Click for next page ( 20


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 19
PROMULGATING A VISION A vision without a task is a dream. A task without a vision is drudgery. A task with a vision can change the worm. In addition to an encompassing vision of its parson d'etre and how it performs its mission, the Social Security Administration (SSA) needs a vision of its information systems derived from and supportive of the bigger picture. The agency has accomplished a great deal under its systems modernization projects and should be proud of these achievements. It has also taken an important step in developing a strategic planning process and setting form some Droaa goals and oDJect~ves. The committee assessed an early version of the Agency's Strategic Plan in its Phase I report (National Research Council, 1990). An updated strategic plan was being prepared during this second phase of the committee's study arid was not available for review. Notwithstanding what may be intended for its revised strategic plan, the agency needs a vision of its information systems that will be promulgated within and outside the agency, will drive subsequent plans and projects for information systems, and will support the quantified objectives selected and needed as a basis for decision making. True modernization requires that, rather than just continuing to patch problems, the SSA ensure that changes to its information systems be relevant to the agency's needs for the foreseeable future and be based on a vision of how the SSA might improve the way it does business. An information system that will be adequate to support the SSA as it approaches the twenty-first century will be substantially different from the system that exists today. ~ ,~ ~ ~ ~ ~ ~ ~ . To what extent an adequate information system for the twenty-first century can be reached via evolution or redesign of systems is not clear. But no such assessment is possible without a clearly defined vision to follow in choosing among the many possible alternatives. For example, redesign focused on software is likely to create redundancies in file systems; focused on file restructuring, it might affect multiple software packages. Redesign can also call for new data interfaces, for greater flexibility in using old and new 19

OCR for page 19
20 applications concurrently. The committee foul that high-~el fechr~c~ decision nag at the SSA suffers from lack of a clear vision and goals specific enough to providle needed metrics and objective criteria for )7uuu~e77t0~ For example, a number of high-level goals and objectives are presented in the SSA's interim systems plan (Social Security Administration, 1990)' but they are not adequately coupled to specific information system goals. The absence of metrics in the SSA's goals and objectives makes it impossible to determine to what extent stated goals have been attained now or to what extent-proposed system changes will directly contribute to their attainment. This subject is dealt with in more detail in Chapter 3. The need to establish measurable goals based on a vision for information systems becomes even greater as the SSA's systems stabilize and at a time when the agency has emerged from the crisis-resolution mode of the past (U.S. Department of Health and Human Services, 1982). The most obvious problems in the SSAts systems have been corrected by its past systems modernization efforts, but from now on new automation activities should be justified in terms of long-range benefits. Proper justification and comparison of implementation alternatives require that the agency focus on the future in modeling its information system requirements. Only then can new system-related initiatives be evaluated in terms of their contribution toward achieving a long-range model of the agency's work and realizing the attendant benefits. A VISION FOR THE 1990s AND BEYOND An information system as large as the SSA's has tremendous momentum and takes a long time to change. Even with unlimited resources, an unconstrained charter, and a clear statement of objectives, moving the present system to a desired contemporary one would take years and be fraught with difficulties and risks. Under the far more restrictive conditions within which federal agencies operate, the task of making fundamental improvements to the SSA's information system is a daunting one, to say the least, but it is not hopeless. Continuing and emerging advances in information technology in areas such as software development tools, fourth-generation languages, and interoperable systems will certainly make the task of achieving fundamental improvement much more feasible than it would have been in the past. Planning for these changes is necessary in order to make the inevitable transition as efficient and painless as possible. To put off inevitable technological change will simply cause it to come later at a much higher cost and with greater difficulty. Even though we cannot know the long-term future of information technology, we can predict with confidence the broad directions of technical development during the 1990s.

OCR for page 19
21 Almost everyone would agree that mainframe computers will continue to come down in cost, at a compound rate of about 20 percent per year. Workstation processing power will improve at an even faster rate--perhaps in excess of 30 percent per year. Continued raDid e , ~ , ~ ~ d. . ~ ~ . ~ . e ~ improvements can ne expected In sucn areas OI technology as magnetic and optical storage, communications, and expert systems. I~nprovements in software development tools watt potentially have a profound effect on implementation of the SSA's information system of the future. New fourth-generation languages and other development tools offer the potential for a tenfold improvement in the productivity of system developers. Such productivity improvement would give the SSA, for the first time, the option of desigriing large portions of its information system from the ground up, using contemporary concepts, hardware., and software. The picture of emerging technology is certainly clear enough to provide reliable guidelines for planning the future information systems architecture for the SSA. Considering SSA's needs, the following are general architectural characteristics for the system toward which the agency should strive: A distributed system, with mainframe computers serving as the hub of the system (partimiarly with respect to the storage of shared data files). _ ~ e e . ~ e ~ ~ ~ LDcal ~ntelIlgent workstations to support service agents ano clerical personnel, prodding local service functions and a user-friendly interface to ease training requirements. A telecommunications network that interconnects all parts of the system, allowing an agent to access client data from any terminal in the system. A high degree of computer "intelligence" embedded within the system (probably through the use ot sophisticated expert systems methodologies), permitting increased functionality to serve clients and increase labor ~ .. . et: :~clency. An advanced office information system to support document creation and dissemination, image storage and retrieval, person-to-person cornrnunications, and decision support systems. c' . ~ ~ .. . .. ~ .. .. . Other characteristics such as integrated image processing and audio response with intelligence show promise for SSA over the longer term. Nothing in these statements about the emerging architecture is farfetched or even very controversial. Large, well-managed commercial Grins are moving rapidly toward such systems. Many will certainly have such systems in place by the end of the decade. The SSA can also develop such architecture if it has the leadership, vision, and will to do so. Given the l a system agency's

OCR for page 19
22 importance and size, the SSA should aspire to no less for itself than that expected from competent commercial enterprises. TECHNOLOGY VISION AND INNOVATIVE TRANSITION For technological evolution within the SSA to proceed in a rational manner, the agency must develop a vision of how new technologies will be introduced in concert with an overall vision of the SSA's mission. This can be accomplished only by careful and sustained planning for technological change. A vision of technological change should guide the transition to new systems, but technic y i~lshoz~ not five ~ysfen~s evoked The vision must take into account not only technological opportunities but also, and most importantly, the functional goals, data requirements, service-level objectives, and resource availability (in terms of personnel. equipment. funding. and legislative and administrative constraints) of the agency. New unanticipated requirements, no matter what the source, must be considered realistically in light of system capacity and performance required to support such new requirements. A systems architecture for SSA needs to be highly adaptable to handle increases in the diversity and inhomogeneity of clients, agency personnel, oversight requirements, and even the technological tools themselves. Innovative transition to the new and improved systems needed must be carried out in an evolutionary, planned fashion under the clear guidance of the agency vision. , ~ ~ ~ C~7 C7 - ARCHITECTURES AND PLANNED TRANSITION TO NEW SYSTEMS Once the agency's goals and plans are sufficiently articulated, it will be possible to evaluate alternative architectures for SSA's information systems of the future. For each architecture, alternative transition plans from the current system must be developed. The selection of a target architecture must involve evaluation of the best alternatives. Increased automation of currently manual processes should be included in overall agency goals and articulated plans. The SSA must evolve toward more complete and wide- ranging automation to reduce current manual operations and to minimize, if not eliminate, paper-based processing. The current systems still require excessive manual interventions by clerical staff, with many paper-based operations. Me committee found the current systems to be inflexible, nonmodular, and burdened with many different protocols and interfaces. SSA should consider data requirements, data flow alternatives, and operational needs independently of the current system. The functions performed and data of the present system can then be compared with the needs of the future. Me tradition per

OCR for page 19
23 should prescribe automation occurs manual operations that account for a huge proportion of costs and personnel them By analyzing its information requirements first, SSA can develop a set of well- defined ir~formation-processing modules for each programmatic function. These should be arcked for each proposed system design based on availability and on the cost of well- understood technology and subject to resource allocation constraints. A transition plan and a method for implementing it must be develoner1 ~n(1 ...1 The characteristics of a target system must include modularity, uniformity of data interfaces, and flexibility for both users and those who maintain and update the system. This can be optimally achieved only by carrying out a careful systems analysis based on the metrics and abstract data requirements of SSA's overall task, which, followed by configuration planning, will lay out alternative transitional mixes of parts of the old and new systems that can become intermediate stages in the migration to a new architecture. The transition could be accomplished through a spectrum of possible paths. Two paths that represent opposite ends of such a spectrum are the following: 1. Independent development of a new systems architecture (either through a contractor or an independent in-house group, potentially at another site) and implementation, testing, and temporary parallel operation with the current system prior to complete cutover to the new system; and Gradual module-by-module redesign and addition to the current system, modified as necessary. The first is obviously "cleaner" technologically, but it is also highly visible because the costs must be identified up front. The second alternative is incremental in nature and may require lower levels of budgeting in the short term, but it still requires careful planning and may not ultimately prove to be either technically or economically feasible if the SSA wishes to achieve transition to contemporary technologies in the mid-term (the next decade). The SSA should not try to adopt unproven technology. It must follow in the footsteps of "comparable" industrial systems for its overall systems architecture, which should use a mix of storage media, human interaction technologies, and intelligent systems to achieve the SSA's goals. However, it should do so only after available evidence of successful results in other applications makes it clear that the technology will adequately fulfill SSA's functional requirements. SSA personnel should take advantage of industrial and academic training that focuses on managing technological change, such as the training offered at Carnegie Mellon University's Software Engineering Institute. The trend in industry is toward a client-server architecture and fRic nrrhifP-f~fr~r~ art be quite effective ageu:y-wide if proper ~mplerneruted ~ ~ ~ ~ ~~I,,~ ~ I, Evolution toward a client-server

OCR for page 19
24 architecture may occur gradually without major disruption to either the current system or the services it provides. During the transition to such an architecture, the functions and major processing now done at the NCC with large centralized mainframes would be gradually off-loaded to remote workstations. The central processors would continue to provide centralized access to SSA's databases and would evolve into database servers. The committee expects that UNIX-based intelligent workstations will be a principal component of such an architecture. IMPLEMENTATION OF A NEW SYSTEM Large organizations--in both the government and the private sector alike--have a history of more failures than successes in trying to develop large integrated computer systems. For a system as large and complex as the SSA's, there is always a danger of striving to do too much and failing in the attempt. It is therefore exceedingly important to plan with great care and to exercise suitable caution in implementing a new system. It is important to recognize, however, that new technology does not necessarily equate with high risk. In fact, quite the contrary mav be true. For example. it would be , ~ ~ em virtually 1mposslule to redesign a major segment of the current SSA system using its existing third-generation software development methodologies with COBOL, CICS, and the like. If the agency does not adopt newer technology, it perforce must confine any new developments to relatively small incremental improvements. Such an approach would guarantee continuation of the fragmented present system. It is impossible at this point for the committee to snecifv the exact nature of a ~ ~ ~ r . . ~ .. ~~ . aes~ran~e tuture system tor the bbA or even to describe in much detail the approach that should be taken to move to a new system. This is a job for an internal planning staff within SSA's systems organization that has adequate representation and knowledge of the SSA's business functions. The staff should be very small and very competent. Some outside expertise will be needed to supplement the planning group, but the effort must come primarily from the SSA itself, because continuity and agency knowledge are essential. The planning group must have sufficient clout to be taken seriously. The committee did not find that such a group currently exists, and the SSA may be hard pressed to staff it. It is essential that managers throughout the SSA, at all levels, participate as partners in the effort. The group should be willing to think anew about how the agency can best carry out its mission, rather than merely automating the present procedures that evolved over five decades in an entirely different environment. A goal of the planning group should be to develop a plan to move from the present system to the new system of the future. Such a system must necessarily be developed over a number of years in a deliberate and risk-averse fashion. The committee recognizes that

OCR for page 19
25 adopting low-risk solutions will constrain the agengy's choice of technologies to those that are proven and evolutionary. It is prudent for the SSA to be technically conservative because of the national importance of its programs and the need to keep them running smoothly and continuously. But SSA's sheer size, momentum, and "sunk" investment make any change difficult. With limited technical resources that may be willing and able to lead and implement new approaches, a risk-averse strategy becomes necessary, but it should still allow for significant improvement. The planning group should resist the temptation to spell out too many details about aspects of the system that cannot be implemented for a number of years. Rather, its focus should be on laying out a broad architecture for a future system, combined with a realistic statement of the detailed short-term steps needed to move toward the long-term goal. Perhaps the single most critical question about any plan for SSA's future information system is whether excellence in informations system design is a realistic aspiration for an agency hemmed in by complexity and constraints. The committee is confident that the architectural characteristics envisioned above ("A Vision for the 199Os and Beyond") are both highly desirable and quite feasible if done with reasonable competence. Needed is a commitment to excellence. REFERENCES Social Security Administration. 1990. Social Security Administration Interim Systems Plan for Fiscal Years 1990-1995. Report prepared for the Senate Appropriations Committee. Washington, D.C.: Social Security Administration. National Research Council. 1990. Systems Modernization and the Strategic Plans of the Social Security Administration. Washington, D.C.: National Academy Press. U.S. Department of Health and Human Services. February 1982. Systems Modernization Plan--From Survival to State of the Art. Washington, D.C.: Social Security Administration.

OCR for page 19