management strategies and processes within organizations adopting the new technology, and effective, ongoing evaluation of new service concepts and goals.
Technologies and Practices
Finding 2.1. Many aspects of e-government development can and should follow the marketplace in selecting technology and associated practices.
Recommendation 2.1. Government should adopt commercial e-commerce technologies and associated practices wherever possible.
Government can make considerable progress toward the e-government vision even without new research results. Businesses (and government) have made significant strides as they have made use of rapidly maturing Internet-based information technologies and responded to the corresponding growth in user expectations. Because the technologies and processes that underlie digital services in government are in most respects similar to those used in e-business, government obviously benefits when it can exploit off-the-shelf components, infrastructure, systems, and successful practices already used in the private sector. In many cases, government can and should follow the example of business with respect to service concepts, technical standards, infrastructural components, and interface design. Recognition of this point is reflected in more than a decade of movement toward increasing use of commercial off-the-shelf technology for government applications. (Of course, for some specific applications and systems, there may be no off-the-shelf option.)
When government systems exploit widely used standards and technologies, government can “ride the curves” of performance growth and enhancement that are characteristic of the broader IT marketplace. Pro-
gram managers must balance these benefits in cost and capability with the added uncertainty that they can experience in the management of systems whose evolution is less under their control compared to that of systems custom-built for specific purposes.
Finding 2.2. Although government should maximize the use of commercial technology as it builds e-government capabilities, government requirements sometimes differ from those found in the commercial world. With respect to requirements for such capabilities as ubiquity (and others listed below), government is a “demand leader.” Targeted research in computer science, coupled with effective technology transition strategies, can contribute significantly to the development of such capabilities.
Although it can generally build on the technologies and services emerging in the e-business marketplace, government leads demand in several areas. Research in these areas can help government agencies better accomplish their missions and also—with the right research-management strategies—stimulate commercial interest and development. Successful stimulation of the commercial marketplace could 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 less costly off-the-shelf commercial capabilities. As the result of such investments in the past, government now can make use, for example, of commercial operating systems with enhanced security (e.g., better process separation), multimedia databases, and packet-switched networking equipment. In the experience of mission agencies such as the U.S. Department of Defense (DOD), National Aeronautics and Space Administration (NASA), and U.S. Department of Energy (DOE), this strategy has been effective not only in areas where government leads demand, but also in areas where aggressive requirements for innovation are shared by government and the commercial sector.
Important areas of leading demand (discussed in more detail in the main text of this report) include these:
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. The breadth of service provided by government must encompass a wide range of individuals’ physical, cognitive, and language abilities as well as their education, income, and geographic location. Near-universal communications access is provided today primarily through telephone and postal mail interactions, but e-mail, the World Wide Web, public information kiosks, and other IT-based
approaches are providing new opportunities to expand the range and accessibility of services. Research in areas such as human-computer interaction, information retrieval, language translation, and speech recognition and synthesis can help increase the ubiquity of such services. Achieving usability and accessibility requires that systems be evaluated in realistic settings, and government applications may provide a range of contexts in which researchers can frame their efforts.
Trustworthiness. Citizens expect government to provide assurances of security—including 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 expected or desired may be beyond what 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, and these calculations sometimes show that certain levels of exposure can be tolerated for certain applications. (Credit card fraud is an example.) 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 citizens’ compliance with government mandates (e.g., paying taxes, completing census forms); equally critical is trust in the safety and reliability of systems on which people’s lives may depend. In military applications, requirements for trustworthiness have led to efforts to promote “high-assurance” technologies for critical systems; similar requirements apply to IT in the transportation and health arenas, in which government is often in partnership with the private sector. In the wake of the September 11, 2001, attacks, there is increased attention to protection of critical infrastructure, including both government and private sector infrastructure. Addressing issues related to trustworthiness involves intermingled considerations of policy, organizational behavior and culture, and technology. Computer science research can contribute tools and approaches for facilitating dissemination of information without compromising confidentiality, as well as for designing and developing systems that provide appropriate, comparatively high overall levels of trustworthiness (as is the intent of the federal interagency High Confidence Software and Systems research program).
Information heterogeneity and semantic interoperability. Like other large entities, governments confront significant organizational and tech-
nical challenges when they seek to use information drawn from multiple sources. Integration is especially difficult in ad hoc situations—such as when federal, state, and local agencies must establish an “instant bureaucracy” to respond to a crisis and support recovery efforts. But integration is difficult in preplanned situations as well—such as when agencies seek to serve customers through aggregation of services (see next bulleted item) or to share information in order to comply with legislative or administrative directives (e.g., mandates to exchange state welfare data or to monitor visa status and U.S. border crossings). Because agency information systems are generally “stovepiped”—that is, they employ different and often incompatible conventions for data format and semantics—such information sharing and data fusion can be difficult. (On the other hand, there can be privacy risks when fusion can be easily accomplished, as is discussed in Chapter 2, in the subsection “Access and Confidentiality.”) Organizational barriers, including a frequent reluctance to share information across agency boundaries, further complicate matters. Technological responses include efforts to employ prevailing commercial standards, develop common standards, and develop improved capabilities to translate information across systems.
Providing software interfaces to services. The provision of application-programming interfaces (APIs) or structured data representations (as in use of the markup language XML) can enable third parties to engage government services on behalf of clients. The Internal Revenue Service (IRS), for example, now allows commercial tax-preparation firms to file returns electronically. Health care researchers, as another example, are actively creating XML representations in order to more readily exchange, combine, and analyze clinical trial data. More generally, appropriate interfaces would also enable citizens and businesses to use software that directly connects with government services. A number of technical challenges exist, though, including protocol design, development of information representations and metadata standards, security and authentication, and development of digital library systems. Government has historically provided data in electronic form, but often the design of the formats makes it infeasible for all but a few specialized contractors to exploit the data readily. Achieving “lightweight” protocols or standards can be difficult, however, because of the often-large number of specific cases that must be handled—for example, those relating to seemingly simple data items such as names and addresses.
Building large-scale systems. Many problems associated with large-scale systems—including delays, unexpected failures, and inflexibility in coping with changing needs—exist in government at all levels. Budgetary constraints also dictate the development of systems that can be sus-
tained at low cost. Continued research activity, involving case studies of particular systems as well as methodology research on architecture, techniques, and tools, is needed to address the difficult challenges, technical and nontechnical, posed in realizing these systems.
Finding 2.3. In building e-government systems and the supporting digital infrastructure, government faces significant nontechnical challenges in the way it acquires IT capabilities.
Best practices in government acquisition already recognize the iterative nature of systems development, due to the simultaneous evolution of technology capability, user requirements, and the environment in which a system is used. For example, the Information Technology Management Reform Act of 1996 (the Clinger-Cohen Act) promotes iterative processes in mainstream IT systems acquisition. Contracts with system integrators that were built around a traditional waterfall acquisition model—in which one first defines the full set of requirements for a system and then builds a system that complies with those specifications—are widely understood to be inadequate except for systems whose requirements are well understood at the outset, and for which there is a corresponding base of experience in design and operation. A characteristic of iterative processes is that life-cycle requirements and/or the engineering design space may not be fully understood at the outset. Iterative processes are also appropriate when unexpected changes in the operating environment are likely.
In electronic business, the frequency of iteration can be very rapid, with consequently high demand on capability for acquisition and engineering management. This fast pace of e-business evolution is shaping expectations for the delivery of government services, creating pressure for an even more rapid pace of evolution in capability and scope. This places even greater demand on government’s ability to manage its relationships with its technology supply chain, including system integrators, vertical suppliers, vendors, start-up technology companies, and researchers. For IT systems, a particular challenge is that there are rapid changes both in underlying technologies and in the expectations of customers. Users seek to avoid getting locked in to rapidly obsolescing technologies and practices—which implies that nontechnical, market-based decision criteria can be significant. Incentives that promote risk sharing between vendors and the government in the development of innovative solutions are also helpful. Complicating the situation is the persistent challenge faced by government in attracting, retaining, and rewarding program managers who can effectively articulate and negotiate the government’s long-term interests in contracts with system integrators and other vendors.
WHY E-GOVERNMENT RESEARCH?
Sustaining the IT Technical Base and Addressing Government Needs
Finding 3.1. By leveraging its dual roles as a user of IT and a long-term investor in IT research, government can increase its awareness of the opportunities afforded by IT advances and influence the development of IT that can help to meet its own needs. This synergy can be stimulated by government through investment throughout the technology supply chain and through the development of relationships with all appropriate segments of the IT research, development, and vendor communities.
Like all major IT customers, government has an interest in the management of its future, which includes maintaining both awareness and influence throughout its IT supply chain. This suggests that there is value for government in developing relationships with diverse segments of the IT research, development, vendor, consultant, and integrator communities.
Government has an additional role, however, as a long-term, patient investor in IT research, particularly with respect to research results that have broad value (Finding 3.4). In this role, the government creates public goods with broad and significant socioeconomic value—such as new algorithms, conventions enabling interoperation, measurement methods, shared testbeds, and the like.
Finding 3.2. A cooperative alliance between researchers and agency end users in defining requirements for new IT-based capabilities has benefits for both groups. Researchers gain a better understanding of the real challenges and obstacles, along with access to data and artifacts that can inform or validate design. Agency users, especially those that lack in-house research capability, gain understanding of emerging and future technologies. By collaborating directly, the two groups can more rapidly converge on requirements that meet real needs and that are technologically achievable—and that may not have been expected by either party in advance of the collaboration.
Interaction between researchers and end users on problems of mutual interest has become increasingly important to progress in many areas of IT. Working on government IT problems in particular offers researchers a number of potential benefits—an important one being access both to subject-matter experts and to associated artifacts that are needed in computer science research. A government setting gives researchers access to
applications with a richness and texture typically lacking in the laboratory, and it may also provide a less constrained environment as it is less subject to proprietary considerations that are often associated with work in the private sector. The National Science Foundation’s Digital Government program, which funds collaborative research between computer science researchers and government agencies, has begun to demonstrate the potential of such collaborations.
For government agencies, benefits stemming from research can extend well beyond the research results themselves. While there is no substitute for in-house IT talent, interactions with researchers are a useful way to tap additional technical expertise, especially top-caliber research talent that is unlikely to be obtainable in-house or through the usual advisory mechanisms. In addition, sustained involvement by a research community in a mission area helps define the metaphors and problems that drive that community. This is evident among researchers who have worked over long periods with DOD, DOE, and NASA, for example. Some of the most significant benefits of government mission research investment can be indirect, because research results can have considerable value beyond the particular application area and because research on hard and long-term problems sometimes leads to unanticipated breakthroughs. Although it is very difficult to do a precise evaluation, these additional indirect benefits have to be considered when assessing the ultimate impact and value of government research programs. Finally, a close collaboration between researchers and agency users can create an environment for stimulating creative solutions to challenges, for example, through rapid iteration of prototypes. (See Chapter 4, “Technology Transition and Program Management: Bridging the Gap Between Research and Impact.”)
Finding 3.3. A sound foundation for e-government and other applications of information technology throughout society depends on a continuing, broad, federal computer science research program.
The historical record shows that broad and sustained federal investment in IT research has nurtured the development of computer technology and the IT industry. Many IT research programs not only reached their intended goals, whether mission or scientific, but also stimulated commercial products, companies, and indeed industries. They also contributed fundamentally to the training of researchers and practitioners, who often transferred technologies to the commercial sector through employment or entrepreneurship. Illustrations can be found, for example, in distributed transaction processing, raster displays, very-large-scale integrated circuit design, and data mining.
Indeed, the rapid rise of e-business in the 1990s—and subsequent e-
government development—was based on a variety of information technologies that were developed incrementally over a period of many decades, eventually reaching a point of maturity, scale, and usability that allowed them to be incorporated directly into the infrastructures of commerce and government. The rapid growth in the past decade was also based on the acceptance of an evolving set of common standards that enabled scaling up, competition, and interoperation. The development of the Internet suite of protocols, along with the establishment of processes for evolving them, is perhaps the most widely recognized example. A significant portion of these technologies and standards resulted directly from ongoing, farsighted government investment by a number of research agencies. Indeed, without this investment, it could be argued that the Internet phenomenon would not have come into existence—it was by no means an inevitable development.
When mission goals are approached strategically, new thinking can be infused into a mission agency’s supply chain of vendors and technology developers, which can yield major dividends for government. For example, products or standards that address government needs can be acquired off the shelf rather than by developing more expensive custom systems. Of course, these products or standards might have emerged without government investment, but quite possibly at the cost of considerable delay.
Finding 3.4. In areas of research whose benefits cannot be fully retained by the investor—areas in which the private sector may not be able to justify investment—government investment can be critical to creating new technology capabilities and opportunities for all participants, including government itself.
Universities and some private sector research laboratories have long led in the development of the core technical underpinnings of e-business, in the education of the IT workforce, and in many aspects of the creation of new technologies and technological commonalities. Building in part on past research results, the U.S. information technology industries have seen tremendous growth. Even with this growth and success, government investment—both from research and mission agencies—in IT research and education continues to enhance the overall competitive position of the U.S. information technology industry. The government role in addressing long-term issues continues to be unique and critical.
This special role of government in IT research investment can be understood in the light of typical industry return-on-investment calculations—which generally (but not always) preclude innovations that are fundamentally long term in character or that have broad and nonspecific
impact. In the language of economics, many of these innovations are nonappropriable—the results of the research diffuse broadly into the technical community and cannot successfully be retained by a single sponsoring organization for its exclusive use. Similarly, particular innovations may be competition-neutral or their impacts may be unpredictable; such innovations, though they could be of broad benefit across the industry, are therefore not in a company’s self-interest to support. (See the discussion of nonappropriability in the subsection “Will Industry Do It?” in Chapter 4.)
Among the products of such research innovation in diverse technical areas are Internet protocol standards, algorithm-analysis techniques, programming-language foundations, and, to a great extent, technologies for interoperation. Government investments leading to such broad benefits not only serve the sponsoring agencies’ needs but also have much wider impact by creating new technologies that can be used by all participants, public and private.
Finding 3.5. There are a number of broad technical areas where government investment in IT research would be particularly likely to have an impact on the creation of more advanced capabilities in e-government.
Enhancements to a number of classes of technologies would greatly facilitate the building of advanced e-government capabilities. These include, for example:
Information infrastructure and e-commerce technologies, which provide the foundation for e-business outside and inside government;
Information management technologies, which permit search and retrieval from the very large volumes of information held by governments and allow integration of diverse sets of heterogeneous information systems;
Middleware, which provides common services and capabilities that “glue” software components together into larger systems;
Human-system interfaces, which provide “every-citizen” usability;
Modeling and simulation, which are important tools underlying government planning and decision making (such as in crisis management); and
Software technologies, which permit construction of more robust, larger-scale, interdependent software systems.
Details of each research area are provided in Chapter 3 of this report.
Recommendation 3.1. The federal government should continue to participate actively in developing a full range of information technologies. At the same time, government should leverage its role as a long-term supporter of IT research to embrace e-government challenges within its broad research programs and to stimulate more targeted technology development to meet particular needs.
E-government research encompasses targeted research to meet particular mission needs, research in technical areas where government leads in demand (Finding 2.2), and broad research of benefit to e-government and elsewhere (Finding 3.5). In each case, government research programs can take steps to stimulate a broader commercial response as well as meet government needs. A historical example of this kind of government stimulus is the development of distribution and security support for operating systems. To meet the demand for large-scale and reliable distributed computing in the Department of Defense, the government invested over a period of many years in underlying technical approaches and engineering-process concepts. The resulting technologies not only address specific government needs but also have formed the basis of mainstream commercial systems.
STIMULATING INFORMATION TECHNOLOGY INNOVATION AND TRANSITION INTO OPERATIONAL SYSTEMS
Within the federal government, some mission agencies—including DOD, NASA, DOE, the National Institutes of Health (NIH), and the Environmental Protection Agency (EPA)—maintain research capability that supports both intra- and extramural research in innovative IT. Government also has a number of mission areas, such as federal statistics and crisis management, that are supported by multiple entities and for which explicit coordination mechanisms have been devised (such as the cooperation between federal statistical agencies and the Office of Management and Budget [OMB], the lead role played by the Federal Emergency Management Agency [FEMA] among federal agencies in crisis management, or the emerging coordination potential of the new Office of Homeland Security). In addition, mechanisms within government coordinate research and development activity in certain key technology areas such as geographic information systems and high-performance computing. Government is taking action to ensure that its long-term needs are addressed in these areas of shared mission, shared reliance on aggressive technology, or both.
The sections above considered basic research and exploratory development that contribute to understanding of core concepts and key design
issues. The next subsections consider additional success factors in this innovation process.
Stimulating Innovation in Systems Building
Finding 4.1. Most government agencies have limited capability (or mandates) to manage IT research programs or to undertake exploratory IT development or aggressive prototyping programs. While some agencies would benefit from developing this capability, for others it may be more effective, with respect to cost and risk, to collaborate with agencies facing similar challenges.
To address their technological needs, most federal, state, and local agencies must choose between acquiring off-the-shelf products and building custom systems. In the former case, costs may be low, but the agency customer has less control over product capability and can exert influence on vendor product plans only in proportion to its (relatively small) share of the market. In the latter case, the agency customer defines the capability but must also bear the total costs of ownership over the lifetime of the system, including explicit management of system evolution and interoperation.
Recommendation 4.1. Consideration should be given to providing specific mechanisms, such as a centrally managed cross-agency IT innovation fund, as incentives to enable government organizations to undertake innovative and risky IT projects.
Agencies, particularly in their acquisition activities, may need explicit incentives and process models in order to stimulate and exploit leading-edge technologies. In many cases, early digital-government capability emerged primarily as a result of imaginative, independently entrepreneurial federal employees taking initiative and accepting risks, but this approach does not scale. Agencies are normally driven by prudence to make conservative decisions in setting requirements and establishing incentives for contractors to deliver within cost and schedule constraints. A centrally managed innovation fund can provide funding resources that help “buy down the risk” for a would-be innovator and a group of experts who can work with contractors to help them develop innovative solutions. Such a fund can also provide opportunities for sharing best practices, experimental integration, scaling up of prototypes, and creation of evaluation testbeds. NSF’s Digital Government grant-making program has demonstrated the concept on a small scale.
Building the Bridge Between Research, Systems Development, and Government Innovation
Finding 4.2. Success in the development of innovative systems depends not only on underlying technical capability but also on research-management strategy. In particular, the transition of new technologies into operational systems remains one of the areas of greatest risk and difficulty.
Even assuming that there are sufficient incentives and rewards for innovation in IT development, huge challenges remain in moving new technology into government operations—from research to prototype development and operational experimentation. This “middle” set of technology development stages is central to successful realization of the vision of digital government, and it can be the most problematic and least controllable portion of the overall process of innovation. There is more to the story than simply understanding the tensions between the risk-averse, operationally focused culture of the acquisition system and the risk-tolerant, future-focused culture of the IT research community.
In particular, managing this set of transitional stages can mean working across the entire supply chain of IT innovation, recognizing that each of the participants—government system integrators, vertical application developers, vendors of major components such as databases and operating systems, technology-focused start-up companies, and university and laboratory researchers, among others—has its own unique set of incentives and interests.
One of the greatest challenges is how to escape from the “specification-first” acquisition model, which one sees even in some prototyping programs that were meant to be exploratory and to undertake risks. Effective innovation in systems development requires simultaneous commitment to developing new technology and to defining new concepts of operation. In this model, iteration (in the sense of the spiral model from software engineering, in which a sequence of prototypes of increasing functionality is created) is undertaken both with respect to definition of operational concept and development of the technology. Ironically, adoption of the spiral model may increase overall programmatic risk (for development projects) unless it is carefully managed—which is one reason that sequential models are often adopted even when incentives exist to use iterative approaches.
Regardless, exploratory IT development and prototyping programs are an important mechanism for better exploiting new IT. Such applied research allows agencies to evaluate potential new capabilities and to assess their implications for operational concepts.
Recommendation 4.2. The federal government should develop more effective means for undertaking multiagency collaborative efforts that support aggressive prototyping, technology evaluation, and technology transition in support of e-government.
The recent emphasis in government on homeland security, including the development of new science and technology in support of the counterterrorism mission, has fostered new attention to cross-agency activities related to information technology. The appointment during the Bush administration of a federal chief technology officer and development of an associate director for information technology and e-government within OMB reinforces both the need and the potential for cross-cutting activity. Circumstances have become more propitious than expected at the outset of this study for the kind of cross-agency program development that the committee believes is required and appropriate.
Government agencies that lack robust in-house IT research capabilities should explore research partnerships with agencies that do have them. Several agencies currently manage IT research programs that aim both to conduct IT research in support of agency mission requirements and to stimulate IT innovation more broadly. These agencies include NSF, the Defense Advanced Research Projects Agency (DARPA), DOE, and NASA, as well as the National Institute of Standards and Technology (which is charged with addressing measurement, standards, and interoperability challenges). Each of these agencies has in-house capabilities, an IT research culture, and existing relationships with the IT research community that could prove useful in e-government innovation programs. NSF has already established research partnerships on a modest scale through its Digital Government program, which could be expanded to include more government partners. Others might also explore development of such programs where there is a natural alignment of existing missions with common requirements. Given these various options for supporting e-government-related research, it would appear more appropriate for agencies lacking in-house IT capabilities to tap some combination of these existing capabilities (possibly adding some coordination mechanisms) than to create a new entity to manage e-government research.
Recommendation 4.3. To ensure success in the development and transition of new technologies, explicit attention should be given to program-management practices. Research program managers should be cognizant of the complexities of technology development and transition processes, and aware, in particular, of the range of program-management models and strategies that could be employed.
By facilitating collaboration among mission agencies and government research organizations, it may be possible to address areas of shared concern, such as security, more effectively. Cross-agency attention is evident in OMB, the federal Chief Information Officers Council, and in congressional activity, but it has only infrequently translated into long-lived relationships being established to do research or to undertake prototyping studies. Details of various program-management strategies are provided in Chapter 4 of this report.