You are on page 1of 149

The Social Security Administration and

Information Technology

October 1986

NTIS order #PB87-136834


Recommended Citation:
U.S. Congress, Office of Technology Assessment, The Social Security Administra-
tion and Information Technology-Special Report, OTA-CIT-311 (Washington, DC:
U.S. Government Printing Office, October 1986).

Library of Congress Catalog Card Number 86-600567

For sale by the Superintendent of Documents


U.S. Government Printing Office, Washington, DC 20402
— —

Foreword

The Social Security Administration in 1982 announced its Systems Moderni-


zation Plan (SMP), designed to restructure and extensively upgrade its data-
handling systems. The agency told Congress that, without this major upgrading,
there might be serious disruption of its services, which are essential to the wel-
fare of millions of Americans. The SMP was one of the most expensive civilian
information projects ever undertaken; it has since become a ‘‘rolling’ 5-year plan
with projected costs currently estimated at nearly $1 billion.
The disruption of services that the Social Security Administration feared in
1982 has been averted, but the SMP and its implementation have been the sub-
ject of continuing controversy and criticism within the Administration and in con-
gressional oversight committees. The Social Security Administration has scheduled
major procurements in fiscal year 1987 that are central to implementation of SMP.
In this special report, OTA examines the objectives and technical strategies
embodied in the SMP and the progress that the Social Security Administration
has made toward its implementation. The report calls attention to some general
problems faced by both SSA and other Federal agencies that are increasingly de-
pendent on communications and information technology in carrying out their
missions.
OTA appreciates the participation of the advisory panelists, workshop par-
ticipants, Federal agency officials, and interested citizens, without whose help
this report would not have been possible. The report itself, however, is the sole
responsibility of OTA, not of those who advised and assisted us in its preparation.

-
JOHN H. GIBBONS
Director

//’/
Social Security Workshop: March 5, 1986

Hale Champion, Chairman


Executive Dean, John F. Kennedy School of Government, Harvard University
Eileen Appelbaum Ken Laudon
Professor of Economics Educational Fund for Individual Rights
Temple University Paul Light
Frank DeGeorge Director of Studies
Deputy Inspector General National Academy of Public Administration
Department of Commerce Raymond N. Palombo
Renny DiPentima Director, Human Resources Planning
Deputy Associate Commissioner J.C. Penney, Inc.
Office of Systems Integration Nancy Shor
Social Security Administration National Organization of Social Security
Professor John Gessford Claimants’ Representatives
Information Sciences Arthur Simermeyer
Claremont Graduate School Acting Deputy Commissioner for Systems
Mark Hardy Social Security Administration
Booz-Allen & Hamilton, Inc. James Strassberger
John D. Harris Steward, Local 1923
Special Assistant to the President American Federation of Government
American Federation of Government Employees
Employees Eileen Sweeney
Ed Howard National Senior Citizens Law Center
Villers Foundation Alan Westin
Jed Kee President
Professor of Public Administration Educational Fund for Individual Rights
The George Washington University

February 12, 1986 Meeting

Renny DiPentima Peter Poetker


Deputy Associate Commissioner Program Manager, Software Engineering
Office of Systems Integration Social Security Administration
Social Security Administration David Schwartz
Richard W. Eckert Social Security Administration
Director, Division of Data Administration Arthur Simermeyer
Social Security Administration Acting Deputy Commissioner for Systems
Ken Laudon Social Security Administration
Educational Fund for Individual Rights Alan Westin
Roger McDonnell President
Assistant to the Commissioner for Systems Educational Fund for Individual Rights
Social Security Administration
OTA Project Staff: The Social Security Administration
and ‘Information Technology—Special Report

John Andelin, Assistant Director, OTA


Science, Information, and Natural Resources Division

Fred W. Weingarten, Program Manager


Communication and Information Technologies Program

Vary T. Coates, Project Director


Ben Amick, Analyst

Administrative Staff
Liz Emanuel Sandra Serbinoff Audrey Newman

Contractors
Ken Laudon
Educational Fund for Individual Rights
Alan Westin
President, Educational Fund for Individual Rights
Reviewers

Arthur G. Anderson David H. Flaherty


IBM Corp. (retired) Professor of History and Law
University of Western Ontario
Eileen Appelbaum
Professor of Economics J. William Gadsby
Temple University Associate Director, Human Resources Division
General Accounting Office
Christie Bentzen
General Accounting Office John Gessford
Professor of Information Sciences
Jerry J. Berman, Esq. Claremont Graduate School
Legislative Counsel
American Civil Liberties Union Carl Hammer
Consultant
R. J. Bogumil
Society on Social Implications of Technology Mark Hardy
Institute of Electrical and Electronics Booz-Allen & Hamilton, Inc.
Engineers John D. Harris
James W. Carey Special Assistant to the President
Dean, College of Communications American Federation of Government
University of Illinois Employees
Hale Champion Robert D. Harris
Executive Dean Deputy Assistant Director for Budget Analysis
John F. Kennedy School of Government Congressional Budget Office
Harvard University Starr Roxanne Hiltz
Robert L. Chartrand Professor of Sociology
Senior Specialist Upsala College
Congressional Research Service Ed Howard
Tony Cicco Villers Foundation
Group Director Kenneth W. Hunter
IMTEC D, General Accounting Office Senior Associate Director for Program
Melvin Day Information
Vice President General Accounting Office
Research Publications Jed Kee
Frank DeGeorge Professor of Public Administration
Deputy Inspector General The George Washington University
Department of Commerce Richard P. Kusserow
Joseph Duncan Inspector General
Corporate Economist Department of Health and Human Services
The Dunn & Bradstreet Corp. John C. Lautsch
William Dutton Chairman, Computer Law Division
Professor, Annenberg School of American Bar Association
Communications
University of Southern California

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the reviewers. The reviewers
do not, however, necessarily approve, disapprove, or endorse this report, OTA assumes full responsibility for the report
and the accuracy of its contents.

vi
Theodore J. Lowi Roger Sperry
John I.. Senior Professor of American Governmental Affairs Committee
Institutions U. S. Senate
Cornell University Terril J. Steichen
Edward F. Madigan President
Office of State Finance New Perspectives Group, Ltd.
State of Oklahoma Paul Steitz
Felix J. Majka Committee on Ways and Means
Assistant Inspector General for Audit Subcommittee on Social Security
Department of Health and Human Services U.S. House of Representatives
Marilyn Gel] Mason George Trubow
Director Director
Atlanta Public Library Center for Information Technology and
Privacy Law
S. Anthony McCann The John Marshall School
Assistant Secretary for Management and
Budget Bernie Unger
Department of Health and Human Services Group Leader, Human Resources Division
General Accounting Office
John McNicholas
Chief, Information Policy Branch Susan Welch
Office of Management and Budget Professor and Chairperson of Political Science
Department
Vince Muzzuca University of Nebraska
Director, EDP Audit Division
Office of the Inspector General Langdon Winner
Department of Health and Human Services Associate Professor of Political Science
Department of Science and Technology Studies
Joe Quasney Rensselaer Polytechnic Institute
Associate Director
IMTEC D, General Accounting Office
Jim Rife
Committee on Government Operations
U.S. House of Representatives

OTA Reviewers
Earl Dowdy
Fred Wood

vii
Contents
Page

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Part I: An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter l: SSA and Information Technology:
Conclusions, Issues, and Options . . . . . . . . . . . . . . . . . . . . . . . . 7

Part II: The Systems Modernization Plan and Its Potential Effects . . . . . 29
Chapter 2: The Status of Systems Modernization at SSA, 1986 . . . . . . . . 33
Chapter 3: Systems Modernization and Related Issues, 1986-90 . . . . . . . . 57
Chapter 4: The Future: The Necessity of Long-Range
Technology Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

Part III: The Case History of Information Technology Management


at the Social Security Administration . . . . . . . . . . . . . . . . . . . . . . . . 89
Chapter 5: Years of Service and Satisfaction, 1935-71 . . . . . . . . . . . . . . . . . 93
Chapter 6: Deepening Problems, 1972-81 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Chapter 7: The Beginning of the Systems Modernization Plan, 1982 . . . . 119
Chapter 8: The Oversight of SMP, 1982-86 . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Glossary of Technical, Institutional, and Legislative Terms
Used in This Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

..
Vlll
Summary
In 1982, the Social Security Administration (SSA) announced a 5-year plan to modernize its information
systems. The Systems Modernization Plan (SMP) was a multimillion dollar response to serious problems
that had developed during the 1970s, and that repeatedly threatened to disrupt SSA’s services delivery
operations. Congressional oversight had identified many of these problems, but had not fully revealed
their persistent, deep-seated, and cumulative effect on the agency’s ability to respond efficiently to con-
gressional mandates and priorities.
This special report explores the factors that led to SSA’s information systems problems. It concludes that
other Federal agencies are vulnerable to similar problems as they automate and modernize data-handling
operations. It also concludes that effective oversight and monitoring of agencies dependent on advanced
information systems is becoming more difficult, as technological decisionmaking and management in-
creasingly requires specialized knowledge.

Issues for Congress


The basic strategy of SSA’s SMP as set out in 1982 is reasonable and defendable in the sense that
it is consistent with accepted systems engineering practices. Some experts argue that alternative strategies
should have been adopted, such as regional decentralization of data processing. However, whether or not
the original decisions were the best ones, the alternative strategies also have disadvantages and risks;
they cannot be shown to offer stronger guarantees of success than does SMP. Achieving SMP’s objectives
now depends on SSA's technical competence, on the quality of its management as it implements the
SMP, and on certain factors outside of the agency’s control, including Administration policy and directives.
SSA has made significant progress toward achieving the goals of the plan in many areas, especially hard-
ware acquisition. In other areas, SSA has fallen behind. The report identifies a series of serious unresolved
problems in software engineering, database architecture, and database integration. Even though the strat-
egy is sound, these problems cast doubt on SSA’s understanding of the nature of some technical questions,
and its commitment to the quality of management that is essential to success of SMP.
This report also raises strong questions about the reliability and completeness of the information that
SSA has provided to Congress and to the public about its progress in SMP implementation. Congress
may therefore want to intensify its monitoring and oversight of all aspects of implementation, to make
sure that the goals of systems modernization, efficient management, and improved services delivery that
Congress accepted in funding the SMP, are achieved. Alternatively, Congress may want to accept the judg-
ment of SSA and its executive branch monitoring agencies as to whether current progress toward solving
major implementation problems is acceptable, and implementation should proceed through the planned
steps and scheduled milestones. This choice has a bearing on two immediate issues: a large SMP procure-
ment in fiscal year 1987, and the proposal to make SSA an independent agency.
The SMP schedule calls for procurements to automate claims filing in field offices in early 1987. If
implementation of SMP were proceeding satisfactorily, the timing of these acquisitions probably would
not be controversial. The benefits of a modernized claims process, with improved service for the growing
population of beneficiaries and greater productivity for SSA, would outweigh the cost of temporarily
underused computer capacity while other procedures are automated. Unfortunately, there is conflicting
evidence about the reality and pace of this progress, and whether SSA fully understands the nature of
the technical hurdles that must still be surmounted. SSA has not convincingly demonstrated that it
is on the way to solving its software and database problems, and some experts have questioned the advis-
ability of proceeding with this or other procurements until such demonstration is made.
The decision about the procurement should be made in the overall context of the SMP and the desirable
long-term goal, adopted in 1982, to modernize the SSA system. Analysis of costs and benefits of any
procurement should include the effects of its timing on all aspects of SMP implementation. Proceeding
with the procurement risks incurring the costs of unused communication and computer capacity while soft-
ware is developed for automating additional field office services. On the other hand, procurement delay
risks losing possible productivity gains in claims processing, as well as possible degradation in service,
should staff reductions take place. It appears that the claims process could be automated, with considerable
benefit to clients and the agency, while the software work continues. The database problems must be solved,
regardless of the procurement decision. The immediate procurement issue should not deflect attention
from, nor should it determine the course of actions related to the larger systems management problems in SSA.

1
Congress may wish to consider giving SSA a cautious “go-ahead” in conjunction with increased moni-
toring and close oversight, and insisting that broad corrective actions be taken over the next year to
strengthen SSA's management control and its systems development capability. A second option is to
stop SMP procurement and contracting, and to take advantage of this interval for congressional re-
evaluation of the goal of providing SSA with state-of-the-art technological systems, and its feasibility,
in the light of current priorities.1 In this situation, however, some potential productivity gains may be
foreclosed for the near future, and the commitment to and impetus of systems modernization may fall
into disarray, and be difficult to recapture. A third option is to require each step in implementation to
be justified in isolation; however, this involves Congress in a difficult process of micromanagement that
may lose sight of the overall objectives of the SMP effort.
Some potential risks to the timely success of systems modernization are outside of SSA’s control. The
current pressure to realize productivity benefits by severe work force reductions before the modernized
systems are ready, could lead to a return of the problems of 1978 to 1982, and thus discredit SMP and
whatever progress has been made. Renewed instability of leadership and organizational restructuring
could also delay progress and disrupt SSA’s still fragile attempts to institutionalize advanced planning
and to improve labor-management relations, both of which are essential to the success of systems mod-
ernization. These factors are as important to SMP success as is the immediate procurement issue.
The House of Representatives has voted for independent status for SSA; in part to reduce some of these
risks. While this might buffer the systems modernization effort against some external pressures, it would
not necessarily resolve the difficult questions about the most effective modes of congressional oversight
and monitoring.

The Long-Range Implications of SSA Systems Modernization


SSA’s ability to respond to congressionally mandated changes in benefits programs has already been
improved, and will be strengthened further when it succeeds with software development and improvement.
However, continuing automation of both operations and management functions-both in SSA and other
Federal agencies-is likely to make effective congressional oversight more demanding and difficult. Infor-
mation technologies can make it easier to select, manipulate, or obscure critical performance data, and
the costs of failures in implementing and managing advanced systems may make it more tempting to
do so. Evaluating management decisions and performance related to advanced technological systems
increasingly requires specialized knowledge.
As the automation of SSA’s operations proceeds, labor-management relations will inevitably be stressed.
Displacement of some workers will occur, and there will be increasing need for retraining and relocation
of other workers, and for reconsideration of personnel recruitment, retention, advancement, and compen-
sation policies.
State-of-the-art information technology, by facilitating increased use of computer-matching techniques
and the sharing of data about individuals with other organizations, and broadening access to data, will
exacerbate present concerns about the privacy and security of social security information.
The findings of the report about the source of SSA’s problems, why they were not resolved, and why
their seriousness was not fully revealed by oversight and monitoring mechanisms before they became
critical, have significant implications for other Federal agencies and for Congress. Many of the problems
were found to stem from factors and forces that may affect all Federal agencies that become dependent
on advanced information systems.
The report describes some mechanisms or actions that have been suggested for facilitating congres-
sional oversight of SSA’s systems modernization. They include independent agency status, broader and
more frequent studies by congressional support organizations, integration of SSA oversight in one com-
mittee of each House, a stronger mandate for subcommittees overseeing government information tech-
nology decisions, and an external blue-ribbon advisory panel on government information technology de-
cisions. The most promising of these options are those that could be generalized to apply to other Federal
agencies with similar problems now or in the future.
*This option was recommended by the General Accounting Office in its recent report on SSA automated data processing acquisi-
tions dated August 1986.

2
Part I

An Overview

Chapter 1 summarizes the report of information technology management in


the Social Security Administration (SSA), highlighting the conclusions of the study,
and the issues related to SSA’S Systems Modernization Plan and its implementa-
tion from 1982 through 1986 and beyond.
Chapter 1

SSA and Information Technology:


Conclusions, Issues, and Options
Contents
Page
Summary of Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Why SSA Faced Potential Disruption of Service in 1982 . . . . . . . . . . . . . . 11
Brief History of SSA and Information Technology . . . . . . . . . . . . . . . . . 11
Underlying Factors in SSA’s Critical Systems Deficiency . . . . . . . . . . . . 13
Why Monitoring and Oversight Failed To Correct SSA’s Problems . . . . . 17
The Basic Strategy of SSA's Systems Modernization Plan Is Sound . . . . 19
Conflicting Evidence on Progress in Systems Modernization . . . . . . . . . . . 20
The Effects of Systems Modernization: Closely Related Issues . . . . . . . . . 24
SSA Responsiveness to Congressional Mandates . . . . . . . . . . . . . . . . . . . 25
Independent Status for SSA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Privatization of or Contracting out Government Operations . . . . . . . . . 25
The Work Force and Labor-Management Relations . . . . . . . . . . . . . . . . . 26
Privacy and Security Concerns . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . 26
Options for Facilitating Congressional Oversight of Systems
Modernization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... , . . . . 27

Table
Table No. Page
1. Major Reported Accomplishments of the SSA’s
Systems Modernization Plan in Its First 4 Years and
Future Milestones by Specific Program Areas . . . . . . . . . . . . . . . . . . . . . . 22

Figure
Figure No. Page
1. Cumulative Projected Total Costs for the Systems Modernization Plan
for 1982, 1985, and 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Chapter 1

SSA and Information Technology:


Conclusions, Issues, and Options

In 1982, the Social Security Administration to prevent emerging problems from be-
(SSA) began an effort to thoroughly modern- coming critical?
ize its data-handling systems, in order to “. . . How sound are the basic strategies of
avoid potential disruption of service through SSA’s Systems h’modernization Plan?
immediate improvement of critical system defi- Is there evidence that the progress on
ciencies, to restore integrity and public confi- SMP to date is reasonable}’ satisfactory,
dence, to improve productivity, and to close and that it will achieve its objectives?
the technology gap”l that had developed How will SSA’s systems modernization
over the last decade. affect, or be affected by, several issues now
before Congress such as the movement to
Projected to cost $500 million and take 5
give SSA independent agency status, the
years to carry SSA “from survival to state of
effort to reduce SSA budget and the size
the art, the SSA Systems Modernization Plan
of its work force, or the possibility}’ of
(SMP) was one of the most expensive civilian changes in social security programs, ben-
information projects ever undertaken. It has
efits, or eligibility determinations’?
since become a ‘‘rolling’ 5-year plan with pro-
jected costs currently estimated at nearly $1 In addition, the report seeks to explore sev-
billion. See figure 1. eral broader questions that are addressed
throughout the report:
This report of SSA and its SMP addresses
the following questions: ● Are the problems that SSA had, and is
having, generic problems that other Fed-
● Why did SSA face ‘‘potential disruption
eral agencies are likely to face in manag-
of service’ through ‘‘critical system defi-
ing information technology’?
ciencies” in 1982? ● What can be learned from SSA experi-
● Why did continual congressional over-
ence that can be helpful in future adop-
sight and executive branch monitoring fail tion, use, and management of advanced
systems for government operations?
● Will advanced information systems facili-
‘ This is an abbreviated quote from SS/4 .S~.ster12s ,JZocfern- tate, or make more difficult, congressional
ization I)lan: From .5’ur~’i~’al to State of the .~lrt, ch. 1, Februar~r oversight of executive agencies?
1982, pp. 1 W 19. Material in this chapter describing the Sjs-
tems Modernization Plan and not otherwise attributed is drawn ● Are there feasible strategies for making
from SSA documents and discussions with SSA officials. oversight more effective?

SUMMARY OF CONCLUSIONS
The necessity of modernizing SSA’s infor- of certain critical choices that SSA made. There
mation technology systems was beyond dis- is serious doubt about how much progress has
pute, yet in the 4 years that SMP has been been made in systems modernization; and
underway, it has never been free of criticism about whether SSA fully understands and is
and controversy. There has been widespread prepared to cope with some persistent prob-
questioning of the basic strategy of SMP and lems, or has dealt, in an open and frank way,

7
8

Figure 1.— Cumulative Projected Total Costs for the Systems Modernization Plan (SMP) for 1982,1985, and 1986
1,000 f
950 0 - - - -

900 ● — C

850 -
800

200
150
100
50
0 L
81 82 83 84 85 86 87 88 89 90 91
Year of SMP
SOURCES. U S Department of Health and Human Services Soc(al Secur(t y Adrm mstrdtlor] S} stems r$fcidero~:a(~o,? Plan from Surv/ka/ tL] Stare (of (he A rf Fehruary
1982, U S Department of Health and Human Serwces Soc(al Securlfy Adrnl m strat Ion Systems ?vfode,nl:~ffor~ f’/dn f98.5 Upd~fe PutIl I( at Ion No 40004
January 1985, U S Department of Health and Human Services Soc Ial Securtv Admlnlst~dtlon SSA S\ stem Modernl.7dt/orl P/d’ 1:786 Lo,Iq Raoqe SI,a/eqIL-
P/afl, Publlcatlon No 40-004, October 1985

with its oversight organizations in terms of tomate the operations of a large data-handling
these problems and SSA’s plans to cope with operation; there are several alternative ap-
them. proaches, each of which has some advantages
and disadvantages. SSA's plan must be evalu-
Some of the doubts about and criticisms of
ated in that context.
SSA’s systems modernization plan can prob-
ably not be answered definitively. Both in the The basic strategy of SSA Systems Mod-
public sector and in the private sector, large ernization Plan is in accord with accepted engi-
organizations with complex data operations neering practice; it is reasonable and defend-
are still struggling to find the best ways to use ble. To reverse this strategy four years into
advanced technology to maximum advantage. the effort probably would be wasteful of invest-
While many lessons have been learned from ments already made, and alternative broad
experience, experts can be found to attack or strategies suggested by critics would not nec-
defend any strategy with plausible reasoning essarily provide any stronger guarantee of
and with equal vehemence. There is no clear success. This conclusion does not however nec-
and indisputable ‘right or ‘best way to au- essarily imply that SSA performance in im-
9

plementing the plan is satisfactory or that the a change in SSA top leadership. Whether this
objectives of the plan will be achieved. At a will strengthen or disrupt these promising de-
minimum, increased monitoring and close over- velopments, remains to be seen. A thorough
sight are necessary if SSA is allowed to pro- reorganization of the agency could, for exam-
ceed according to its current schedule; but it ple, interrupt or destroy these still tenuous
is essential that decisions about specific SMP management improvements.
procurements be made in the context of the
broad plan. Other impending events that will affect the
likelihood of success in systems modernization
The success of SMP in meeting its reason- are largely out of SSA’s control. For example,
able objectives will depend not only on the the attempt to reap the anticipated benefits
basic soundness of the plan, but also on: of increased productivity, in the form of severe
1. whether SSA has the technical compe- work force reductions, before the systems are
tence to implement the plan, in place to provide these benefits could pose
2. whether SSA exercises good management significant risks to continued progress. The
in carrying out the plan, and Administration is insisting on immediate re-
3. whether certain conditions are obtained duction of SSA’s work force. If SMP’s prom-
at SSA at this critical stage in its imple- ised increases in productivity are not yet in
mentation. place to support such reductions (as they prob-
ably are not), any subsequent temporary ex-
There are some serious questions about pansion of the volume of work (e.g., implement-
whether the progress toward implementation ing a legislatively mandated change in benefits
of the plan can be considered satisfactory. In or coverage) could again lead to huge backlogs,
some areas, the implementation of SMP is far which could discredit SMP before it is com-
behind schedule, and although SSA makes pleted, and would almost surely result in a
strong claims to have solved, or to be well on breakdown of a promising but still embryonic
the way to solving, serious technical problems joint agreement designed to reduce crippling
in achieving its goals, it has not been able to labor-management tensions.
demonstrate to independent experts that this
is the case. There are disturbing signals that SSA’s problems in data-handling built up
SSA either may not understand the serious- slowly, but became evident when the agency
ness of these problems, is relying on "solu- was several times able to respond to congres-
tions” that are likely not to work, or is cover- sional mandates only with extraordinary ef-
ing up the seriousness and persistence of the forts and with long-lasting, deleterious after-
problems. effects, For at least a decade, SSA's frequently
changing leadership was unable to solve
SSA appears to have just begun to develop chronic organizational problems, and the
some management procedures and mecha- agency failid to communicate effectively these
nisms to improve its capability at advanced problems to Congress throughout several Ad-
systems development, for example, to remove ministrations.
long-standing friction between the systems de-
velopment and operations components of the Many of the problems that in the decade be-
organization, to improve the recruitment and fore 1982 drove SSA to the brink of crisis were
training of systems personnel, and to use an common to many large organizations with sim-
innovative and constructive approach to labor- ilarly complex operations and rapidly growing
management relations. These management im- workloads. In the case of SSA, however, they
provements, if they are developing as SSA offi- persisted and were exacerbated almost beyond
cials describe, are still highly fragile. As this the point of solution by, on the one hand, cer-
report is being completed, there has just been tain characteristics of the organization and
10

failures of its management, and on the other appears to have contributed to this situation
hand, by external constraints and pressures, and still complicates attempts to understand
such as frequent changes in organization and SSA’s problems. This defensiveness was ex-
in its top leadership. treme but is not unique to SSA. The highly
technical decisions that must be made with re-
Government agencies are properly and nec-
gard to advanced computer systems pose spe-
essarily subject to constraints, accountability
cial difficulties for most congressional over-
requirements, and oversight that do not affect
sight committees. Special mechanisms may be
private sector organizations. These greatly
needed to facilitate oversight of major tech-
complicate agencies’ decisions about technol-
nological decisions by Federal agencies.
ogy and forbid some routes to modernization
that private sector organizations have found The issue of restoring SSA to the status of
productive. In addition, governmental agen- an independent agency is now before Congress;
cies are insulated against suffering the imme- the House of Representatives has already
diate marketplace penalties for bad decisions; passed a bill (H. R. 5050) to this effect. One of
this allows them at times to persevere in faulty the reasons that has been put forward in ad-
management practices and to ignore or con- vocating this action is to improve SSA’s re-
ceal emerging problems until they become in- sponsiveness to congressional policies related
tractable. The report indicates that in some to systems modernization. Whether or not this
aspects of systems modernization, SSA did action is advisable on other grounds, independ-
this persistently in the 1970s and may well be ent status is unlikely to solve either SSA sys-
doing it today. tems development problems or problems with
The problems that SSA has demonstrated congressional oversight of that process.
will be likely to afflict other government agen- For long-term success in achieving systems
cies as they adopt, and struggle to use, ad- modernization and allowing SSA to use infor-
vanced information technologies. mation technology efficiently and effectively
A number of Federal agencies are like SSA in carrying out its mission, a strong systems
in that they handle huge volumes of highly planning capability is crucial. SSA’s effective
standardized data, deal with individuals planning horizon is limited to 5 years forward,
directly, and are now absolutely dependent on and is focused on achieving the state of the
information technology systems to perform art of today technology, not on being at the
their missions: the Internal Revenue Service, leading edge of information-handling technol-
the Bureau of the Census, and the U.S. Treas- ogy as it continues to develop rapidly. SSA
ury are obvious examples. They are vulnerable officials have said that they do not want to
to the same pressures that caused SSA to fall be on the leading edge but rather in a position
behind technologically, and those that the of average industry performance. This means,
agency encountered in modernizing and man- however, that in 20 years it may again be a
aging systems, to the extent that these are decade behind current practice, unless it con-
structural or generic problems. tinually forecasts and monitors emerging tech-
nological capabilities with a view toward their
Congressional oversight procedures did not future utilization.
detect or understand emerging problems in
SSA —for reasons that involved the priorities There are several other areas in which Con-
of SSA, the Administration, and congressional gress may wish to clarify its policies or priori-
committees themselves. Similar problems in ties with regard to SSA practices; among these
making congressional oversight effective are are safeguards for the privacy and integrity
almost certain to occur in the future, and to of client information. This report of the SSA
become progressively more troublesome. A systems modernization efforts is intended to
defensiveness on the part of SSA career offi- help in foreseeing and understanding those
cials, described in other chapters of this report, problems.

WHY SSA FACED POTENTIAL DISRUPTION


OF SERVICE IN 1982
The first question addressed in the report In 1946, SSA was placed within the Federal
is, ‘‘why did SSA face potential disruption of Security Administration; thereafter, it was
services . . . through critical system deficien- headed by a single commissioner rather than
cies’ as it stated when announcing its Systems a three-person board. A few years later, it be-
Modernization Plan in 1982. The reasons for came a component in the new and massive De-
these deficiencies bear on the extent to which partment of Health, Education, and Welfare.
such problems are unique to SSA or may be
In the late 1950s and 1960s, as information
of more general governmental concern. The
technology steadily improved, SSA developed
steps through which SSA moved toward serv-
a special relationship with IBM, and worked
ice disruption will be briefly summarized be-
closely with that company to adopt and adapt
low. (Part II of this report is a more detailed
systems to fit its needs. This was in no way
case history. )
unusual, since IBM was then the clear leader
in the field, and computer systems in most Fed-
Brief History of SSA a n d eral agencies were predominantly IBM. SSA
Information Technology was a leader in use of information technology
through much of this period, although as late
Three Decades of Healthy Progress
as 1971, SSA operations, like those of other
In its first few years, 1935 to 1939, SSA be- data-handling organizations, were still heav-
gan to establish a reputation as a highly effi- ily paper-based.
cient and well-managed agency. It was able to
recruit a well-qualified staff, set high stand- Early Signals of Emerging Problems
ards for data security and privacy and for
responsiveness to client needs, and maintain After Public Law 89-306 (’‘the Brooks Act’
low administrative costs. SSA then enjoyed was passed in 1965, it was clear that SSA
relationships with computer vendors and its
a high degree of autonomy. Its needs for data-
handling equipment stretched or exceeded the methods of procurement would have to change.
limits of then available technology, but the But by this time IBM had able competitors.
agency was able to work closely with manu- Competitive procurement should not have
facturers to push the state of the art. Commis- caused major troubles. By the end of the 1960s,
however, there were emerging but not fully rec-
sioners were experienced managers and main-
ognized problems at SSA. By virtue of hav-
tained a good balance between attention to
ing been one of the first users of large com-
daily operations and insistence on long-range
planning for technological development. puters, SSA also had the oldest system; it was
no longer at the leading edge and would fall
From 1940 to 1970 there was steady growth steadily farther behind. If new systems were
in SSA operations. Congressional support for not necessarily to be IBM equipment, there
social security programs had become broadly would be problems of compatibility. Software
bipartisan. Programs were expanded and new conversion and updating would be a growing
programs were added; but as the volume of necessity. Documentation of changes in the
work expanded the work force grew propor- software would become more essential; but the
tionately, at least in the first two decades. SSA importance of this had not been fully realized
employment tended to be a lifetime career, and earlier, and as the software had been adapted
workers and managers had a strong shared to accommodate changes in benefits or eligi-
commitment to the social programs for which bility determination procedures, these modifi-
the agency was responsible. cations had not been well documented.
12

Because SSA had been at the forefront of ware. Repeatedly the time allowed between the
computer use, its system designers, managers, passing of a law mandating changes and the
programmers, and analysts had to learn their time at which they were to go into effect proved
craft on the job. People at SSA tended to stay to be inadequate. The changes could not be
there, and some allowed their skills to become made in an orderly and efficient manner, and
obsolete. Because promotions were based on were accomplished only at the cost of heavy
seniority, and because SSA no longer had state- overtime for the workers, high error rates, and
of-the-art technology, it became difficult to disruption of other activities (e.g., quality con-
make room for, or to attract, highly trained trol, new software development, and above all,
newcomers. long-range planning). Backlogs became a recur-
rent problem.
In the late 1960s and early 1970s, these and
other emerging problems were not highly vis- Instability of leadership and repeated and
ible to Congress, and perhaps not even to incomplete reorganizations of the agency, per-
SSA’s top management. As the work force haps intended to solve the problems, instead
grew, labor-management relations inevitably made them worse. Between 1973 and 1981,
became more complex. The training and ori- SSA had seven different commissioners or act-
entation given to new workers was somewhat ing commissioners, with an average tenure of
diluted and their commitment to the goals of 1,1 years. There were four drastic reorganiza-
social security programs was less personally tions, none of them fully completed in the sense
and directly translatable into standards of per- of establishing clear jurisdictional boundaries
formance and loyalty to the agency. In the mid- and program account ability before the next re-
1960s, with the rapid growth in most employ- organization. In the course of these, the activ-
ees’ workload, conditions deteriorated, and ities of the major social security programs were
there were signs of serious tension. After 1962, split apart and distributed between functional
union membership and collective bargaining divisions of the agency; program coherence was
for Federal workers made labor-management lost and performance measures were obscured.
problems more adversarial. Advanced systems planning was fragmented,
its professional resources drained to bolster
Beginning in 1968, there were several forced
over-stretched and failing operations; finally
reductions in SSA’s labor force, although the
advanced planning was almost completely dis-
work had increased with the expanded social
continued. Mechanisms for decisionmaking
programs of the Great Society. The increased and for review and control of technology pro-
workload strained the operating efficiency of
curements, which had been institutionally sep-
the agency and further stressed the workers,
arated, were merged so that important checks
while recurring announcements of layoffs
and controls were lost.
caused resentment and feelings of insecurity.
However, growing use of computers allowed In the larger world, it was increasingly rec-
the work to be absorbed without serious dis- ognized that software development, rather
ruptions or delays. than advances in hardware, was the key to the
effectiveness of future computer systems. In
this area SSA was now falling further and fur-
A Decade of Deepening Problems
ther behind. The complexity of its operations
During the 1970s, SSA’s problems deepened and the frequent changes in procedures that
and became intractable. From 1972 to 1981, were necessary required frequent modifica-
15 new laws made changes in retirement, sur- tions of codes and software programs, but
vivors, and disabled insurance programs. Four these were done piecemeal and under pressure,
of these made significant changes in entitle- with little attention to uniformity, standards,
ments and benefits. This often required exten- documentation, or knowledge of the just be-
sive changes in coding and revisions in soft- ginning discipline of software engineering.
Promotion by seniority for systems person- sional oversight committees blamed SSA for
nel was by now taking its toll. The long tenure poor technological decisionmaking, for mis-
of SSA upper managers carried with it experi- estimating the resource requirements of the
ence, loyalty, and dedication to the mission new program, or for failing to inform Congress
that was SSA’S strength, but the managers of the impending crisis. They suspected that
also developed a deep defensiveness and a sus- SSA had been prevented from making its prob-
picion of both consultants and new adminis- lems and resource needs known by its parent
trators who criticized established procedures agency and the Office of Management and
or tried to introduce management innovations. Budget (OMB), whose overriding priority was
to reduce SSA work force and budget. Em-
The passage of the Supplemental Security
ployees of SSA say that agency officials had
Income (SS1) Program, at the end of 1972,
repeatedly warned the Administration and, in-
evoked a crisis—still spoken of at SSA as a
directly at least, Congress that the prepara-
disaster–that made the agency’s problems
tion time and the work force for the new pro-
only too visible. This program was very differ-
gram would prove inadequate. All parties to
ent from other SSA programs, although the
the debacle agree that the oversight process
differences were not widely recognized at the
had somehow failed to reveal the extent of
time. It involved much greater interaction be-
SSA’S problems in meeting congressional
tween SSA service representatives and clients,
mandates, and those problems had become in-
often under conditions of distress and urgency,
tractable.
and these interactions often took on the char-
acter of prolonged negotiations, or became con- The Systems Modernization Plan introduced
tentious and adversarial. in 1982 was designed to solve these problems.
(The progress that SSA has made since then
SSA had two planning groups studying SS1
is discussed in ch. 2.) Neither SSA’S problems,
proposals before Congress passed the bill, but
nor severe criticism of the agency in Congress
there was considerable doubt until the last mo-
and elsewhere, has disappeared. Stringent en-
ment that it would pass, and in any case the
forcement of the Disability Amendments Act
planners had no resources to do more than min-
of 1980 and the Debt Collection Act of 1982
imal paper studies. After the bill passed, SSA
brought SSA strong criticism. Pressure to re-
had 14 months to get ready. At this point it
duce the SSA work force troubles its employ-
chose to put in place a new telecommunication ees. There are many concerns about the agen-
system to link service representatives with
cy’s ability to respond to future congressional
headquarters. There were only one or two ter-
mandates for changes in its programs or pro-
minals per office, the systems failed frequently,
cedures.
and the communications traffic exceeded ex-
pectations and soon saturated the communi-
cation lines. The communication system be- Underlying Factors in SSA’S Critical
came a bottleneck in SSI operations rather Systems Deficiency
than a facilitator.
Problems Common to Large
Moreover, the number of people trained to Data-Handling Organizations
operate the system and to provide client serv-
From this brief overview and more detailed
ices was grossly inadequate. Long lines of
discussions in Part III, broad factors that con-
clients formed, waited for hours, and were sent
tributed to SSA’S nearly disastrous situation
home to come again. Huge backlogs developed.
can be identified. Some were problems to which
SSA reputation and public relations suffered
many private sector organizations had also
severe damage, and its chronic problems were
fallen victim. The restructuring of the com-
now visible to Congress.
puter industry in the 1960s had caused wide-
Congress had not anticipated this outcome, spread confusion and floundering in systems
and was surprised and indignant. Congres- planning.
14

SSA’S problem was not that its computers long-term systems development; top-level ex-
were “old” or “obsolete.” It was that the work- ecutive officers, who were not technologically
load had become too large, too complex, and sophisticated, did not insist on its importance;
too dependent on automated processing to be and political decisionmakers did not want to
handled by SSA’S existing work force with ex- encourage demands.
isting technology. In this situation, every addi-
tion to the workload became a potential cri- Special Problems for Federal Agencies
sis. Information processing was: Box A summarizes some conclusions from
● pieced together, program by program, a series of OTA assessments of Federal Gov-
with manual handling and mechanical ernment Information Technology.z This spe-
flow of data in between automated steps, cial report on the Social Security Administra-
with no agencywide planning or design- tion found that many of the generic problems
ing of a system that could allow the work identified in these earlier studies could be ob-
to flow smoothly, and little backup for sys- served in this agency. Many of the Social Secu-
tems that “went down” when the work- rity Administration’s problems are not particu-
load was heaviest; lar to it, but typical of problems in large
● the computer systems were driven by Federal Government organizations. The con-
heavily patched, inefficient software, with clusions in box A were based on cross-cutting
years’ of changes and revisions that were examination of many Federal agencies, but can
poorly documented; and also be read as a diagnosis of SSA problems
● the data in 50 years of SSA files was in information technology management.
categorized and recorded variably and un- Excessive Instability of Leadership.— While
systematically, with data definitions that much of the SSA work force, up through the
differed across files; there was no “cor- levels of middle management, suffered from
porate (agency) database. ” a lack of “new blood, ” the top level of man-
Advanced computer systems cost millions agement was continually changing. Commis-
of dollars and several years to procure and im- sioners are political appointees; in recent years
plement. These investments cannot be lightly they came and went almost yearly. Most had
abandoned. Continuing modernization of hard- little understanding of advanced technology
ware requires continuing software upgrading resource needs and constraints, or technology-
and conversion. When technological capabil- oriented management, but sought to gain con-
ities are improving rapidly, the leaders in a trol by reorganizing the agency.
field, having sunk large costs, may be over- Frequent, drastic reorganizations broke up
taken and left behind by more recent adopters the earlier coherence and accountability of ma-
of the technology. Government officials are not jor programs, but failed to provide what may
as free to take risks as are corporate entrepre- have been better–a rational structure based
neurs; nor can they independently undertake on a redesigned work flow and technology-
to raise capital for new ventures on the gam- based functions. There were no organization-
ble that this will pay off in the marketplace. wide systems or system development planning,
Thus Federal agencies are particularly vulner- because operations and systems development
able to falling behind the state of the art in
technology. 2
U. S. Congress, Office of Technology Assessment, Electronic
The greatest management failure at SSA surveillance and Civi] ~Jiberties (Washington, DC: U.S. Gov-
ernment Printing Office, October 1985); Ikfanagernent, Security,
was lack of planning and advanced develop- and Congressional Oversight [Washington. DC: U.S. Govern-
ment. Professional competence in computer ment Printing Office, February 1986); Electronic Record Sys-
technology was scarce and had to be devoted tems and Individual l+it~acy (Washington, DC: U.S. Govern-
ment Printing office, June 1986). Conclusions in box A are
to solving immediate operational problems; the quoted or abstracted from Management, Security, and Congres-
budget did not provide adequate resources for sional O}ersight, pp. 1-5.
Box A.—General Problems in Federal Government Information Technology and
Information Policy As Identified in Previous OTA Assessments

Strategic Planning: Failure to:


. include strategic as well as operational planning in 5-year plans;
• identify innovative opportunities for use of information technology;
● connect planning effectively to implementation;
● involve users, clients, and the interested public in the planning process;
explicitly consider the implications of information technology use for protection of data secu-
rity and privacy; and
• have an effective vision of the future, with strategies for using new technology to further
government missions.
Information Availability and Quality
● There are serious deficiencies in the scope and quality of information available to Congress
and to agencies themselves, which can hamper effective congressional oversight and agency
decisionmaking. There is a need to specify the types of information that should be reported
to assist oversight of information technology, and to strengthen data quality standards and
procedures.
Innovation
● Where there are examples of agency innovation, such as use of electronic mail, videoconferenc-
ing, and computer-based decisionmaking, the exchange of this experience and learning with
other agencies is irregular or nonexistent. Many agencies view innovations as too risky to try.
Procurement
● Government information technology procurement is subject to multiple and sometimes con-
flicting efforts to simultaneously expedite the procurement process (e.g, through General Serv-
ices Administration’s delegation of procurement authority), increase the level of competition
(e.g., the competition in Contracting Act), and more clearly demonstrate a significant return
on investment in information technology (as now required by OMB).
● The “success” of procurement is closely tied to the government’s ability to plan and define
technology needs and to match technology to those needs. There still appears to be a need
for: better training of procurement staff, greater senior management involvement in and under-
standing of the planning and procurement process, improved mechanisms to exchange procure-
ment experience and learning, and possibly a procurement and management troubleshooting
team to assist with serious trouble spots.
Information Resources Management (IRM)
● This concept was intended to bring together previously disparate functions—such as com-

puters, telecommunications, office automation, and the like-and to establish the importance
of information as a resource. Actual implementation of IRM varies widely and has been only
partially or minimally implemented in many agencies.
. IRM is essential for large, long-term investments in equipment and its related training and
recruitment demands. Chief executives are not in control long enough to realize the return
on investment in resources spent for long-range planning and development.

were constantly forced to compete for profes- ficult by the imperatives and constraints that
sional resources and management attention, are special to public sector organizations.
and because reorganizations were generally not SSA’S performance in coping with technologi-
completed and “set,” before a new commis- cal change in the face of these necessary gov-
sioner and his team took over. ernmental rules was particularly poor; but the
explanation need not involve conspiracy, mal-
Private sector organizations had similar feasance, or even special ineptitude. These pit-
problems. SSA’S problems were made more dif- falls will continue to beset governmental agen-
cies as they attempt to reap the benefits offered man to Congress-its chief executive officer—
by advanced information systems.] effectively represents the Administration,
rather than the agency.
In some cases, new commissioners arrived
with a political directive (i.e., immediate bud- Lack of Control Over Systems Procurement.–
get reduction) that had to take priority over Competitive procurement became a serious
finding long-term solutions to chronic opera- problem chiefly because SSA had been accus-
tional problems or improving service delivery. tomed to working closely with the vendor to
In addition, the scale of SSA operations and develop systems tailored to its needs, did not
the extent to which they had become depen- clearly understand its technological require-
dent on technological systems made it diffi- ments, and already was struggling with soft-
cult for new leaders to understand SSA prob- ware loaded with poorly documented patches.
lems quickly. This was compounded by the By the time competitive procurement laws were
defensiveness of the long-tenured middle man- passed, technological choices were broad, the
agers. Committed to “getting the checks out computer manufacturing industry offered
on time” and barely able to cope with grow- many alternative systems and vendors; there
ing backlogs, many of them feared any inno- were many large computer-using organizations
vation, seeing it as carrying a risk of disrupt- whose needs provided the stimulus for further
ing daily operations. technological innovation. SSA poor procure-
ment procedures rather than the legal require-
Lack of Control Over Changes in Its Workload
ments for maximum competition caused it seri-
and Commitment of Resources. –Corporations
ous troubles and opened the way both for
—if well managed—consider many factors be-
defective systems and for fraud and abuse by
fore seeking a greater market, offering new
SSA officials.
services, or making significant changes in their
operations. They pay particular attention to Nevertheless, the ability to schedule procure-
the timing and to the manpower, skills, and ments at the best time for the organization,
equipment that will be required. Often this to carry them through quickly, to choose with-
analysis is made easier by studying the experi- out constraints, and to gamble on innovative
ence of similar or competing organizations. For but unproven technology, gives private sec-
SSA, assumption of new programs, provision tor organizations a large advantage over gov-
of new services, and changes in benefits and ernment agencies in making technological
entitlements are mandated by Congress. In choices.
government there are always at least three sets
Impatience in Collecting the Return on Invest-
of actors in this situation: Congress, which
ment.–Putting in place radically different
mandates changes in mission and responsibil-
technology for carrying out operations requires
ity; the agency, responsible for performance;
adjustment of the flow of work, changes in in-
and the Administration, which can to some ex-
ternal jurisdictions, and acquisition of a skilled
tent constrain the flow of resources, and to a
work force and management team. Attempts
large extent control the communication to Con-
to grasp the benefits too quickly–to sharply
gress of resource needs. An agency’s spokes-
cut labor costs before the automation is ready
to pickup the load–can lead to overloads and
‘For example, the Internal Revenue Service had serious trou- disruptions that discredit the systems and un-
bles with its 1985 tax processing; a contractor review (accord- dermine management’s commitment to the
ing to Computerworld, which obtained the report from a con- technology. The expectation that this will hap-
gressional committee), said that IRS lacks a strategic plan, and
its processing system is ‘‘inefficient, fragmented, and difficult pen at SSA is now causing renewed resistance
to maintain.” Mitch Betts, “ IRS Systems Need Revamp, Au- to and criticism of the SMP both within SSA
ditors Say’, Computerworld, Mar. 24, 1986. p. 1. and in oversight organizations.
WHY MONITORING AND OVERSIGHT FAILED
TO CORRECT SSA’S PROBLEMS
During the 1960s and 1970s, SSA was pro- cur frequently. The commissioner and his
gressively less able to respond to congressional aides presented in documents and several
mandates without herculean efforts, resulting hearings, a dramatically strong picture of
in large backlogs, high error rates, deteriorat- SSA’S “crisis” in operations. Congress
ing cost-effectiveness, and worsening work- had not heard such strong statements in
place conditions. Repeatedly, congressional the past, and heard them in 1982 only af-
oversight committees were unpleasantly sur- ter an internal struggle at SSA. Some SSA
prised by these outcomes, as revealed in later middle and upper level managers then dis-
hearings. Yet oversight mechanisms, during puted (and still dispute with some bitter-
this time, had been exercised diligently. Sev- ness) the accuracy of these statements—
eral factors contributed to these unpleasant either because the statements were exag-
surprises: gerated, or from reluctance to reveal past
shortcomings, or to protect their power
● Differences in priorities between Con- base within the agency.
gress, the Administration, and SSA itself All information that congressional over-
constrained SSA communication of re- sight committees receive is of necessity
source needs. affected by the political objectives of the
The White House and both Houses of Administration, and by the attention and
Congress have been controlled by the concerns of the congressional committee
same party for only 4 of the last 18 years. that poses, or fails to pose, crucial ques-
Administration constraints on the budget tions to the responsible officials.
sometimes overrode estimates of the Conflicting priorities among oversight
number of man hours necessary to make committees further obscured SSA devel-
changes that Congress mandated, whether oping problems.
the Administration supported or opposed Oversight of SSA is shared by several
those legislative initiatives. Thus in 1972, committees in both the House and Sen-
the overwhelming demands that the Sup- ate. Appropriations committees have one
plemental Security Insurance Program set of concerns—accountability and effi-
would place on SSA staff were not clearly ciency; other committees focus on social
communicated to Congress, although programs and the special concerns of the
enough analysis had been done within aging or of disabled workers; others are
SSA to make clear that new and more chiefly concerned with competitiveness in
time-consuming ways of dealing with procurements. In addition, since 1981 the
clients would be necessary. House and Senate have been led by differ-
In 1982, both the Administration and ent parties, which emphasizes differences
Congress supported the large investment in priorities in guidance to SSA, SSA offi-
in SMP but with differing perspectives on cials repeatedly assured each committee
its justification and expectations of its that the agency was attempting to re-
outcomes. The Administration supported spond to its chief concerns, without much
automation as a means of increasing discussion of the conflicting directives
productivity and trimming the Federal la- that this implied.
bor force. Many Members of Congress Alternative channels of communication
give greater emphasis to improved respon- failed to reveal the cumulative, interac-
siveness to clients. tive, and long-range nature of emerging
The way in which SMP was justified to problems.
congressional oversight committees in GAO performed scores of studies of
1982 illustrates a problem that may oc- SSA for various congressional coremittees
18

during this period. Most addressed spe- Most congressional oversight commit-
cific, narrowly framed questions posed by tees do not have either Members or staff
the committees, usually having to do with with the specialized training and experi-
technical performance or compliance with ence to fully understand or challenge what
a particular law. It was difficult for any they are told about the increasingly com-
one person —even by reading all of the plex and esoteric field of advanced infor-
reports—to get an integrated, coherent mation systems. Experts themselves dis-
picture of the situation that was steadily agree on many critical issues of design,
developing. (GAO’s new series of agency capabilities, choice, implementation, man-
management reviews takes a more in- agement, and lifecycle costs; and few ex-
tegrated approach.) perts can discuss these questions in jar-
Special studies by national commissions gon-free language understandable to the
focused on the viability of the social secu- nonspecialist. Staff members who have
rity Trust Fund and did not probe SSA made themselves expert on one aspect
management issues; others (such as the (such as competition in procurement, or
Grace Commission and the National Acad- systems capabilities) may still not be ex-
emy of Public Administration) looked at perienced in problems of management.
management issues but did not give great Relatively few are able to ask hard ques-
attention to quality of services or to the tions about the likely course of technologi-
long-range future. cal development over the next 10 years.
● SSA officials themselves had little moti- Thus, it is increasingly likely that many
vation to call attention to accumulating important questions about long-range re-
problems. turn on public investments will go un-
Congress does not confine its attention asked by Congress and thankfully un-
in hearings to the Commissioner. It heard answered by government agencies.
from many other SSA officials or former ● SSA own estimates may have been un-
officials, vendors, and outside experts dur- reliable.
ing this time. There are also less formal Though SSA people say that their re-
channels of communication between bu- quests for resources were repeatedly cut
reaucrats and Congress. SSA officials who by the Department of Health and Human
testified were nevertheless either politi- Services or OMB before reaching Con-
cal appointees or under their control, and gress, these estimates of requirements
so inclined not to dispute the Administra- were themselves often the focus of dispute
tion position. Those who held positions in between operations officials, systems de-
operations and those who held positions velopment people, and the Commission-
in systems development, moreover, often er’s office; it is not clear whether they were
lacked a comprehensive or disinterested credible projections.
view of the problems that were develop-
ing. Some others who appeared at hear- As Federal agencies become more dependent
ings either had no credible source of in- on large computer systems both for operations
formation about SSA’S internal problems and for internal administration, critical infor-
or (especially vendors or potential contrac- mation that Congress needs for effective over-
tors) had a vested interest in possible con- sight will more and more be embedded in large
gressional actions. databases. Management information systems
● Members of Congress and the staff of can be made to extract, combine, manipulate,
most oversight comittees lacked the spe- and format data to produce performance meas-
cid”zed expertise to challenge statements ures, accounting categories, benchmarks, and
about advanced data-processing capabil- trend projections tailored to almost any pur-
ities, options, or resource requirements. pose. The temptation to present such informa-
19

tion ‘ ‘in the best possible light (or, according stakes riding on investment in systems that
to purpose, in the worst possible light) has al- promise high, but delayed, return on invest-
ways existed; but computers can make it eas- ment, It will be more and more difficult for
ier to get away with it by removing such proc- most Congressmen and Congresswomen to
essing a few steps further from easy perusal, challenge what they are told by agency offi-
everyday experience and plain common sense. cials about their technological choices and
At the same time, the motivation to conceal problems.
mistakes or failures is increased by the high

THE BASIC STRATEGY OF SSA’S SYSTEMS


MODERNIZATION PLAN IS SOUND
Another question addressed by this report 5. To Add Automated Management Tools:
is the soundness of the basic strategy of SNIP. To these primary goals were added, some-
That strategy is: what later, the development of automated
techniques for managing and scheduling
1. To Upgrade computer Capacit&v: TO con-
computer operations, and development of
solidate SSA scattered computing sys- information systems for use in manage-
tems and sites, greatly increase its total ment and administration.
computer capacity, acquire more modern
computers, develop a local network for Most elements of this strategy are noncon-
high-speed data transfers, and acquire bet- troversial, but there are several points at which
ter peripheral equipment. the strategy has been questioned, as is dis-
2. To Integrate Its Database: To rational- cussed in more detail in chapter 2. One debate
ize and integrate files into an SSA data- centers on whether centralization of process-
base, by moving files onto disk storage, ing was a sound choice; the alternative is dis-
achieving direct (random) access to data, tribution of data-processing operations to re-
developing an overall ‘ ‘database architec- gional centers. A second critical question is
ture, ” and establishing a data dictionary whether SSA should have bought or developed
to standardize the definition and form of all new software, rather than choosing to pre-
separate units of data, serve, modify, or rewrite millions of lines of
3. To Institute Modern Techniques of Soft- code. (In practice, SSA now appears to be tak-
ware Engineering: To retain and upgrade ing a middle course, rewriting some code and
as much as possible existing software, re- developing new software to modernize some
writing as much code as necessary; enforce operation s.)
consistent standards for all future soft-
ware; improve and modernize all software The questions and criticisms regarding basic
documentation (reference manuals, user SMP decisions are serious ones, but they do
and training manuals, records of changes); not have definitive ‘‘right’ answers to which
develop new software applications. all experts can agree-in general or with re-
4. To Build a Data-Communications Utility: gard to a specific organization such as SSA.
To re-engineer and consolidate three ma- In terms of the basic SMPstrategy, the choices
jor telecommunications networks into a that SSA has made may not be demonstrably
modern, expanded conduit for two-way the “best” choices but they are reasonable, in
transmission of data and interactive com- accord with well-established engineering-
munications between service representa- tices, and defendable. There is no guarantee
tives in the field offices and the headquar- that alternative choices or strategies, urged
ters processing operations. or implied by critics, would be more assured
20

of long-range success or involve fewer prob- broader question of how long-range technol-
lems or risks, or indeed fewer doubts and ogy planning and development— which will al-
criticisms. ways be beset by uncertainty and risk—can
The main thrust of SMP strategy is to pro- best be evaluated in ways that are both useful
to Congress and supportive rather than threat-
duce first a modernized claims process, with
ening to public servants with a difficult mis-
service representatives in field offices using
sion to perform.
interactive terminals to access SSA’S head-
quarters database to answer the clients ques- What is needed is a mechanism or mecha-
tions and transmit an application to headquar- nisms for providing both agency officials and
ters for final processing. Other program Congress with an independent and disin-
procedures and management activities will also terested source of expert advice and evalua-
be automated as the plan proceeds. It can be tion, separate from monitoring and investiga-
argued that SSA should be looking much fur- tory functions and also apart from both
ther ahead at developing technological capa- regulatory responsibilities and political objec-
bilities and new ways of accomplishing its tives. While there would often be a lack of con-
mission, rather than automating today’s pro- sensus among such expert advisors, the range
cedures with today’s technology. This, how- of options available for consideration by agen-
ever, involves a degree of innovation, and cies and by Congress would possibly be broa-
perhaps risk-taking, that few government dened and the relative advantages and dis-
agencies are willing or able to assume. advantages of the options clarified before
choices are made. Such mechanisms already
Substantial progress has been made in some
exist in the several congressional support agen-
areas of the SMP and there have already been
cies, but their assistance is often sought after
large expenditures of time and effort which
basic decisions have been made and imple-
should show results in the near future. Tore-
mented. Also, since they are located within the
quire SSA to begin again with a different strat-
legislative branch, their assistance and advice
egy does not appear to be justified even the
is usually not available in helping agencies
uncertain strength of the critical arguments.
frame action proposals to be put forward to
To conclude that the basic strategy is sound the Administration or to Congress. One alter-
and should not be abandoned, however, does native is to create new mechanisms for this
not necessarily mean that SSA can and will kind of public service. Some possibilities are
carry the systems modernization plan to a suc- outlined in the options section of this chapter.
cessful conclusion. Neither will it answer the

CONFLICTING EVIDENCE ON PROGRESS


IN SYSTEMS MODERNIZATION
There appear to be serious implementation while SSA proceeds with hardware procure-
problems to which SSA–in spite of strong ment and upgrading, it consistently downplays
claims of accelerating progress-does not yet the problems it is having in defining a data-
appear to have a credible solution. Denial that base architecture and making decisions about
these problems exist, or unsubstantiated as- database integration and management. Only
sertions that they have been solved (in ways persistent challenges to the statements of SSA
that to outside experts are not clear or con- officials and comparison of those statements
vincing) leave considerable room for doubt that with information gleaned from workers, con-
SSA understands its technical difficulties or tractors, and monitors, reveal the existence of
is addressing them adequately. For example, some of SSA’S persistent, unresolved techni-
21

cal problems. GAO’s forthcoming manage- Major programmatic systems computers


ment review, based on examination of records have been upgraded and “mean-time to fai]-
not directly available to other congressional ure ” increased from 270 hours to 19,000 hours.
support organizations, may answer at least Telecommunication processors and some de-
some of the questions raised in this report cision support systems have been installed.
about the adequacy of SSA response to tech- Major files have been moved from tape stor-
nical problem areas and the amount of sound age to disk storage. The six Program Service
progress that has been made. (See table 1.) If Centers are still using very old, too small com-
GAO confirms the apparent gaps between puters and a procurement award for their
SSA’S official statements and the degree to replacement has been held up by a challenge
which SMP objectives are being realized, this under the Competition in Contracting Law. In
will underscore the increasing difficulties in general, however, the capacity upgrade pro-
monitoring highly technology-dependent gov- gram, which will account for about 24 percent
ernment operations. of SMP expenditures, is approximately on
schedule, with other procurements to be com-
pleted this year.
As of the end of fiscal year 1986, it appears
that about 20 percent of SMP’S currently pro- The soft ware engineering program has had
jected costs have been expended. Major pro- serious problems and is behind schedule. SSA
curements for the SMP, especially for placing claims to have completed essential early steps:
the interactive data communication terminals definition of its functional requirements for
in field offices, will occur in fiscal years 1987 data-handling and software applications, de-
and 1988. Some SSA critics say that before veloping software engineering standards, and
such steps are taken, Congress should insist preparing a basic Software Engineering Tech-
on a complete reexamination of the assump- nology manual. The agency says that it is mak-
tions and strategy of the SMP with a view to ing a strong effort to enforce new tools and
aiming it in radically different directions, or standards for software development.
formulating a quite different plan. At a mini- Some new software systems and applications
mum, they say, the procurement of 22,000 to have been put in place. SSA apparently judi-
39,000 interactive terminals should not be done ciously abandoned vague promises to rewrite
this calendar year as scheduled—instead, ter- all old code, but software improvement has be-
minals should be phased in over the next 2 to gun. SSA is, however, still far from having
5 years, or the procurement should wait until made its existing software ‘maintainable and
all field office services are redesigned and auto- transferable, as was to have been achieved
mated. Even though there are no convincing by this time.
arguments for reversing SMP’S basic strategy
at this stage, it is not certain that SSA is mak- There is evidence that in some areas the func-
ing satisfactory progress in development of tional requirements are not well enough devel-
software; it would not be unreasonable to move oped to be an effective guide to systems re-
more slowly in making major procurements design, and that use of software engineering
until there is proof of acceptable progress in tools and standards is not yet stringently en-
all areas. This decision, however, should be forced. Software will not be ready for full and
made in the light of its effect on SMP as a efficient use of the new interactive terminals
whole. To stop a pivotal SMP procurement on being procured for field offices for several
the grounds that the 1974-82 operational cri- years.
sis has been surmounted would effectively be About 21 percent of total SMP projected
a rejection of the concept and objectives of sys- costs are allocated to the software engineer-
tems modernization. ing technology program; expenditures in this
program have been running somewhat ahead
The most unequivocal progress in imple- of projections.
menting SMP has been made in upgrading the Database integration is also far behind
capacity of large primary computer systems. schedule; SSA now says that by 1987 this part
22

T a b l e .—Major Reported Accomplishments of the SSA’S Systems Modernization Plan


in Its First 4 Years and Future Milestones by Specific Program Areas

Accomplishments a
Software engineering program:
● Piloting a modernized claims system at two dis ri ct ● Design, develop, and implement LAG software
offices ● Upgrade SET
• Initiated functional requirements for LAGs ● Continue software improvement
● Completed operational software improvements ● Implement claims modernization nationwide
● Designed new debt management system

● Began piloting critical payment SET

● Designed AWR

• Established PDTF/TTSF
Database integration program:
• Implemented nationwide, on-line query capabiIity on ● Implement on-line omnibus query capability on all
several major master files for district offices major master files
● Converted major fiIes from tape to disk storage ● Complete data purification through verification and
Ž Implemented data administration tool validatlon
• Began piloting target database architecture ● Develop and implement database architecture
● Initiated data purification
● Provide database support for LAGs
● Initiated database support for LAGs

Data communications utility program:


● Continued with procurements of DCU network and
● Implement DCU backbone network nationwide
terminals ● Acquire and install new terminals for district offices
● Replaced SSADARS host computers

Engineer future network components and expanded
• Upgraded telecommunicate ions Iines and software capabilities
● Completed general DCU design

• Planned DCU and TAP implementation


Capacity upgrade program:
● Completed seven phases of DASD installation ● Install new hardware at PSCS
● Converted on-line, programmatic, and test systems to ● Upgrade programmatic, telecommunications, and test
MVS/XA Operating System capacity
● Implemented local computing network and
HYPERchannel facilities
● Installed high-speed printers at Baltimore sites
● Replaced programmatic and TTSF host computers
System operations management program:
● Implemented new tape management system Institutionalize and enforce data center

Continue to
. Increased job run frequency for critical system standards
● Implemented automated job rerun capability ● Complete user service agreements
● Improved off site storage process ● Expand use of automated tools to on-line and decision
support systems

Complete NCC integrated control center
● Modernize computer operations at PSCS
● Implement on-line operating environment
Administrative/management information engineering program:
● Established information center ● Implement new systems to provide reliable MIS
● Initiated office automation projects for SSA ● Increase office automation
components ● Implement SSA’S portions of FAIMS
● Initiated MID project to provide reliable management ● Develop management information database
information for SSA architecture
● Initiated projects to define management Information ● Provide telecommunications support for management
requirements for MID informat ion
● Implemented end-user computing guidelines
● Developed framework for an integrated MIS
● Installed FAIMS database management system on the
TTSF
a
The text raises questions about some of these accomplishments
KEY AWR —Annual Wage Reporting NCC —National Computer Center
DASD —Direct Access Storage Device PDTF –Program Development and Test Facility
DCU —Data Communications Utillty PSCs — Program Service Centers
FAIMS —Financial and Administrative Integrated Management System SET –Software Engineering Technology
LAG — Logical Application Groups SSADARS—SSA Data Acqusition and Response System
MID –Management Information Design TAP —Terminals Acquisition Project
MIS – Management Information System TTSF —Test and Time Sharing Facility
MVS –Multiple Virtual Storage XA – Extended Architecture
SOURCE U S Department of Health and Human Services Social Security Administration SSA Systems Modernization Plan 1986 Long Range Stralegic Plan Publication
No 40004 October 1985
23

of the implementation may be back in step. In 1984a new component was added to SMP,
Some tasks have been accomplished. Master the systems operation and management pro-
files have been transferred from magnetic tape gram, to develop automated procedures for
to disk storage, and the number of tapes in use scheduling and managing major computer op-
for storage has been significantly reduced. erations. It has already implemented auto-
Field offices have been given limited access to mated job scheduling and several other appli-
the master data files through a file manage- cations, and is on schedule. This will account
ment and access system, although processing for less than 3 percent of SMP costs. Another
is still sequential, and random (direct) access element belatedly added to the plan is the ad-
to records is still beyond SSA capability. A ministrative/management information engi-
data dictionary has been developed, but it will neeringprogram, to develop management in-
take years to rewrite all of the 50 years of formation systems and other administrative
records to make them fit the data dictionary tools, and to encourage personal computer ap-
categories and standards. plications and use. This effort is one of the more
advanced elements of SMP, although hardly
Real database integration is, however, still
avant-garde. Not including this element in the
far in the future. SSA has still not settled on
original SMP was a blunder that may have sig-
a database architecture, although the agency
nificantly increased the costs of the manage-
says that a “target” database architecture has
ment information systems development. The
been defined. This in turn further delays the
program will probably account for about 20
rewriting of software. It is difficult to tell
percent of SMP costs.
whether SSA has, in fact, made any signifi-
cant progress toward real database integra- SSA reports that the backlogs and high er-
tion. Recent statements that it has taken a big ror rates that marked the crisis period have
step toward solving the architecture problem largely been overcome, that SMP has already
by deciding to use an already available data significantly improved performance, and that
management system appear to be almost the basic steps have been accomplished to al-
meaningless on close examination. low continuing and steady progress in the later
phases of SMP.
This program was originally expected to
spend about 14 percent of projected SMP According to SSA, significant progress has
costs; it appears so far to have accounted for been made in developing new mechanisms for
about 4 percent of expenditures. strategic planning and for resolution of the per-
sistent conflicts between operations and sys-
The data communications utility program
tems development personnel, through their
appears to be about on schedule. Troublesome mutual involvement in the systems modern-
data transmission backlogs were greatly re- ization effort. There is said to be an active pro-
duced during the first year of SMP by replac-
gram of outreach to the systems users to fur-
ing the host computers, adding trunk lines, and
ther define changing functional requirements.
upgrading telecommunications. The backlogs The agency has expanded its training pro-
have now been eliminated. The design of the
grams as it implements the SMP, and has re-
communications utility has been completed, cently recruited some senior computer systems
and by early 1987 SSA plans to put over 22,000
experts. A joint agreement with the union was
interactive terminals in field offices, to mod-
signed last year, which appeared at that time
ernize its claims process. The timing of this
to hold great promise for improving labor-
move is controversial; many critics argue that
management relations.
terminals cannot be used to full capacity for
several years, and a full-scale procurement One critical test of SSA’S claims of improved
should not go forward at this time. The data management and resolution of internal con-
communications development program has ex- flicts over systems modernization will come
pended 7 percent of SMP costs to date, but during the next 6 to 9 months, as the claims
will eventually account for about 28 percent. modernization project is implemented, If these
24

improvements are real, they are more hopeful ance indicators have been changed and, there-
signs of progress than acquisitions or quanti- fore, results of SMP (in terms of comparison
tative measures of performance, because such with past performance) cannot be demon-
management innovations would signal a change strated. Such private comments may some-
in SSA organizational culture and behavior. times be based on biased judgments, or on ob-
Such changes are probably essential to the suc- solete information; progress of SMP has, if
cess of SMP. SSA claims are accepted, quickened in this fis-
There are, unfortunately, reports that these cal year in spite of some unanticipated delays
new mechanisms have already disappeared or because of challenges under the Competition
in Contracting Act.
gone dormant. The joint agreement with the
union is being severely strained by SSA offi- Both SSA statements about progress or re-
cial position with regard to work force reduc- sults and the statements of its critics are dif-
tion, and there is widespread disappointment ficult to evaluate since those who do not have
with the present lack of activity in putting its a vested interest to protect also do not have
provisions to practice. Since the announcement independent access to operational data or close
of a change in leadership of SSA, the expecta- familiarity with SSA’S complicated tasks.
tion of another drastic reorganization has GAO is now undertaking an extensive man-
raised fears of a protracted period of uncer- agement study of SSAJ that will provide
tainty, confusion, and possible internal power another expert judgment; however, both GAO
struggles. The present organizational structure auditors and OTA assessors are in part depen-
is probably far from ideal; however, it has the dent on information selected and presented by
advantage of allowing for agencywide rather SSA.
than program-by-program design of an auto-
The more fundamental difficulty for Con-
mated work flow, and its continuation for a
gress, however, has been and will be the ne-
while could help avoid the disruption and tur-
cessity of making judgments about complex
moil caused by repeated reorganizations of
technological strategies for which there are no
SSA over the last decade. Assessment of the
categorically right or wrong answers and on
likelihood of progress in systems moderniza-
which even computer experts disagree.
tion in the near future must take into account
these troublesome uncertainties.
‘Recognizing that good management is essential to the ef-
Many critics of SMP are skeptical of SSA’S fectiveness of a department or agency in achieving its mission,
GAO in 1982 launched a new initiative, to perform reviews of
ability to achieve its objectives. Some individ- the overall management of selected Federal agencies. These re-
uals inside and outside SSA and in monitor- views are to facilitate effective congressional oversight by show-
ing organizations privately dispute some of ing how breakdowns or problems in agency management struc-
tures and systems contribute to long-standing programmatic
SSA’S claims of progress, say that bad news and administrative problems. The GAO management review
is being concealed, or suggest that perform- of SSA is not yet complete.

THE EFFECTS OF SYSTEMS MODERNIZATION:


CLOSELY RELATED ISSUES
Systems modernization is likely to affect and tion, privatization of government services,
to be affected by a number of questions and Federal labor-management relations, and data
issues now before Congress: suggested modifi- privacy and security concerns. These are dis-
cations in social security programs, independ- cussed briefly below, and in more detail in chap-
ent status for the Social Security Administra- te 3.
25

SSA Responsiveness to 4 to prevent


. measures that they believe will
Congressional Mandates adversely affect the quality of social secu-
rity services, such as overly zealous at-
The ability of SSA to respond efficiently and tempts to cut disability rolls.
quickly to congressional changes in programs,
entitlements, and benefits has improved be- Stable, experienced leadership could contrib-
ute significantly to success in systems mod-
cause of the elimination of backlogs of work,
and should be further improved if the systems ernization, if that modernization is a high pri-
modernization plan meets its objectives. So far, ority of the appointed leaders. Independent
however, the elimination of large backlogs and status might do relatively little to facilitate
achievement of a smoother flow in daily oper- congressional oversight, because it has also
ations has been made possible largely by the been hampered by other factors, as described
hardware improvements–the capacity up- above, including SSA own tendency to hide
grade. Significant further improvement prob- its problems. Independent status must neces-
ably depends on resolution of the technical sarily be limited to a few agencies, yet most
difficulties with software development and of the problems that SSA has had in manag-
database management, and redesign of post- ing information technology are likely to affect
entitlement systems. It is, therefore, possible other government agencies as their informa-
that assignment of responsibility for large new tion needs expand and as they first adopt new
programs (e.g., as support for immigration con- information systems. Congress can clearly not
trol measures) at this stage could complicate make all agencies that suffer from these prob-
lems–or from specific Administration direc-
and delay implementation of some SMP steps
by suddenly increasing its volume of opera- tives—into independent agencies.
tions, or requiring new data to be collected and Systems modernization is thus not, in itself,
managed. Some congressional sources have an argument for giving SSA independent
suggested a moratorium on legislative changes agency status. However, if Congress decides
until SSA is closer to completion of its sys- that Administration personnel and privatiza-
tems modernization, but this is probably not tion directives are likely to disrupt SSA oper-
essential. The changes that appear most likely ations before systems modernization can be
to be proposed over the next few years, accord- achieved, this option will become more at-
ing to congressional committee staffs, appear tractive.
reasonably small and could probably be assimi-
lated without the large backlogs that occurred Privatization of or Contracting Out
in the 1970s.
Government Operations
Independent Status for SSA The Department of Health and Human Serv-
ices has directed SSA to develop a plan for
The House of Representatives has (in July
contracting-out the equivalent of 8,600 full-
1986) passed H.R. 5050, a bill to give SSA in-
time positions, about 12 percent of its work
dependent status, and similar proposals are ex-
force, in operations such as the processing of
pected to come before the Senate. Support for
annual wage reports and running the National
the measure comes from some who hope:
Computer Center (where central beneficiary
1. to give SSA more stability and continu- records are maintained). Privatization of de-
ity in leadership; termination of disability status (now done by
2. to facilitate congressional oversight of the States) has also been proposed. OMB,
SSA by removing the “political filter” GSA, and GAO have found that privatization
that they believe distorts communications of some government services can result in sig-
with Congress; nificant cost-savings and improved services.
3. to protect SSA from OMB work force re- But there are serious concerns to be consid-
duction and privatization pressures; or ered in privatizing social security operations:

61-085 0 - 86 - 2
26

these include the additional risk to confiden- Privacy and Security Concerns
tiality and security of SSA’S personal data on
160 million Americans; questions of public con- Systems modernization will facilitate and
fidence in the fairness of eligibility and bene- probably encourage data-sharing programs
fits determinations; the level of competition and computer-matching programs that have
that could be expected; the large amounts of expanded under OMB directives and GAO rec-
time and labor that would be necessary for con- ommendations. SSA is now considering their
tractors to learn and master SSA operations; use for front-end verification of eligibility,
possible disruptions from periodic recomple- which has not been done in the past. These
ting of the contracts; the likelihood of conflicts activities are considered useful for elimination
of interest for many or most competent com- of waste and fraud, although SSA has not sys-
petitors; opportunities for fraud; additional tematically evaluated their cost-effectiveness.
difficulties of congressional oversight; and There are, however, growing concerns about
difficulties of contractually specifying a re- intrusions on personal privacy when data col-
quired level of quality of service. lected for many specific legitimate purposes
is aggregated and used for other purposes,
An important question is whether privati- and/or shared with other Federal and State
zation would reduce the return on investments agencies.
already made in SSA’S systems modernization.
Congress will want to consider carefully Security measures for SSA’S main com-
whether privatization initiatives are likely to puters and databanks have generally improved
prejudice the objectives that it has sought in since 1982 with consolidation of processing
supporting SMP, i.e., improved qua.hty of serv- activities in the National Computer Center and
ices and equity as well as efficiency in use of improved backup of files. SSA does not, how-
public resources. ever, have procedures and policies to assure
privacy and security for data in personal com-
puters. Opportunities for violations of privacy,
for fraud, or for inadvertent loss of data will
The Work Force and Labor- increase as SSA places interactive terminals
Management Relations in field offices and puts a new data communi-
cations utility into use. Although SSA plans
Increasing tension between labor and man-
to use standard techniques of restricted access,
agement during the decade of mounting prob- passwords, audit trails, etc., for protection,
lems in the 1970s has worsened since 1982 with
many of the planned control systems have not
the threat of severe work force reductions. Em- yet been developed.
ployees and their union take the position that
improved productivity should be translated Other new technologies which SSA is using
into enhancement of services and better work- or will use in the future, ranging from personal
ing conditions rather than immediate elimina- computers to satellite transmission and in-
tion of jobs. tegrated services digital networks, will also in-
crease the opportunities for unauthorized ac-
SSA and the union recently agreed to a Joint
cess to, misuse of, or theft and loss of data.
Statement of Common Purpose toward labor-
SSA has done relatively little as yet to imple-
management relations. After a promising
ment, or even plan for, privacy safeguards for
start, that agreement is now said by union
some of these technologies.
sources to be breaking down. Labor-manage-
ment relations are likely to worsen if systems Systems modernization will tend to inten-
modernization is directed toward immediate sify concerns about the privacy and security
labor force reduction. of SSA data.
27

OPTIONS FOR FACILITATING CONGRESSIONAL OVERSIGHT


OF SYSTEMS MODERNIZATION
In spite of the attentions of a half-dozen con- 2. Increasing the number of GAO audits, or
gressional committees and frequent hearings, studies from other sources, is a second op-
emerging problems were allowed to become tion. GAO audits and several national
chronic and Congress was repeatedly surprised commissions have been invaluable in sup-
by SSA’S serious difficulties in implementing porting congressional oversight but have
congressional mandates. At present, OTA has not entirely solved the problem. National
identified a large number of unanswered ques- commission studies usually provide only
tions and unresolved issues about which there a snapshot of the situation at a given time,
are strongly conflicting critical charges and and are in addition usually slow, costly,
SSA claims regarding SSA information sys- and necessarily rare events. GAO studies
tems development and management. There are have in the past been technical and nar-
disturbing signs that SSA’S statements on rowly focused, responding to the specific
some of these questions cannot be taken at face perspective and concerns of the request-
value. Some of these questions are in that area ing committee or of the Administration,
of uncertainty where there are no definitive reflecting the fragmentation of oversight
answers and even experts may disagree among responsibilities. They thus tended to over-
themselves. It is increasingly difficult for non- look intensifying interactions between
specialists to challenge the actions or the problems, as well as the effects of one con-
statements of agency managers, who must gressional directive or legislative require-
both be given support in carrying out difficult ment on other competing congressional
assignments and be held accountable for their concerns. GAO is however now undertak-
actions. New ways of supporting and assist- ing a broader management study of SSA
ing congressional committees in their diffi- which will be available to Congress later
cult oversight role may be needed, including in 1986, and will provide additional insight
sources of advice and evaluation that are not into current information technology ma.n-
associated with investigation, regulation, and agement problems and progress.
assignment of blame for inevitable mistakes. 3. Designation of one committee in each
At a minimum, Members of Congress and their House, or a joint committee representing
staff are concerned that they have access to both Houses, for comprehensive oversight
information about agency needs and agency of the Social Security Administration is
problems. This information can be provided a third option. This would tend to simplify,
most easily by the agency, but is often filtered integrate, and intensify oversight of the
or distorted to fit executive branch policies and agency, and allow Members and staff to
priorities. With particular regard to SSA, there expand the attention they can give the
are a number of options that address the over- agency and deepen their knowledge of its
sight issue: needs and problems. However, it might
lose the benefit of different points of view
1. Independent agency status for SSA has and specialized knowledge that can be
been proposed as one approach to this brought to bear by other committees.
problem, but it is likely to be at best only 4. A more tightly focused mandate for the
partially effective. Executive branch pri- subcommittees on government informa-
orities have been only one factor in over- tion technology management presently in
sight problems; some of the trouble has each House is a fourth option. At present
come from inside SSA. Moreover, this so- the responsibility for information technol-
lution is a special or limited answer that ogy is in each house combined with other,
cannot be applied to all agencies that may somewhat disparate responsibilities. A
present similar problems. more tightly focused mandate would in-
28

crease the attention given to this subject. Federal agencies in technology-related


This would, however, tend to cause con- choices in an advisory and impartial way,
sideration of technological issues to be and thus could provide a counterweight
divorced from considerations of each to Administration pressures for actions
agency’s special mission, the quality of its that are not realistic in terms of techno-
services, and congressionally proposed logical capabilities.
changes in missions. Such a group could be located within an
5. An external advisory body of nationally existing congressional support agency.
recognized experts on advanced informa- This may however not be the best strat-
tion technology could be established to as- egy, because:
sist all oversight committees now con- ● it is difficult to attract into government

cerned with Federal agencies that are service people of the prestige and stand-
becoming dependent on information tech- ing that would make for greatest credi-
nology to carry out their missions. For bility, to assure them of independence,
best use, this body of experts should not to give them the resources necessary to
be charged with monitoring, investiga- keep their expertise and their prestige
tion, or routine assessment, but should be at the desired high level; and
available, staffed, and ready on a continu- ● within a congressional agency, they

ing basis to translate for Congress in would be viewed by the executive agen-
discussions of technological issues and op- cies as investigatory and threatening
tions, to evaluate agency and Administra- rather than as advisory and helpful.
tion positions on basic technological Thus a blue-ribbon panel, selected
choices and strategies, and to alert Con- from industry and academia, with a
gress to technological trends that might small but highly expert staff, may be
offer alternatives. They could also assist preferable.
Part II

The Systems Modernization Plan


and Its Potential Effects
Chapter 2 describes the Social Security Administration’s Systems Modern-
ization Plan, discusses the status of its implementation, and identifies some per-
sistent technical and management problems. It concludes that the plan is rational
and defendable, but there are serious unanswered questions about the implemen-
tation of the plan. SSA does not appear to recognize the seriousness of some of
these implementation problems, or has not been forthright in discussing them
with monitors and oversight institutions.
Chapter 3 discusses the implications of systems modernization and further
automation for SSA’S relationships with Congress, SSA’S employees and clients.
It considers questions about the future status of SSA, including proposals to make
it an independent agency, and to privatize some of its functions. Other public pol-
icy issues, such as the privacy and security of personal data processed by SSA,
are discussed in relation to the SMP.
Chapter 4 highlights the increased need for comprehensive long-range plan-
ning within SSA, to define goals and priorities and thus provide a context and
rationale for technological systems planning.
Chapter 2

The Status of Systems


Modernization at SSA, 1986
Contents
Page
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
SMP’s Principles and Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
SMPProjects and Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
The Claims Modernization Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
The Software Engineering Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
The Database Integration Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
The Data Communications Utility Program . . . . . . . . . . . . . . . . . . . . . . . 43
The Capacity Upgrade Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
The System Operation and Management Program . . . . . . . . . . . . . . . . . . 47
The Administrative/Management Information Engineering Program . . 47
Planning and Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Evaluating the Progress ofSMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Staff Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
SSA Claims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Critics of SSAand SMP.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
How Well Has SSA Performed? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Imperatives for SSA.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Figures
FigureNo. Page
2. An Overview of the Social Security Administration’s
Systems Modernization Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3. organizational Chart of Social Security Administration asof
Apr.4, 1986, Focusingon Systems Implementation . . . . . . . . . . . . . . . . . 37
4. The SSA Characterization of the Database Architecture
To Be Used in SMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5. Available Computer Capacity and Projected Workload Requirements
of SSA’s Computer System, as projected in 1982 . . . . . . . . . . . . . . . . . . . 46
Chapter 2

The Status of Systems


Modernization at SSA, 1986’

INTRODUCTION
Commissioner John Svahn, newly ap- for the fact that the district offices had to
pointed,’ began in 1981 to work out a strate- serve all of the programs.
gic plan to develop modern information sys-
SSA had always tended to automate on a
tems for the Social Security Administration’s
project-by-project or program-by-program ba-
(SSA) data processing. A planning group was sis, which resulted in poor integration at the
formed under his direct control, to guide the
service delivery level. This had been institu-
planning and its presentation to Congress. tionalized by the previous commissioner, who
The Systems Modernization Plan, hereafter had adopted a “partitioning strategy” that
referred to as SMP, was to be an integrated segmented the several programmatic areas, re-
long-range plan for thoroughly upgrading designed them, and procured hardware for
SSA’S data-handling operations, with new or them separately. SMP explicitly rejected this
improved software, hardware, and telecommu- approach.
nication systems and increased computer ca-
The SMP also differed in other ways from
pacity. Unlike previous SSA systems devel-
earlier SSA practices. It was designed as a dy-
opment efforts, SMP would be an agencywide
namic 5-year plan that would be reconceptu-
plan emphasizing integrated service to all pro-
alized yearly to account for new developments
grams and offices. In the recent past, work on
in technology. (The published 1982 plan, how-
improving systems had been done in specific
ever, did not say explicitly that it would con-
SSA program areas, with little consideration
tinue beyond the initial 5 years.) The plan pro-
vided explicitly for help from external expert
The material in this chapter (not otherwise attributed) is consultants and contractors. The solicitation
drawn from the 1982, 1984, 1985, and 1986System Moderniza- of vendors for a telecommunications upgrade
tion Plan, and accompanying documents, provided by SSA; sev-
eral briefings by top-level SSA officials in the Office of Sys- (later won by Paradyne) was already underway;
tems and the Office of Operations; and written and oral response it was assumed that this would fit into the
by SSA officials to OTA drafts or inquiries. This material was SMP to develop a “data communications util-
augmented by or compared with information gleaned from more
than 65 interviews conducted by OTA and its contractors, with it y, that is, an efficient conduit for transmis-
knowledgeable people both within and outside of SSA; congres- sion of data between headquarters and field
sional hearings and reports; reports of the General Accounting offices or other points on an SSA network.
Office (GAO); and the transcript of an OTA workshop in which
both SSA officials and advisors to OTA participated. The OTA The 1982 SMP was focused almost entirely
contractor for this case study was The Educational Fund for
Individual I.iberties, New York City (Alan Westin and Kenneth on delivery of services; internal administrative
I,audon, Principal Investigators). systems got little attention, and no provision
‘Commissioner Svahn was a former insurance company ex- was made for developing management infor-
ecutive. From 1976 to 1979 he had, however, been employed
by the firm of Deloitte Haskins & Sells (later a major subcon- mation systems. These features were added to
tractor on SMP). the plan later.

33
34

SMP's PRINCIPLES AND STRATEGY


At a cost of nearly $500 million (which by gued for starting fresh, with all new procedures
1986 nearly doubled, to $990 million) SMP was and systems, but SMP rejected this approach.
one of the most expensive single civilian in-
The plan calls for salvaging prior invest-
formation systems projects ever undertaken.
ments by building on existing systems. This
The plan set out ‘governing principles, ” which
means that SSA will look for immediate short-
in reality are generalized aspirations: imme-
term solutions that are compatible with long-
diate improvement to avoid disruption of serv-
range goals. SSA has been criticized in the past
ice; improved client service; assured account-
both for patchwork fixes to problems rather
ability and auditability; improved timeliness
than system redesign, and for redesigns that
of service; improved productivity and manage-
failed to take into account the critical prereq-
ment control; and closing of the technology
uisites for an orderly transition.
gap (i.e., modern systems). Nine principles
were to be followed—these are important ele- In the past, the same personnel had respon-
ments or descriptors of the systems modern- sibilities for both systems development and
ization strategy for purposes of evaluation: operations, and there was seldom time for mod-
ernization planning. A single organizational
1. improvements would be “incremental and
unit, separate from operations, would now be
evolutionary responsible for planning, management, and
2. systems modernization would be kept
control of the modernization program.
separate from operation and maintenance
of existing systems; A system integration contractor (Electronic
3. a systems integration contractor would Data Systems of Dallas [EDS]) was hired to
assure project continuity; provide continuity throughout the duration of
4. “proven state-of-the-art technology” the plan, and to coordinate across SMP pro-
would be used; gram areas. Redirection of development efforts
5. the effort would “build on existing sys- in midstream and frequent turnover at the top
tems” to salvage past investments and had hampered past efforts.
minimize risks;
SSA would not be able to work with manu-
6. design changes would be limited to “crit-
facturers to develop innovative systems de-
ical, user-defined needs”;
signed to meet its needs, as it could do in its
7. system architecture would be reconfig-
first decades (see ch. 5). SMP called for proven
ured to take “full advantage of advanced
state-of-the-art systems from industry. This
technology’
meant that no “unproven technology” would
8. acquisitions would be planned to permit
be used. This strategy was reinforced after
technology upgrading within a “code
SSA suffered from its experience with a tele-
compatible architecture”; and
communications system upgrade procurement
9. a single group would plan, manage, and
(the Paradyne contract, to be described inch.
operate the modernization program. 6). The phrase “state of the art, ” on the other
This was a conservative strategy, following hand, was a signal that SSA would use con-
well-established systems engineering practice, temporary software development technology,
and designed to satisfy SSA’S critics in Con- and structure and document software in ac-
gress and elsewhere, while not disturbing its cordance with modern standards.
supporters. SSA’S Office of Advanced Sys- The plan limited design changes to “criti-
tems, before it was abolished in 1979, had ar-
cal, user-defined needs, but said that systems
architecture would be reconfigured to take full
‘U.S. Department of Health and Human Services, Social advantage of advanced technology. SSA’S as-
Security Administration, Office of Systems, Systems Modern-
ization Plan: From Survival to State of the ,Art, Publication sumption was that with relatively simple
No. 41-002, 1982, p. 1-19. Hereafter cited as SMP 1982. reconfiguration of existing computer systems
35

and some purchases of new equipment in the Combined with strong governmentwide em-
first phase of its modernization, a large amount phasis on budget-trimming and work force re-
of labor-intensive operations could be elimi- duction, the announcement of the systems
nated and performance immediately improved. modernization effort in 1982 caused SSA em-
ployees considerable anxiety. As in any orga-
Upgrading technology in such a way as to nization acquiring new technology, many work-
be compatible with SSA’S old computer codes ers were concerned about their ability to learn
would be difficult, since the agency had to to use it. At the same time, most employees
avoid both demanding an architecture and soft- were eager to have technology that could help
ware that was compatible with only one kind them overcome the constantly increasing back-
of equipment (IBM), as required by the Brooks logs and recurring crises, and the union was
Act; and massive reconversions of software. not opposed to more automation. SSA how-
SMP began in 1982, although it incorporated ever failed to keep its employees well informed.
some improvement projects that were already In early documents there was no mention of
underway. In the discussion below, some rough the touchy subject of effects on the level of
indications are provided about the allocation employment. In a brochure published some-
of resources among and between SMP program time in 1984,4 SSA states that “an overrid-
areas, to indicate something about the relative ing consideration” was that “all current SSA
importance of tasks and objectives. However, employees must be assured of job security,
this gives only a very poor indication of the but the promise is not part of the formal doc-
distribution of effort and resources; some ob- umentation of SMP. Only in 1985 did SSA
jectives have been shifted from one program announce an” aggressive’ plan to inform em-
to another between 1982 and 1986. Even the ployees about SMP, and in January 1986 it
overall SMP expenditures indicated by SMP distributed to field operations employees a sim-
publications are only approximate, since some plified “Field Edition” of the plan. Questions
projects have been included under SMPin 1 or about job security were still not addressed
2 years and not included in other years, for rea- directly.
sons that are not clear. This is one problem that
complicates any external evaluation of prog- ‘U.S. Department of Health and Human Services, Social
Security Administration, S.vstems Modernization Program–
ress in implementing the SMP. An Overview, no date.

SMP PROJECTS AND PROGRAMS


The 1982 SMP called for four program areas: hardware configuration to achieve final in-
software engineering, database integration, tegration, and the testing and certifying of
data communications utility, and capacity up- the redesigned system. By 1988, with this
grade. The three chronological phases of the achieved, SMP would evolve into a continu-
program were labeled survival, transition, and ing 5-year planning and enhancement cycle.
state of the art (see figure 2). The survival
phase consisted of actions to survive the im- One important objective of the SMP is mod-
mediate crisis, which is described in chapter ernization of the claims process, which is per-
haps the primary point of interface between
6. The transition phase would bring SSA up to
SSA and its individual clients.
a “contemporary data processing capability. ”
These phases were each to take 18 months, and The Claims Modernization Project (CMP)
to be completed by 1985. The state of the art
phase, the final 2 years, would develop the new This project is in effect a plan, or a depic-
software, new databases, new communications tion of the major desired outcomes of the SMP,
utility, distributed processing, and the final and it is therefore described first. CMP is
36

Figure 2.—An Overview of the Social Security Administration’s Systems Modernization Plan

Survival Transition State of the art

Software Software improved Software salvaged New benefit payment


engineering and under control and redesigned system implemented

I
Database Tape dependency Master files Database management
integration reduced converted to disk implemented

Data Message backlog Equipment upgnaded Data communications


communications eliminated and and communications utility implemented
utility reliability increased utility designed

1
Equipment
Capacity reconfigured and Expanded, Advanced system
upgrade workload capacity architecture installed
requirements met

1
Computer center Expanded
Systems move completed automated control
Production
operation environment fully
and operating of computer
management standards set operations
automated
I J
Administrative/
management
information
engineering

18 months
T AMIE
program
established

18 months
Reliable
admi nistrativeimanagement
information systems
initiated

24 months

aNot Included In original SMp


SOURCE U S Department of Health and Human Servtces, Soctal Security Admlnlstratton, SSA Systems A40dem/zatKm P/an 7985 Update, Publlcatlon No 40-004, January
1985

SSA’S closest claim to a vision of how it wants by file folders. Clients must wait for service
to do business, SSA’S version of “the office representatives to send messages to headquar-
of the future. ” Since SMP is a 5-year plan, this ters and receive information back about the
is, however, a near-term future. client records by way of the one or two SSA
Data and Retrieval System (SSADARS) ter-
At headquarters, the four major programs
minals, manned by a data technician, in the
(old-age insurance, survivors insurance, dis-
back office. CMP will make the field offices into
ability insurance, and supplementary security
modern, automated offices in which represent-
income) are fragmented and spread over 10
atives use on-line, interactive systems for both
functional offices under four deputy commis-
the initial claims interview and later for case
sioners (see figure 3). They come together in control. There are prototypes in two field
the district offices, where SSA meets its cli- offices,s where research is being conducted on
ents. These district offices are now largely
paper-based operations, with cases represented “York, PA, and Baltimore, MD.
37

Figure 3.–Organizational Chart of Social Security Administration as of Apr. 4, 1986


Focusing on Systems Implementation
,
Office of the
Commissioner

I
I 1
I ●

Deputy Commissioner, Deputy Commissioner,


Deputy Commissioner, Deputy Commissioner,
Operations Management & Systems Program & Policy
‘Assessment J
I I
v t

Office of Planning
Support & Integration Director

Office of System Operating


Associate Commissioner

— Regional Commissioners Office of System Requirements


(lo) Associate Commissioner

Office of System Integration


i Associate Commissioner I
- Office of Information Systems
Associate Commissioner

SOURCE Social Security Administration, 1986

its impact on the labor force. By 1988 SSA ex- This office of the future is still an objective,
pects the system to be in place nationwide. not an accomplishment, but by January 1987,
the first phase of implementation will begin.
A claims representative will interview the There is a prototype model office at headquar-
client, asking questions prompted by a desk- ters, and there are pilot sites in SSA regions.
top computer screen. The results will be trans- Two pilot sites are already working with bor-
mitted directly to Baltimore and will go rowed GSA terminals, of the kind the U.S. Sen-
through a communications processor located ate is now retiring from Senators’ offices. In
in one of the six program service centers where large service organizations of the the private
the claim is further processed for earnings in- sector, CMP would not be regarded as an “of-
formation. The results of the interview will also fice of the future’ at best it would seem moder-
be printed out locally for the client and for use ately up to date.
in case control.
This will eliminate the need for the claims Eventually, SSA’S batch-processing based
representative to queue up for the one or two claims system will be fully redesigned and
office SSADARS terminals as she or he now automated, as will postentitlement control and
does. It will also eliminate most of the Data audit functions. The first phase of the project,
Review Technicians, the people who now key however, deals with initiating claims. In the
data into SSADARS. (Some will be retrained first z years, at least, the interactive terminals
as service representatives. ) will be used chiefly for tasks related to initial
38
—. —

claims filing. Software for postentitlement pro- ernization that both the Administration and
cedures will not be ready; the postentitlement Congress accepted in 1982. The analysis of the
process has yet to be redesigned. SSA’S plan costs and benefits of this procurement (or
is to procure all of the interactive terminals rather, the timing of this procurement) should
at once (in 1987) so that CMP provides a model include the effects of its timing on other as-
for and a basis of automating other operations. pects of SMP implementation. The risk of pro-
An alternative would be to procure terminals ceeding with the procurement is basically the
now only for some pilot sites, with automation risk of incurring the cost of unused telecom-
and modernization of all 13,000 field offices munications and computer capacity while soft-
to be phased in after 1990, or when software ware is developed for automating additional
development is complete. The advantages of field office services. Delaying the procurement
automating at least a portion of the work im- indefinitely may involve other costs, includ-
mediately, providing improved service in all ing foregoing possible productivity gains in
rather than a few communities, giving employ- claims processing, plus the risk that reductions
ees experience in using the equipment, (and for in staff may cause a deterioration in services
SSA, locking in the allocation of no-year fund- to claimants. The August GAO reort did not
ing) must be weighed against the costs for tem- address these broader considerations. Strong
porarily underused capacity, and the uncer- corrective actions by SSA, DHHS, OMB, and
tainty of relying on software that is not yet Congress may be necessary to resolve persist-
developed. SSA insists that immediate pro- ent SMP implementation problems, but they
curement is necessary for smooth progress in should not be driven by this procurement in
SMP implementation. isolation from the broader and more important
issues. The option of a cautious go-ahead for
In August 1986, GAO recommended that
the procurement, with intensified monitoring
SSA not proceed with the full procurement un-
and oversight, should be considered.
til there is a full evaluation of the complete sys-
tem, although GAO did not clearly specify how
this was to be accomplished.G GAO’s recom- The Software Engineering Program (SEP)
mendation was based on the grounds that the Software engineering is a new discipline that
agency has extricated itself from the 1982 cri- aims to improve software through providing
sis and ‘these procurements are not supported better tools, concepts, and methods for soft-
by documented deficiencies in current ADP ware development and testing, and insisting
operations. GAO also said that there were on their consistent and systematic use. SSA’S
deficiencies in specifying functional require- software engineering program however was de
ments for system components, and in SSA’S signed to retain (so far as possible) and upgrade
cost-benefit analysis for the procurement. The existing software so that an entirely new code
GAO report did not, however, place this recom- would not be necessary. It would also develop
mendation in the context of SMP as a whole, requirements for new software and new appli-
or SSA’S related management problems. cations, and reconfigure the database architec-
The decision about the procurement should, ture so as to take advantage of new technol-
however, be made in the context of SMP as ogy. Finally, it would develop standards and
a whole. To argue that because the 1982 crisis productivity tools for software development.
was surmounted (i.e., Phase I of SMP succeeded), Special emphasis was to be put on modern pro-
the second phase of SMP is not necessary, gram documentation, standardization of pro-
amounts to rejecting the goal of systems mod- grams, and conversion to high-level languages
where possible.
‘U.S. Congress, General Accounting Office, ADP Acquisi- SSA developed a software engineering tech-
tions: SSA Should Limit ADP Procurements Until Further Test-
ing is Performed, report to the Chairman, Committee on Gov-
nology manual between 1983 and 1985, but it
ernment Operations, U.S. House of Representatives, was found to be incomplete and lacking in nec-
IMTEC-86-31, August 1986. essary provisions for quality assurance and
39

compliance, according to GAO; work in this a level of detail that can effectively guide
area is continuing. systems redesign and development. Internal
reviews of specific functional requirements
There were in 1982 some 12 million lines of
repeatedly speak of incomplete functional
poorly written and undocumented program
decomposition, improperly partitioned and
code. There were about 6,000 COBOL* pro-
poorly named data categories, ambiguities and
grams, 1,500 assembly language code pro-
contradictions between data flow diagrams,
grams, and another 1,000 miscellaneous pro-
and many other technical flaws.
grams. Over the years SSA had translated old
manual procedures into software using now A baseline Software Engineering Technol-
outdated programming languages, and then, ogy manual has been prepared. An interim
converted them line by line to COBOL, pre- Debt Management System and a pilot of a
serving the inefficiencies of the older technol- Modernized Claim System are in operation. A
ogy. The old code is being cleaned up and re- software improvement process has begun.
written as it is needed, according to SSA. Over the next 5 years the program will design
and develop Logical Application Groups, de-
The software engineering program has fallen
scribed as methods and systems for enhanc-
far behind schedule. However, SSA claims to
ing security controls and auditing capability.
have accomplished the systematic definition
of its information requirements, for the first The redesign of the batch-oriented claims
time in SSA history. This was done using top- system to a contemporary interactive system
down Business Systems Planning, a technique aims at allowing immediate eligibility and en-
for analyzing an organization’s “business func- titlement determinations, automated compu-
tions’ and defining the needs for software ap- tations of benefits, and enhanced control and
plications. A second technique, Critical Suc- audit functions. SSA has established in its cen-
cess Factors Identification, was also used. tral office a model district office and a test proc-
Establishing the information requirements essing module to evaluate software for district
was a critical first step to help the agency con- offices. Data-entry screens have been designed
ceptualize the uses of data in its procedures, for district offices and the processing center.
and to lay out a general plan for a systems ar- A project is underway to obtain at least 22,000
chitecture (the hardware and software that interactive terminals for district offices (the
would be used to modernize and automate claims modernization project, as described
these procedures). Thirty-five SSA analysts above. ) Field offices are now pilot testing some
then interviewed over 200 managers and work- interactive systems. These projects will be fur-
ers throughout SSA to get a detailed picture ther discussed below.
of the agency’s business and information re-
The annual wage reporting system was also
quirements and an evaluation of existing soft-
to be redesigned, and employers were to be en-
ware, which was inventoried for the first time.
couraged to report wages on magnetic media
More than 180 systems-related problems and rather than paper. However, this project be-
needs were identified. This work is continuing,
came unnecessary when new Internal Revenue
with groups of users from the field office peri-
Service regulations required that all organi-
odically brought into headquarters to work
zations with more than 500 employees file
with the Strategic Planning and Integration reports on magnetic media by 1986, and those
staff.
with over 250 employees do so by 1987.
There is, nevertheless, considerable doubt
In 1981, SSA had $2 billion in outstanding
among many SSA systems developers and
debts owed by people who received over-pay-
expert observers as to the adequacy or qual-
ments. A new Interim Billing and Follow-up
it y of the functional requirements, as defined,
System was put in place in 1984 as a first step
in some areas; some are still not developed at
in improved debt management. This is sup-
Common Business-oriented Language. posed to be replaced by the new Debt Man-
40
—— . —

agement System by the end of 1986, which sional needs expressed in new legislation.7
should further reduce the average age of receiv- GAO cited delays in the SMP database man-
ables and maintain better accountability over agement program, and software efforts. It said
all debt collections. The new system will pro- that SSA had failed to document existing code
vide on-line access to information about over- (over 10 million lines) as originally promised
payments, bills and notices that have been in the SMP and instead had chosen to ignore
sent, and resolution agreements. It may not this problem while developing entirely new sys-
be usable in all program areas by the end of tems in the absence of software standards and
1986, however. There are problems in complet- enforcement. While praising SSA for its hard-
ing the design for, and implementing, the new ware acquisition program, GAO concluded
National Debt Management System, because that SSA had made little progress “in im-
it must interface with postentitlement systems proving its ability to respond to legislative
and procedures which are still to be redesigned changes that require software modifications
and automated; thus the functional require- to existing systems. ”
ments for the debt management system are
Within the executive branch, SMP’S soft-
incomplete.
ware program has also come under criticism.
The automated enumerations screening proc- In the Department of Health and Human Serv-
ess, begun in November 1984, gives SSA the ices (D HHS), the Assistant Inspector General
capability to process requests for social secu- for Audit, Felix J. Majka, conducted a review
rity numbers in 1 day; currently only 3 percent of the claims modernization project from late
of requests require any clerical intervention. 1983 through May 1984, and found numerous
deficiencies. The HHS Inspector General,
The 1982 SMP called for existing software
Richard Kusserow, issued reports in February
to be “made maintainable and transferable”
1985 and again in June 1985, criticizing SSA
and to be fully documented by 1985. This has
for wasting over $1 million in the procurement
not been accomplished. All future development
of useless software. Kusserow criticized SSA
of software is to use software engineering
for “poor planning and management of a soft-
technology -e.g., strict rules, procedures, and
ware replacement effort. ” He pointed to the
criteria to make sure that it can be fully un-
Claims Automated Processing System upgrade,
derstood, added to, improved, and corrected
saying that software purchased from a ven-
when needed. The software engineering tech-
dor was unusable. A similar result occurred
nology was supposed to be ready for full “in-
with an upgrade of the Manual Adjustment
stitutionalization” by 1986. It is not complete,
Credit and Award Process (MADCAP), and the
and what has been introduced is not always
conversion of earnings program software.
strictly enforced. However, SSA is installing
modern techniques to measure compliance, Critics inside and outside SSA point to the
which should then improve. software program as most behind schedule and
suffering from poor management. In inter-
The software engineering program was esti-
views conducted by OTA, critics said:
mated in the 1982 program to cost $103 mil-
lion, or 21 percent of the total SMP 5-year cost. Senior management has seriously under-
In the first 3 years, 28 percent of SMP expend- estimated the difficulty of examining, docu-
itures went to this program. Its total cost menting, and rewriting 10 million lines of code
according to the 1986 SMP will be about $200 found in SSA’S major problems.
million through 1990, still about 21 percent of Standards developed to control software
the projected total development are not being enforced.
On August 30, 1985, GAO released a report 7
U.S. Congress, General Accounting Office, Social Security
to the Senate Committee on Appropriations Administration's Progress in Modernizing Its Computer Over-
concerned with SSA’S ability to meet congres- ations, IMTEC 85-15, Aug. 30, 1985.
41

The functional (detailed) requirements of a volume which increases by billions of bytes


SSA’S major systems have not been produced each year. In 1982, SSA had limited access to
on schedule. We are about 18 months behind its most important systems and production
here. files, which were on magnetic tape. Use of over
The Business Systems Plan was a nice ex- 500,000 reels of tape required extensive sched-
ercise, but it did not lead to redesigning ma- uling and a large clerical staff just to file and
jor SSA processes. move the tapes. Over 30,000 production oper-
These criticisms, and those of GAO and the ations each month required 150,000 tapes to
Inspector General in 1985, may be too severe be handled several times, causing human errors
in 1986, since SSA says much progress has that were estimated to consume each month
been made in the past year. This claim, how- about a quarter of available computer hours.
ever, is difficult to document and relies on SSA It was very difficult to determine the num-
assertions. SSA has discovered, as have many ber of data elements maintained on the vari-
business organizations, that software engineer- ous databases. There was no single formal data
ing is not a scientific formula but a set of tools dictionary with standard definitions of all the
for better programming. Installing these tools data elements.
does not guarantee that they will be used or
that good code will in fact be produced. Some One purpose of the DBI program was to re-
private sector studies indicate that even in- duce the use of magnetic tape through the use
tense application of the tools brings only mod- of shared Direct Access Storage Devices and
est gains in productivity; other experts argue to establish a data administration function (i.e.,
that it can double productivity. Getting SSA a data dictionary) for logical definition of data
programmers with 20 years’ experience to use elements and files. This would make it possible
new tools is indeed a major problem in itself, to use available hardware and software tech-
but SSA is now improving its monitoring of nology to create a modern integrated database.
how much of the new code is produced in ac- In its first phase, the DBI program placed
cordance with software engineering standards. the Master Beneficiary Record and Supple-
The promise, implied or explicit, to document mental Security Record master data files onto
10 million lines of old code was probably mis- disk storage, and provided on-line access to this
guided in the first place, and the “failure” to data for field offices, through the one or two
pursue this objective rigorously is probably Paradyne data communications terminals that
wise. New operational procedures related to each office already had. SSA says that this
the claims modernization process will avoid the project is on schedule. The number of tapes
need for cleaning up some of the old code, and in active use has been reduced from 500,000
the rest can be done as needed. to 250,000. More than 360 disk drive units have
been installed.
SSA has made considerable progress in im-
proving its software, but just as clearly this A file management and file access system—
is the area in which SMP is most behind, and the Master Data Access Method, or MADAM
may be seriously floundering. A critical prob- —was developed to handle more than 500,000
lem seems to be the need for more expert per- queries a day. Data has been separated from
sonnel in this area. applications programs, so that it can be used
and updated independently; this is essential
for modern data management. For the user,
The Database Integration (DBI) MADAM appears to be a modern database
Program management system; the user asks for data,
and gets it, without knowing how to use vari-
The DBI program has achieved its first and ous separate software programs. In fact, how-
second phase objectives, essentially on sched- ever, MADAM extracts the data from a vari-
ule. One objective was to improve the manage- ety of separate files rather than from one
ment of over 1 trillion bytes of data per year, integrated database. The other new software
42

program that performs in this way is the criti- Figure 4.— The SSA Characterization of the
cal payment system. The ability to update all Database Architecture To Be Used in SMP
files at once, automatically, must await a more Database
modern database management system. handler
1
The earnings systems, enumeration systems, Application DRMS
and postentitlement claims systems update interface interfaces
the major master files, now on disks, using
batched sequential access. SSA is still work-
ing toward modern data administration, with
a completely integrated database. Application DRMS
programs
SSA’S recently developed data dictionary de
scribes over 50,000 data elements. A data dic-
tionary, one of the first steps in data adminis-
Application DRMS
tration, defines the data elements—that is, the interface interfaces
pieces of information—that should go into a
SOURCE: U S Department of Health and Human Services, Social Security
database and dictates the form they will be uni- Administration, SSA Systems Modernizatton Plan 1986 Long-Range
formly given and their labels, or the terms used Stategic Plan, Publicatlon No 40-004, October 1985
to call them up, so that retrieval and process-
ing is easier. Although the new data diction- that SMP is behind schedule in developing an
ary is widely cited by SSA as a major accom- integrated database because of delays in
plishment, it is valuable only if it is rigorously procurement. 8 In late 1984, a $9.8 million re-
used. This may not be the case; OTA was told quest for proposals for database architecture
by some people at SSA that the dictionary was development was issued, but only six vendors
often not adhered to in writing programs and bid, and those were judged technically in-
“new uses and new data definitions are pop- competent. The procurement was withdrawn
ping up all over the place. ” and canceled in May 1985. The major vendors
The data dictionary, even if rigorously used, did not bid, reportedly, because the venture
does not solve SSA’S problem. The agency al- was too risky and SSA allowed only 4 weeks
ready has about 80 million records on RSI rolls to write a proposal. Some potential vendors
and 10 million on SS1 rolls, accumulated over said that SSA asked for an “overly ambitious
50 years, with data categories defined in many architecture, ‘g and complained that the Re-
different ways over the years. The attempt to quest for Proposals was vague and confusing.
purify or clean up SSA’S data is staggering; SSA throughout 1985 said that it had moved
one master file run through a data cleaning pro- ahead with developing an architecture on its
gram reportedly “produced 3 billion lines of own, and had made up the time lost on the
print and 120 million invalid values. ” failed procurement.
The DBI program has defined a “target data- Yet in 1986 SSA was still struggling to de-
base architecture”–that is, the general kinds velop a database architecture. In April 1986
of structure, software and hardware, that are SSA told OTA that it had “re-examined” a
needed for organizing its databases, but it has database management system produced by
not yet worked out what that architecture will Cullinet, IDMS/R (Integrated Database Man-
be (see figure 4). When the new database ar- agement System/Revised), which is already in
chitecture is decided on and implemented, it use in HHS, and decided that it would adapt
should have tools to assure that all databases the SSA database architecture to use this soft-
can be updated in synchrony; that has not yet
‘U.S. Congress, General Accounting Office, IMTEC 85-15,
been accomptished. op. cit.
UW.S. Congress, General Accounting Office, Social Security
GAO contended, in a report to the Senate Ati”m”stration Computer Systems Modernization Effort May
Appropriations Committee in August 1985, Not Achieve Planned Objectives, IMTEC-85-16, Sept. 30, 1985.
43

ware package, which according to SSA is com- costs. In the short term, there are also risks
patible with its existing software, including and costs. MADAM is apparently a very com-
MADAM. Whether or not this could solve plicated and poorly documented system, so
SSA’S architecture dilemma is far from certain. that only a small group of people are suffi-
IDMS/R is not one of the newest database ciently knowledgeable to operate it, yet it is
management systems available, but it is widely the basis of SSA’S data management. This con-
regarded as a good system, and it has replaced stitutes a peculiar vulnerability to smooth
IBM database management system in many operations if there is any short-term emer-
large corporations. But some information in- gency, sudden work force reduction, or dras-
dicates that SSA is not, in fact, structuring tic reorganization.l”
an architecture that can use IDMS/R but
The DBI program was allocated about 14
merely “layering” IDMS/R over MADAM—
percent of projected SMP costs in 1982, or $65
that is, using information retrieval and data
million. But according to the 1986 Plan, its to-
management systems to obscure the fact that
tal cost will be less than $3o million (3 percent
it still has no firm plan for database integra-
of SMP) although SMP costs as a whole have
tion. These changing and conflicting reports
doubled. This revision occurred after the failed
provide an excellent example of the near impos-
request for proposals for a contractor to de-
sibility, for those not inside an agency with
velop a database architecture, when SSA
hands-on access to its systems, of distinguish-
decided it would be done in-house; presuma-
ing what is being done in implementing infor-
bly it represents the estimated difference be-
mation technology plans from what an agency
tween in-house and contractor efforts. When
reports it is doing.
the 1985 SMP Update was published, 3% years
Failure to settle on a database architecture into the plan, this program had expended about
in the near future could have severe conse- $7.8 million, or 4 percent of total expenditures.
quences in terms of lost productivity. Fourth
generation languages operating in a modern The Data Communications Utility
database could save thousands of hours of (DCU) Program
programmer time. Many applications could be
written in more efficient advanced languages. The DCU program is to reengineer the three
However, existing programs, those already major telecommunication networks to consti-
written, will be compatible with the proposed tute a data communications utility; that is, a
database architecture. It is in the area of lost conduit for transmission of data between and
productivity that SSA would pay a price for among processing points. With the existing
failure to develop a database architecture. SSADAR system, there are only one or two
communications terminals in each office, oper-
The most controversial accomplishment of
ated by a data technician, and service repre-
the data integration effort is perhaps the Mas-
sentatives have long waits for sending and re-
ter Data Access Method, or MADAM, the file
ceiving messages. In its first 7 years, the
management system that SSA developed when
SSADAR system failed frequently, and was
it converted from tape to disk storage. Many
experts thought that SSA should have sought
or adopted off-the-shelf software for this pur- 1’)As one internal critic said, “If these people get sick, die,
leave, or just get mad, then all of SSA on-line operations could
pose, which would be maintained by vendors, go down. ” Another official charged that “the people who built
rather than developing its own, which it must MADAM . . . refuse to gi~’e management the schematic dia-
maintain (that is, improve, modify, and up- grams and documentation on how MADAM works. All the~
give us is the commands and a users manual. ”Se\’eral SS.4
date). MADAM may well be incompatible with officials concurred in the conclusion that a few people ha~’e used
future mainframe operating systems, database their exclusive knowledge of MA I)Ahl to resist efforts to de-
management systems, and fourth generation \.elop a database architecture without MADAM, and that
hlADAM will hake to “be built around” in designing the ar-
languages. Thus SSA incurs future risks of chitecture. 1 n short, MADAM has become a focus of internal
incompatibility and long-term maintenance tension and dispute as well as external criticism.
44

sometimes inoperable for long periods. Dur- will be let by late summer of 1986, and that
ing the first half of 1981, it was ‘down’ about installation will begin in the fall of 1986.1s
11 percent of the time, or about 1 hour of each But critics have raised serious questions about
working day, on the average. The most imme- whether this program should proceed as planned.
diate objective was to increase the reliability There are in fact two separate controversies
of communications (“the mean time to failure”) surrounding the program: whether the basic
by 20 to 30 percent, and to reduce the amount strategy is sound, and whether SSA’S pacing
of “downtime.” of its implementation is reasonable. In regard
to the basic strategy, two questions are often
Communications software improvement be-
raised:
gan within the first year of SMP. The objec-
tives were: 1) to eliminate the daily return mes- 1. Should SSA be planning to decentralize
sage backlogs; 2) to achieve an acceptable its processing rather than to rely on in-
response time, even if the 1982 volume of daily teractive communication between field
transactions doubled; and 3) to be able to serve offices and processing computers at head-
the needs of all SSA users (including those quarters?
using the new interactive terminals). 2. Can SSA be sure that the traffic between
district offices and field offices can be
The two host computers (IBM 370/ 168s) were
handled?
replaced in 1983, trunk lines were added, and
telecommunication monitors upgraded. These The 1982 SMP strategy is basically one of
immediate improvements significantly reduced creating a highly centralized system. This runs
or eliminated long communications backlogs. counter to a strong trend for large organiza-
tions to decentralize their operations as much
The 1982 plan was that by the end of 1985,
as practical, in both the private sector and the
communications software would be improved
public sector; for example, the State of Utah
to make it “maintainable and transferable, ”
began to move toward distributed processing
replacement concentratorsll and processors
for government operations in 1979, well before
would be installed, the concentrators’ software
the SMP was formulated.
would be converted, local intelligence would
be installed at district offices, and specifica- Distributed data processing was in fact a
tions would be completed for the final data part of the SMP strategy as first announced
communications utility (i.e., communications in 1982. How the SMP strategy came to be
lines, etc.). one of complete centralization of processing
is somewhat mysterious. The 1982 SMP in-
The general design of the communications
cluded “installing] local intelligence at all Dis-
utility has been completed, and in 1987, three trict Office terminals. ” This was a response
very high-capacity machines will increase tele-
to GAO criticism in 1979 of SSA’S planned
processing capacity by seven times over. This
procurement of Paradyne (dumb) terminals,
will be essential as the on-line claims modern-
which predated the SMP. In order to satisfy
ization project, already described, comes to
GAO’s criticism and still proceed with that
fruition.
procurement, SSA agreed, in 1980, that the
The DCU program is essentially on time. It Paradyne terminals would be enhanced in
is expected that contracts for procurement of memory capacity at some time after they were
the 22,000 to 39,0001 interactive terminals installed, to allow distributed processing. Be-
1’Concentrators are the minicomputers which receive data
and query messages from field office terminals, through modems; }{The schedu]e calls for award of a contract in August 1986
and then condense, edit, and reformat the messages and send (competition closed in January 1986), installation of a test site
them on to two main host computers. The concentrators also in the National Computer Center in September, and beginning
send response messages to the proper field office terminal. of installation at 20 pilot sites (claims field offices) in October.
“The procurement is to be for 22,000 terminals with an op- All terminals must operate without fault for at least 30 full days
tion for an additional 17,000; with peripherals, etc., about 60,000 out of 90. On acceptance, 500 will be installed the first month,
devices will be procured. 1,000 the second month, and 1,500 each month thereafter.
45

cause of persistent problems with the Para- Neither of these are insurmountable difficul-
dyne terminals the vague plan to upgrade the ties, given modern data-processing and tele-
terminals was dropped. (This situation is de- communication capabilities. However, this
scribed in ch. 7.) does not necessarily mean that SSA’S choice
of centralized data processing is wrong or un-
After 1982, mention of distributed process-
reasonable. It is true that there are limits to
ing was quietly dropped out of SMP descrip-
the efficiency of enormously large databases
tions. Strangely, this decision—or nondeci-
dependent on a few central computers. Cen-
sion—seems almost to have gone unnoticed.
tralization increases the vulnerability of na-
At late as January 1985, the HHS Inspector
tionwide operations to a breakdown at the hub,
General, in a memorandum to Acting SSA
while decentralization would provide some re-
Commissioner McSteen, said:
dundancy and limit the effects of regional in-
We also found [in a review ending May 1984] terruptions or failures. On the other hand,
that SSA had decided to centralize computer centralization allows for more management
processing even though the original SMP controls, better security, and greater redun-
called for local processing (decentralized). dancy, or better backup systems. Most large
Documentation to support this decision, how- financial corporations, in fact, are not decen-
ever, was not available . . . SSA said that the tralizing their data-processing operations. This
basis for deciding to process centrally was doc-
umented, however, we have not been able to is one of the many points on which experts dis-
obtain this documentation.14 agree, and SSA’S decision does not fly in the
face of accepted professional practice.
SSA officials now say somewhat vaguely
that they are studying the distributed proc- For the present, field offices will by means
essing option and will ‘move in this direction’ of communicating terminals be given the same
in future planning. They claim, however, that functions, capabilities, and access that they
to add ‘local intelligence’ would cost approx- would have with distributed logic, according
imately $25,000 per field office, or about $40 to SSA. The communications network will be
million, and that both technically and economi- be capable of accommodating processing at
cally their strategy is the more defensible any of the communications node, and so will
choice. The agency has, to this point, held to not be a hindrance to any future decentraliza-
a belief that centralized control is necessary tion of processing capability.
to protect the integrity and security of its data. As to the manageability of traffic under
SSA systems planners argue also that distrib- SSA’S plan, some critics point out that the
uted data processing would force them to SSADAR system was designed in 1974 to han-
choose between: dle 20,000 messages and 80,000 data transac-
1. maintaining seven or more compete data- tions per day, and that within 1 year the host
bases in regional centers, with the diffi- computer capacity was saturated, while since
culty of assuring that they are simultane- then the transaction loads have increased over
ously updated and rigorously consistent; 500 percent. They argue that the system could
or again become overloaded as the traffic from
2. dividing the beneficiary files between re- up to 39,000 terminals is phased in. Just as
gions, with the difficulty that beneficiaries highway improvement often encourages addi-
are highly mobile and may turn up at un- tional traffic and ultimately results in more,
expected locations for service. rather than less, congestion, the use of the com-
munications network could exceed expec-
tations.
‘ ‘U.S. Department of Health and Human Services, Office of SSA is confident that it has adequately pro-
Inspector General, Memorandum to Martha A. McSteen, Act-
ing Commissioner of Social Security, ACN 15-52654: Audit jected and modeled traffic on the system for
Report– SSA Redesign of the Claims Processing System Un- the foreseeable future. Basically, it has deter-
der the Systems Modernization Plan (SMP), Jan. 30, 1985. mined the maximum number of transactions
46
.

Figure 5.—Available Computer Capacity and Projected Workload Requirements


of SSA’S Computer System, as Projected in 1982
10 ‘

9 - - - - - Projected workload
t ● — 9
Critical workload
6 Available capacity

SMP
-I-

I I I I I I I I 1 I I
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989

Year
SOURCE U S. Department of Health and Human Services, Social Security Admin istration System Modernization plan from From Survival State of the Art, Publication
No 41-004, February 1982

that a service representative can complete per tures. Big investments are scheduled for fis-
hour, and planned a system that would accom- cal years 1987 and 1988 ($184 million). By
modate all field office personnel making max- 1990, this program is projected to cost $273
imum use of the system at the same time million, about 28 percent of total SMP costs.
(which assumes that the number of field offices
and service personnel will not be increased). The Capacity Upgrade (CU) Program
The second controversy about the program
The CU program directly addressed the cri-
has already been discussed above, under the
sis under which SSA in 1982 could no longer
claims modernization project; it concerns the
meet the elementary, basic demands of its pro-
timing of the procurement of the interactive
grams for computing. Figure 5 illustrates the
terminals.
historic growth and projected workload of SSA
The DCU program was originally estimated computers. In 1982 SSA estimated that it
at $160.5 million or one-third of total SMP needed 5,000 CPU (central processing unit)
costs. By September 1985, it had expended hours per month to handle its workload plus
$12.9 million, or 7 percent of SMP expendi- its backlog, and that its installed capacity pro
47

vialed only 3,000 CPU hours, which was effec- which has delayed the replacement. In general,
tively reduced to 2,000 by “insufficiency of however, hardware acquisition and capacity
operations staff. ” The CU program was to upgrading is on schedule.
reconfigure and consolidate the computing
New hardware and system software must
sites distributed around SSA headquarters, to
also be acquired for the National Debt Man-
acquire much higher capacity and more mod-
agement System, and the Logical Applications
ern computers, eliminate magnetic tape files
Groups. The Test and Time-Sharing Facility
and switch over to direct access devices, de-
must also be upgraded.
velop a local computing network for high-speed
data transfers, and enhance the peripheral The CU program was planned in 1982 to ac-
equipment (such as printers). count for 28 percent of SMP, $132.5 million.
By September 1985 it had spent $72.7 million,
The programmatic systems computers have
or 41 percent of all expenditures to that time.
been upgraded and a separate test and devel-
Other large procurements are planned for 1987.
opment facility was purchased. Computers
By 1990, $237.8 million will have gone into ca-
that averaged only 270 hours “mean-time to
pacity upgrade, or 24 percent of the expanded
failure” (MTF) were replaced with machines
SMP budget.
that average 19,000 hours MTF. National Ad-
vanced Systems telecommunication processors
have been installed, as have smaller systems The System Operation and
for decision support, development, and man- Management Program (SOMP)
agement of the larger systems. Additional
The SOMP was not in the original SMP, but
hardware upgrades are planned in 1987, at
was added to develop automated tools and pro-
which time capacity will be far in excess of
cedures for managing computer operations. It
workloads anticipated in the SMP.
has implemented automated job scheduling at
By 1986 computers at the National Com- the National Computer Center, as well as com-
puter Center, used for programmatic, admin- puter monitoring, training, and an integrated
istrative, and test work, had all been replaced control facility. The small program is projected
and modern disk storage had been largely to cost $27.6 million by 1990, or less than 3
achieved, although SSA still has an enormous percent of total SMP costs, is on schedule.
library of tapes in active use. Operating sys-
tems software has been modernized, laser The Administrative/Management
printers installed, and several terminals added Information Engineer-
for software program testing.
ing (AMIE) Program
The computers in the six Program Service
Centers still must be replaced. Four of these AMIE was added to SMP in 1984. The SMP
are IBM 360/65s that are obsolete by any rea- originally focused on data-processing needs to
sonable criteria. They have smaller capacity carry out primary program responsibilities,
than many personal computers, but are still and gave little attention to managing SSA’S
running major program activities, although resources or providing executives and manag-
constantly threatening a breakdown. 15 The ers with information needed for decisionmak-
1982 plan called for this replacement to be ac- ing and policy. Recognizing that SSA badly
complished by the end of 1985, but a procure- needed a management information system,
ment contract award was protested under the
Competition in Contracting Activities law, ” faced with accepting the possibility of using reconditioned equip-
ment which meets its specifications but would put SSA several
‘;The Deputy Commissioner for Systems sa~w wryly that years behind state-of-the-art technology, or rewriting the speci-
‘‘only SSA and a few Third World Countries still use these com- fications to require some newly developed features, which would
puters. ” not only further limit competition but would significantly de-
A potential supplier protested because the specifications lay the acquisition. It chose to revise the request for proposals
did not make allowance for reconditioned equipment. SSA was to allow vendors to offer reconditioned equipment.
48

Acting Commissioner Martha McStean in using fourth generation database language


April 1984, added this program to develop (IDMS/R). This led to the belated recognition
management information systems software, that it may be possible to use IDMS/R for
automate and modernize administrative prac- SSA’S overall database management, as al-
tices, and encourage end-user development of ready discussed.
new applications. The AMIE was allocated $311.4 million or
An agencywide survey was completed to de- over 37 percent of SMP in the 1985 Update
termine management information needs. An of SMP; the 1986 version scaled this back to
information center was developed to spur $197 million or 20 percent. The cost might have
microcomputer applications; microcomputers been considerably less if management infor-
have been piloted in 20 field offices to study mation needs had been integrated into the
their uses. A Financial/Administrative In- original plans.
tegrated Management System was installed

PLANNING AND INTEGRATION


SMP is a rolling 5-year plan, meant to be up- technologies. This may, however, be a rational
dated each year. Both the planning and the ef- choice for SSA at present.
fort to integrate SMP across programs is the
Other critics, including people within SSA,
task of the Office of Strategic Planning and
maintain that little or no integration is occur-
Integration, within the Office of Systems. This
ring, and that the integration contractor is
OSPI has a staff of 100. There are weekly meet-
often diverted to other tasks. It should be
ings between representatives of the SMP pro-
noted, however, that ‘‘integration’ is a loose
grams described above, with the integration
and relative term, and can only be demon-
contractor. In addition, efforts are being made
strated by long-term results of SMP implemen-
to involve operations people in systems plan-
tation.
ning, since they are the ultimate users of the
systems. Some critics, in fact, argue that oper- The integration role contract will be recom-
ations considerations are determining the pleted when it expires in the fall of 1986 and
directions for SMP, and that this guarantees provision has been made for a 3-month over-
that the emphasis will be on rocking the boat lap with the old contract, so that there will not
to the least extent possible; that is, minimum be a lapse in this function should a new con-
change in SSA procedures and customs rather tractor be selected.
than deriving maximum benefit from advanced

EVALUATING THE PROGRESS OF SMP


It is difficult to measure precisely the several hundred documents, supplemented
progress of a very large organization in a near with more than 50 interviews with current and
billion-dollar effort over 8 years (1982 to 1990), former SSA employees at all levels, with con-
an effort with multiple goals, strategies, and gressional committee staff people; with offi-
areas of effort. One measure is increased pro- cials at the General Accounting Office, the
ductivity, or to be more accurate, achievement General Services Administration, and the Of-
of work force reduction goals. Other indicators fice of Management and Budget; with com-
of progress are more qualitative or judgmental. puter vendors and contractors; and with other
This study relies on inspection or analysis of well-informed observers.
49

There were significant differences in assess- basis of SSA performance data) that claims-
ments of progress to date, between critics of processing times and backlogs decreased.
SSA and its defenders, between representa-
GAO said that “the evidence is inconclusive’
tives of various oversight and monitoring
as to the effect on service to the public. Union
groups, and within SSA management. It
representatives and field office personnel said
should be noted that while SSA claims to have
(both to GAO auditors and to OTA) that serv-
made great progress in solving some of its
ice declined; they reported longer waiting
problems, much of that progress appears to
times, a “less caring attitude” on the part of
have been made within the past 6 to 9 months,
employees, and increased error rates. SSA said
while this study was underway. For example,
that service improved, but GAO said that SSA
SSA has recently shown signs of moving to
performance data was incomplete. For exam-
improve management procedures and to change
ple, SSA does not collect data on waiting times
its corporate culture; it has initiated new train-
for clients, or on client satisfaction.
ing programs, recruited highly trained new
programmers, started new management plan-
ning activities, and consulted outside experts. SSA Claims
Some of the skeptics may not be well informed SSA managers point to the SMP as the first
about developments during that period. At the long-range, dynamic plan for meeting SSA’S
same time it should be noted that all of the information-processing needs, and say that the
information about these developments neces- goals and strategy of the plan are now closely
sarily comes from an interested party, i.e., SSA integrated into operations. The Acting Com-
management. missioner, as early as 1983, claimed significant
benefits from the plan, in terms of decreased
Staff Reduction processing time and other quantifiable output
measures.’” In addition, she spoke of “quali-
Among Administration goals for SMP (cited
tative enhancements, ‘‘ including a general ra-
earlier in this chapter) was increased produc-
tionalizing of SSA procedures.
tivity, for which work force reduction is often
used as an indicator, although it is an input SSA points to a number of surveys of both
measure and not an output measure. In its the general population and beneficiaries, which
1986 budget request SSA formally announced indicate that the public continues to hold SSA
the plan proposed earlier by OMB to reduce service in high regard, as both courteous and
the work force by 17,000 full-time equivalents efficient. In a GAO survey, 78 percent of a sam-
(FTEs) or 21 percent of its 1984 staff, by 1990. ple of SSA clients rated service as good or very
This was to be achieved largely through sys- good, and only 7 percent said it was poor; 51
tems modernization and privatization of some percent said its performance was somewhat or
activities. much better than that of other agencies.
GAO concluded in March 1986” that the SSA’S top managers argue that SMP is a
agency was “essentially on target with its complex, multifaceted program that is now in-
planned cumulative FTE reductions. ” In part, stitutionalized within SSA and has had a pro-
however, this resulted from the fact that ex-
pected increases in agency workload did not ‘“Prepared statement of Acting Commissioner Martha A.
materialize (e. g., anticipated inquiries about McSteen, U.S. Congress, Social Securit.v: How Well IS lt Serv-
taxation of benefits); work-year savings from ing the PuM”c? Hearing Before the Special Committee on Aging,
U.S. Senate, 98th Cong., 1st Sess., Nov. 29, 1983, pp. 8-12.
systems and procedural changes were 24 per- ‘gU.S. Congress, General Accounting Office, Sociai Securit.v:
cent less than expected. GAO reported (on the Quality of Services Generafly Rated High by Clients Sampled,
HRD-86-8, January 1986. The report also noted, however, that
18 percent found SSA mail difficult to understand, 30 percent
‘-U.S. (-on~~ess, General Accounting office, Socia~ Securit~ :
r
found explanations unclear or “somewhat clear, ” and 58 per-
.lct ions aJ)d I)lans To Reduce Agenc>’ Staff, briefing report to cent had some negative comments about SSA service (e.g., long
con~r~’s~ional requesters, II RD-86-76BR, March 1986. waiting times, many telephone busy signals).
50

found impact on SSA’S organizational culture. the agency directly but familiar with its prob-
They point to a number of initiatives not de- lems and critical of its behavior. Two other
scribed in SMP documents that are vital to kinds of critics are noteworthy: higher moni-
its efforts at renewal. Among these efforts are: toring authorities in the executive branch, offi-
cials of OMB and HHS; and Members of Con-
● development of a strategic planning func-
gress and staff concerned with SSA oversight,
tion that will drive the development of in-
who have come to distrust its statements over
formation technology;
recent years.
● enlargement of training programs in the
systems area to assure that new software Many of the most adamant critics, however,
tools receive wide acceptance and new admit that their knowledge of events at SSA
standards are actually utilized; and is outdated by 12 to 18 months, so that they
● development of new ways of handling con- have no direct knowledge with which to evalu-
flicts between operations and develop- ate SSA’S strong claims of recent progress. The
ment, disagreements among organizational critics’ positions should be viewed in the con-
subunits, and organizational conflict. text of SSA’S statements, summarized above.
Whether these three points represent cur- One of the major themes of critics was that
rent determined efforts, aspirations to be tack- SSA as an organizational culture was incapa-
led at some future time, or merely lip service ble of bringing about the kinds of change rep-
paid to critics, cannot yet be determined. Pri- resented by SMP, because of the hostility of
vately, some SSA observers say that they de- SSA management to newcomers and the fact
pend entirely on the attention and insistence that powerful SSA senior managers are re-
of a few key people and that they began to fade cruited from within, and promoted up the
as soon as it was learned, in early 1986, that ranks, in long insider careers. While this cre-
a change in top leadership and in internal orga- ates loyalty and dedication, it also creates a
nization is to occur. Whether or not this is ac- strong antipathy to criticism, however well
curate, the future strength of these essential meant, and often an inability to learn from it.
conditions will depend in large part on the pol- Critics felt it also creates a culture that does
icies and the capability of the new Commis- not value innovation, and as a result, outside
sioner. consultants and advisors are ignored or avoided,
and internal conflicts are resolved in favor of
those who resist change.
Critics of SSA and SMP SSA as an organization is said by the critics
Many critics of SSA are convinced that SMP to lack a modern, analytical approach to man-
will fail, not because of the technology nor the agement problems. The early decision in SMP
ambitious objectives, but because of SSA’S to salvage 65 to 70 percent of the 10 million
“organizational culture, ” its long history of lines of COBOL code, for instance, never had
mismanagement, interference from outside, po any analytical support, it “was drawn out of
litical pressures, and its sheer size. Those who thin air. ” An SSA contractor complained of
have generally been critical of SSA in the past having “our work ignored. They [SSA man-
are usually skeptical of the possibility of SMP agers] steer us away from important prob-
improving agency performance. Past support- lems. ” Contractors complain of slow decision-
ers of SSA tend to be optimistic about SMP. making, fallback of up to 2 years in the SMP
schedule, and sluggishness because of the sheer
The strong critics include some former size of SSA. As one noted, “there isn’t a club
managers brought into the agency in the late big enough to beat SSA. Below the level of
1970s and early 1980s, who failed in their ef- Commissioners you can’t get an answer from
forts to change information processing at SSA; anyone. ” Many critics describe an alignment
as well as outside observers not associated with of internal interest groups opposed to change.
51

As a plan of action, SMP is widely perceived hiring new young workers from outside. One
to: vendor notes:
● be primarily oriented towards hardware Imagine what it’s like–everyone started
acquisition, and out there and ends up there. Bank systems
fail to provide a vision of how SSP will people come and go, insurance and airline sys-
do business in the future. tems people switch jobs frequently. But not
at SSA. They never get new ideas and proce-
A continuing theme of SSA critics is that dures carried in on the backs of people.
the in-place systems personnel are a principal
impediment to successful implementation of
SMP. A former employee notes that “in-place How Well Has SSA Performed?
systems workers have impeded efforts to re- SSA’S performance in the first years of SMP
form systems and have a stranglehold over new looks considerably more promising than many
projects. of its critics will allow. There remain major hur-
Many SSA employees are critical of the im- dles to be surmounted if success is to be
plementation of SMP. For example, the (AFGE achieved. But the struggles that SSA is hav-
union) Local 1923 Report has carried a num- ing in modernizing its systems are not unique;
ber of stories about the failure of SSA to bring they are similar to problems that other large
workers into decisions related to SMP, and to organizations in both the public and private
require that managers be trained along with sector have had, or are now having.
workers in the new techniques and procedures The history of the Social Security Adminis-
necessary with the modernized systems. The tration illustrates some general principles of
union newsletter of March, 1986, commented: organizational behavior. Organizations do not
. . . in the whole SMP, not a dime has been innovate in areas of strategic importance un-
spent on the process of managing the human less there is some substantial environmental
side of change in (Operations). If the right change; they innovate when they are driven
questions don’t get raised, if the necessary dia- to it by serious and persistent problems or by
logue and consideration of reality and quality crises. In nearly all organizations, there are
are not brought into the process, SSA will substantial forces resisting change, rooted in
never have an adequate system for building prevailing values, norms, and interest groups.
the data processing system on which so many Organizational innovation must involve more
Americans depend.
than adopting new technology. To use it ef-
Leaders in the Local welcome the new em- fectively requires changes in habits, behavior,
phasis on training, but are critical of some of values and norms, and power relationships.
the ways it is being carried out. They claim Technological change nearly always brings
that training opportunities have been deter- fights over who gets, and uses, the technology
mined by generic job type rather than by the to what purposes. Effective managers can take
individual’s needs, that there is little or no op- advantage of external circumstances to solid-
portunity for project teams to be trained to- ify power, disarm internal opposition, and tilt
gether, that there is little or no training in how the internal conflict among groups towards
to manage projects using new technology, and successful use of the technology. Ineffective
that managers have received, at best, only cur- managers may be thwarted by those who
sory training about the new technology and quietly but stubbornly refuse to adapt work
that where such opportunities have been of- processes and procedures to make use of new
fered, managers have been reluctant to attend. technological capability.
On the other hand, outsiders frequently per- The problems that SSA faced and faces in
ceive that SSA is spending too much time and innovating are particularly difficult. Few pri-
resources on retraining employees rather than vate firms have a business environment of com-
52

parable size, complexity, or operational de- by the immense size and complexity of the
mands. The few private firms that have operations. Because it is a government agency,
achieved the level of software sophistication it had little control over changes in its serv-
needed by SSA, or that have successfully in- ices or the volume of its operations, and was
tegrated all elements of their systems devel- not free to take risks in technology invest-
opment, deal in much simpler environments. ments; at the same time, as a government
Large organizations operating in complex envi- agency, it and its managers are insulated
ronments, such as multidivision companies, against the full penalties of failures and of un-
typically have a hodge-podge of systems de- productive behavior, and some of that behavior
veloped at different times by different people is allowed to persist.
and using different languages. This is the case, About some of the basic decisions and strat-
for example, with General Motors, which is try-
egies in the plan itself, there is room for con-
ing to pull together its many data processing
siderable doubt and debate among systems ex-
“baronies” and expects this effort to take a perts. However, for the most part these are
decade to accomplish. Some other large gov- areas where there are no clear and certain
ernment agencies, such as IRS, have, in un- “ r i g h t answers, and almost any decision
dertaking systems modernization, made mis- would have vigorous critics.
takes or suffered problems that for a time
seriously compromised their mission. As will be seen in the case history, in Part
III, some of the greatest hurdles that systems
It seems clear that SSA has been handi-
modernization at SSA face are not deficiencies
capped in undertaking the SMP by the after- in the plan but long-ingrained suspicions and
effects of years of instability or lack of ex- hostility between operations components and
perience in its top layer of leadership, an systems development components, between
organizational culture that emphasizes relia- newcomers and oldtimers, and between career
bility and regularity in daily operations but
people and political appointees, all of whose
resists change and innovation, failure to at-
efforts will be necessary if modernization is to
tract and hold new recruits in some critical succeed.
professional categories, and most importantly

IMPERATIVES FOR SSA


The opportunities for improvement in SSA’S alization of the work flow to accommodate
management of information technology in the changes in processing systems;
next few years would be enhanced by: ● enhanced capability to recruit competent
and well-trained systems designers, man-
● pacing work force reduction to match real agers, and programmers (which probably
gains in technological capability; i.e., depends now on pay and classification
avoiding abrupt reductions that disrupt schedules);
or threaten smooth operations and pro- ● continued funding for SMP itself, and for
duce excessive resistance by workers and SMP-related support functions, such as
managers to further automation; technical and management training;
● a period without major changes in SSA ● strong commitment by top leadership to
programs and adrninistrative responsibil- achieving the goals of SMP and to build-
ities, or, if such changes are mandated, ing a cooperative relationship between
provision of ample time to plan and im- managers and workers;
plement the changes; ● insistence by top leadership on real coop-
● absence of major reorganizations other eration between operations and systems
than those that reflect and support ration- development personnel;
53

● an emphasis on continued strategic plan- recent and current progess in SSA. SSA ex-
ning, and assurance of resources dedicated cessive defensiveness, attempts to deny any
to this activity; and and all failures, and resistance to outside ad-
● early resolution of the issue of independ- vice encourages its critics to suspect and ex-
ent agency status for SSA. pect the worst. In the past, there has been
strong tension between the institutional drive
These desirable conditions imply certain
to secure the resources to make much needed
responsibilities for SSA leadership, for the
changes, and the defensiveness of those peo-
Administration, and for Congress. For all par-
ple who are struggling to cope, not always suc-
ties, they would require placing long-term ob-
cessfully, with day-to-day problems. This has
jectives ahead of the desire for immediate reali-
at times distorted or obscured the picture that
zation of the benefits sought through systems
is presented to Congress. These distortions—
modernization. Congressional oversight will be
whether in the past or in current efforts to re-
most effective if it is directed toward insist-
write history-are now important chiefly to
ing that the agency and its executive branch
alert Congress to the need to probe deeply and
monitors strive to create the necessary condi- target questions carefully in order to assess
tions for progress, rather than focusing on as-
reliably the degree of improvement in service
signment of blame for problems in the past.
delivery. Much improvement is clearly possi-
SSA is changing as SMP is implemented, ble through the use of new information tech-
although these changes may not be quite as nology, and is the best way of justifying the
rapid, nor as deep and smooth, as SSA sug- significant resources invested in SMP from
gests. Most congressional staff people have not 1982 to 1990 and beyond.
had the opportunity to be well informed about
Chapter 3

Systems Modernization and


Related Issues, 1986-90
Contents
Page
SSAand Congress: Accountability and Responsiveness . . . . . . . . . . . . . . . 57
Monitoring and Oversightof SSA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
SSA’s Ability To Implement Changes Mandated by Congress . . . . . . . . 59
SSA and Its Employees: Labor-Management Relations . . . . . . . . . . . . . . . 61
SSA and Its Clients: Issues of Debt Collection
and Financial Management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
SSA and the Administration: Independent Status and Privatization . . . . 66
Possible Independent-Agency Status for SSA. . . . . . . . . . . . . . . . . . . . . . 66
Privatization of or Contracting Out Major SSA Operations . . . . . . . . . . 69
SSA and the Public: Issues of Data Privacy and Security . . . . . . . . . . . . . 71
New Information Policy Directives for the SSA . . . . . . . . . . . . . . . . . . . . 71
SSA’S Data-Sharing Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
SSA’s Computer-Matching Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Future Information Systems and Possibilities for Abuse . . . . . . . . . . . . 76
Security of SSA Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Table
Table No. Page
2. SSA Computer-Matching Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Figure
Figure No. Page
6. Cumulative Percentage Reduction of Full-Time Equivalent
Employees From 1984 to 1990 in the Social Security
Administration Staff by Fiscal Year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Chapter 3
Systems Modernization and
Related Issues, 1986-90

The likelihood of success in systems mod- its own employees, in the context of systems
ernization for the Social Security Administra- modernization. Third, it considers SSA’S re-
tion depends in part on the support of its em- sponse to a major Federal initiative, improved
ployees, its clients, its overseers in Congress, debt collection and financial management,
and other institutions with which it interacts. which significantly affected SSA relationships
To the extent that SSA succeeds in moderniz- with its clients.
ing both its information systems and its man-
Fourth, the chapter discusses SSA relation-
agement, this will change the way the agency
ships with the Administration and with the
does its business, and will affect its relation-
private sector, in terms of possible major
ships with Congress, its clients, its employees,
changes in SSA status, such as making it an
and with other institutions, such as State gov-
independent agency, or privatizing part of SSA
ernment. This chapter explores some of these
operations. Finally, the chapter looks at a
relationships now and in the next 5 years.
growing issue in SSA’S relationships with the
It surveys, first, two issues in SSA rela- general public: concerns about the confiden-
tionship to Congress: the monitoring and over- tiality and security of data as affected by ad-
sight of SSA, and SSA’S ability to respond ef- vanced information technologies and current
fectively to changes mandated by Congress in practices of data-sharing and computer-match-
social security programs, coverage, and bene- ing, capabilities that are likely to be facilitated
fits. Next it considers SSA’S relationships with by systems modernization.

SSA AND CONGRESS: ACCOUNTABILITY AND RESPONSIVENESS


Monitoring and Oversight of S S A ous charges of irregularities and improprieties
in at least one SMP contract award. ’
SMP has already had both positive and neg-
ative impacts on SSA relations with Congress
and the White House. SMP has been regarded
by most Congressmen as good news and Con- 1U.S. Congress, liearings: (’oncra(t Irr(’gularities and .?lis-
gress has responded with generous funding. management Plague SS.4 ‘.s b’.ixten].< ,jlodernization 1@q-an2,
I.egislation and National %curit~r Subcommittee of the I louse
However, there is continuing concern over the Committee on Go\’ernment operations, No\’, 6 and No\T. 20,
wisdom and cost-effectiveness of some of the 1985.
basic SMP decisions, and over SSA’S procure- Recently a General Accounting Office in~estigation alleged
that the Commissioner who initiated SNIP, John S\’ahn, inl-
ment procedures. In addition, there is congres- properly allowed employees of his own former emplo~er. I]eloitt,e
sional concern over whether SMP-related em- Ilaskins & Sells (DHS) to use SSA office space next to his own
ployment reductions and office closings will office for a number of months just when Shl P contracts we’re
being de~’eloped, in which DHS had an interest as potential cont-
result in poorer service to clients. Congres- ractors. S\’ahn was also accused, along with other SS,4 em-
sional oversight committees have been particu- ployees, of improperly accepting restuarant meals from Ikloit te
larly critical of SSA apparent lack of assess- Haskins & Sells during this period. Deloitte Ilaskins & Sells,
a Ilig Eight accounting firm, has since become the largest sub-
ment of the impacts of systems modernization contractor in the Shl P Program and was a major force in rec-
on service levels. Finally, there have been seri- ommending F] 1)S as the major integration contractor.

57

61- 085 - 86 - 3
58

Because of SSA’S size and importance, and getary control). Thus emerging problems like
the large share of Federal expenditures that those of the 1970s can become unmanageable
it administers, a small army of people is com- before Congress is able to come to grips with
mitted to monitoring and auditing SSA to as- them.
sist either Congress or the Administration in
oversight. A significant amount of SSA man- Some political scientists and some computer
agement time is spent in answering detailed enthusiasts have argued that computer tech-
requests for information from oversight bod- nology will facilitate congressional oversight
ies. SMP has added to the volume and com- by making information more readily available,
plexity of these activities. and by allowing Congress to demand reports
tailored to its oversight needs. However, it ap-
There are inherent difficulties involved in pears at least equally likely that computeriza-
congressional oversight of a program like tion of data may make oversight more diffi-
SMP. Several committees have an interest in cult. In the short term, it is very difficult, for
different aspects of it. The House Committee example, to compare SSA’S performance today
on Government Operations maintains a stern with that of several years ago; as work is reor-
eye on information technology procurement ganized and automated, measures of perform-
and other aspects of its management. The ance have necessarily been redefined. More im-
House Ways and Means Subcommittee on portantly, and in the longer term, oversight
Social Security has broad responsibility for becomes more difficult because administrative
administrative performance, but does not have decisions become more highly technical and in-
the technical expertise to evaluate information volve issues of technological capability, multi-
systems and their management. Other com- year investments, and systems management
mittees focus on service delivery, and the in- strategy that laymen—which includes most
terests of special groups in society such as the congressional representatives and their staff—
aged and disabled. find difficult to understand. Seeking and com-
This tends to separate consideration of tech- paring the judgments of technical experts and
nological issues from consideration of service working to comprehend these evaluations is
quality issues. In addition, the critical prob- extremely demanding of time, effort, and at-
lem of software development or procurement tention; it is all the more difficult because sys-
has probably received less attention than other tems experts constitute a highly concentrated
aspects of information technology use and community of people with a great many po-
management. tentially overlapping vested interests in the
actions of SSA, a major purchaser of computer
The difficulty of achieving effective over- systems.
sight is one factor in a growing movement to
split SSA off from the Department of Health The temptation —some would argue, the
and Human Services (DHHS) and make it a duty (given the imperative of administrators
separate, independent agency. (This option will for institutional suMval and maintenance) -to
be examined further below.) Many Congress- select and manipulate data related to organiza-
men and staff people suspect that they do not tional performance when justifying programs
get complete or accurate information from SSA and budgets, is and has always been strong
about its resource needs, particularly on ques- for agency officials. When those budgets in-
tions of its ability to respond effectively to clude multiyear and no-year investments in
changing legislative mandates and changes in equipment for which a favorable return on in-
benefits programs, because the agency’s an- vestment is years away, and for which there
swers must be “vetted” through DHHS and are many irreducible uncertainties in cost-
the Office of Management and Budget (OMB), benefit analysis, that temptation is much
which may manipulate them to suit the Ad- stronger. When the performance data is em-
ministration’s policies and priorities (i.e., bud- bedded in voluminous computerized databases
59

and can be endlessly recategorized, combined, social order and seek to expand social security
and disaggregated by sophisticated manage- into areas such as national health insurance,
ment information systems, it becomes much a wholly nationally administered disability
easier to present a favorable picture—or an un- program, or a Federal program for covering
favorable one, if the object is to demonstrate catastrophic health care of the elderly. These
a need for further modernization of systems. positions are not necessarily related to party
affiliation. Some additional responsibilities
Thus the task of oversight of a huge orga-
have been considered for SSA; for example, a
nization whose mission performance is entirely
role in proposed immigration regulation.
dependent on advanced technology that is
seemingly describable only in esoteric lan- Most national policy makers, however, prob-
guage, becomes much more difficult. ably expect that Social Security will be main-
tained generally in its present form during the
This difficulty is also a problem for agency
next 5 years, with at most some relatively mi-
officials, who must struggle to explain their
nor changes in programs or some realignment
technological resource needs to congressional
of SSA’S various administration responsibili-
committees in ways that do not oversimplify
ties for non-SSA programs. The spectrum of
and distort them and yet do not conceal the
possible changes that might be required of
technological and administrative problems SSA, ranging from no change to radical change
involved in meeting congressional mandates.
in agency status, and their relationship to
Responding to a large volume of oversight in-
SMP. are discussed below.
quiries also reduces the time that administra-
tors can spend in solving problems within their Moratorium on Program Changes
organizations. or Adjustments
One option is to conclude that SSA needs
SSA’S Ability To Implement Changes a breathing spell in its operational and
Mandated by Congress systems-development work. As recently as
September 1985, a GAO report concluded that
Social Security as a national program was SMP software development was not yet im-
born in a period of strong party cleavages over proving SSA’S ability to implement legislative
having such a federally managed function in changes in programs,: although this may no
our society, but over the next five decades, so- longer be true, since SSA has at least reduced
cial security achieved a virtually nonpolitical or eliminated most of its backlogs. It has been
and bipartisan status. Since the late 1970s, suggested that Congress avoid making changes
however, there have been a series of debates for the next 2 years, or until systems modern-
over the size, scope, and organization of So- ization is further advanced.
cial Security. It is likely that this debate will
continue during the next 5 years, both before This option is not likely to be acceptable to
and after the 1988 elections. those who believe that substantive changes are
necessary. As one experienced congressional
Some believe that current budget deficits aide put it:
and economic limitations make it essential to
cut back on the system of Federal retirement, We gave SSA a huge bundle of money for
SMP precisely so that it could handle the
disability, and welfare programs. Suggested changes that Congress is going to make in
solutions range from turning social security basic social programs. We expect the agency
over to the private sector or creating a worker to keep up with us; that’s what ‘ ‘moderniza-
option to select among competing private and tion” is all about.
public retirement plans, to cutting programs
back in scope, benefits, and costs. Others see ‘U.S. Congress, (;eneral Accounting Of ficet Social .!+curit~
..idministration “.5 Comput(’r fi)’stems ,Ifockrnization ~;ffort Ala,}
the Federal program of retirement, disability, \Tot .4chie\e i}lanned objecti~’es, (;A() 1 X! TI;C-85-16, Sept. 30,
and income-support as the hallmark of a just 1 9/+5,
60

Another aide added: delay ongoing redesign of processes or require


further redesign. Each proposed change should
We don’t forego tax reform because IRS
may have computer problems, and we aren’t be carefully studied in advance to determine
going to lose timely opportunities to im- what resources SSA would need to make the
prove social security just because SSA has changes, in the context of already scheduled
a backlog. work force reductions.
It should be noted that SSA has not asked One major program change recently under
for such a legislative moratorium. The agency discussion is that of complete federalization
says that progress with SMP has already sig- of disability programs, instead of the current
nificantly increased its capacity to fulfill legis- arrangement under which States make disabil-
lative directives. ity determinations.;; State determination of
disability (Disability Determination Services,
Program Simplification DDS) shows great variability in quality and
accuracy, in procedures and organizational
Major and minor program simplifications are
structure, and in physician participation. In
possible that could make both computer and
the recent effort to purge disability rolls (see
field operations easier; for example, simplifi-
below) some States refused to do reexamina-
cation of the formulas for recomputing bene-
tion under SSA guidelines. GAO has advised
fits or changes in the earnings test for eligibil-
the Congress’ that:
ity. SSA has been working for several years
on concepts for formula readjustments to sim- From a purely operational perspective, a to-
plify benefits calculations, but is not ready to federal structure for disability determi-
tally
suggest them. One problem is that they might nation appears to be the preferred option.
require compensatory or transition payments It would give SSA direct control and account-
to soften the losses to various categories of ben- ability; eliminate State political influence; pro-
eficiaries. Proposals for program simplification vide greater organizational uniformity; assure
changes may however surface in the next 5 standardized salary and qualifications for per-
years. sonnel; eliminate the time spent in negotiat-
ing with States on compliance; allow closer
Program Modifications working relationships between district offices
Several congressional and administration and determination units; and allow SSA to se-
sources provided a “shopping list of program lect the number, location, and size of offices.
modifications that various interest groups or
GAO has advised Congress that federaliza-
Members would like to see enacted. These in-
tion of determinations would be likely to add
cluded restoring eliminated benefits to student
a large number of employees to Federal rolls
dependents of deceased, retired, or disabled
(11,000, according to GAO assumptions about
workers; expanding retirement coverage to
productivity). It could also cause the loss of
State and local employees; including partial
some trained and experienced examiners who
disability under SSA coverage or expanding chose not to work for the Federal Government,
rehabilitative or work-reconnection efforts; ad-
and would make the determination process vul-
dressing women’s equity problems through
nerable to Federal hiring freezes or other bud-
measures such as earning-sharing between hus-
getary measures. Claims processing might be
band and wife; and correcting the “notch” or
‘‘inequity’ problem that arose between bene-
ficiaries born pre- and post-1916, as an un- ‘f Under the Disability Insurance Program of 1954 and the
Supplemental Security Income Program of 1972 Congress man-
anticipated consequence of formula changes dated State responsibility for determinations of disability, with
made by the 1977 amendments. Such new or oversight by SSA.
expanded programs would produce a tempo- “U.S, Congress, General Accounting Office, Current Status
of the Federaf/ State Arrangement for Administering the So-
rary burst of additional work to make neces- cia) Security Disability Programs, Report to the Honorable Jim
sary changes in benefits formulas, and might Sasser, U.S. Senate, HRD-85-71, Sept. 30, 1985.
61

disrupted during the changeover period, and Non-Social Security Program Developments
a new policy and system for purchasing medic-
SSA could be asked to take over adminis-
al services might have to be developed.
trative responsibilities for new non-SSA pro-
GAO did not address the possibility of in- grams, as has happened repeatedly during its
corporating the determination process into ex- history. Under national immigrati~n reform,
isting SSA field offices, rather than maintain- for example, employer access to SSA for veri-
ing separate facilities; thus it did not speak fication of job applicant identities could be
explicitly about the effects of federalization of mandated. lf SSA were given this role, there
the process on the level of demand on SSA com- would be pressure to enhance the accuracy of
puter and telecommunication systerns, or the SSA records, such as matching accounts with
effect of these systems capabilities on produc- death records to detect invalid accounts, and
tivity of determination examiners and support identifying accounts used by more than one
personnel. These questions would have to be person, This could represent a significant vol-
addressed in further analysis of the effects of ume of additional work for SSA, especially
this program change on SSA technological and without an integrated database in place; it
personnel resources, and on the quality of fu- would probably require the development of en-
ture disability determination services. tirely new software systems.

SSA AND ITS EMPLOYEES: LABOR-MANAGEMENT RELATIONS


SSA began its SMP with hostile labor rela- vendors and those who would compete for
tions, in large part due to the deteriorating awards if SSA operations were contracted out,
working conditions and heavy overtime de- and the interests of those who depend on SSA’S
mands of the 1970s. In the early 1980s labor services—the beneficiaries.
and management refused to negotiate a con-
tract for 18 months, and ultimately accepted The relationship between the union and SSA
some compromises (December 1981) only with is buffeted by the maneuvering of all of these
great bitterness on both sides. Since then, parties. OMB pressure on SSA to drop 17,000
the union has filed up to 800 unfair labor prac- employees over 6 years, and to privatize oper-
tice charges each year. Until 1983, the labor ations equivalent to 8,600 jobs, as discussed
relations management of SSA would not even below, are good examples. As SSA managers
call the union, for fear of being misquoted or readily acknowledge, in the recent past, only
maligned. the extraordinary efforts and commitment of
SSA workers have allowed the agency to sur-
Both SSA and its union agree that SMP will
mount repeated crises in its operations. But
lead to new levels of productivity. The ques-
SSA must of course respond to the Adminis-
tion is whether this will be used to enhance
tration and Congress as they look for a return
service levels, improve the quality of worklife,
on what will by 1990 be the billion dollar in-
and raise the skill levels of workers; or whether
vestment in SMP. Under these circumstances
the productivity gains will be used solely to
the management is under great tension, and
reduce the size of the work force, speed up
many employees are resentful and suspicious.
work, and lower skills requirements and status
of jobs.
A union official estimated as early as mid-
This debate is not merely between SSA and 1984, that SSA workers were facing a net re-
the union. Also involved are Administration duction of 10,000 field office jobs, one-third of
policies, congressional interests, the stakes this work force, with virtual elimination of
that other unions have in office automation is- the position of data review technician and
sues, the interests of SSA’S contractors and changes in the claims representative job (some
62

managers were calling for its elimination and Under the Civil Service Reform Act (CSRA),
replacement by clerical staff).s management retains the right to introduce new
These expectations proved justified; the net technology and to change jobs and work meth-
ods. The union cannot force SSA to bargain
job loss in 1985, as reported by SSA field
on technology adoption or work standards. But
offices to GAO, was 949 full-time equivalent
the CSRA does require management to an-
jobs, or 2.4 percent of all 1984 jobs’ (see fig-
nounce its plans and give the union an oppor-
ure 6). This included 297 data review techni-
tunity to bargain over the means and condi-
cians, 275 clerical positions, 329 claims repre-
tions of proposed changes. The union can force
sentatives, 86 service representatives, and 140
management to pay attention to working con-
nonceiling employees or other positions. This
ditions, health and safety concerns, retraining,
was a total of 1,127 jobs eliminated, but 178
skill levels, and job classification.
“other positions” were created, including 123
joint data review technician/service represent- Some observers say that this gives the union
ative positions. a way to slow down, impede, and even prevent
SMP from proceeding if it so chose, at least
On the other hand, it was also predicted that
long enough to stir up the ire of Congress and
many of the 1,386 SSA field offices would be
the public and bring the whole project down.
closed. SSA is reviewing the status of these
On the other hand, the union has not opposed
offices, but as of February 1986, the 228 re-
new technology nor does it want SSA to fail.
views that had been conducted had not re-
Workers have generally not complained about
sulted in any closings.
the advent of new technology; rather, they
complained about the terrible workloads im-
“According to a letter from John Harris, Special .Assistant posed by new programs for which the agency
to the National President of the American Federation of Gov-
ernment Employees (A FGE), July 1984. was unprepared, the lack of technology with
“U.S. Congress, General Accounting Office, Social Securit~: which to handle this workload, and the de-
Actions and Plans To Reduce Agency Staff, briefing report to mands on workers to work overtime.
congressional requesters, HRD-86-76BR, March 1986.
These tensions led management and labor
Figure 6.—Cumulative Percentage Reduction of to try a new approach, in the common recog-
Full-Time Equivalent Employees From 1984’ to 1990 nition that both union and management need
in the Social Security Administration Staff by to make SMP a success. In 1985 SSA and the
Fiscal Year union reached an unusual agreement, which
mirrors the recent agreement between the
UAW and General Motors in GM’s Saturn Car
Division in Tennessee. The similarity is more
than superficial; key advisors to SSA and the
American Federation of Government Employ-
ees were also key advisors to UAW and GM.T
“The Joint Statement of Common Purpose, ”
was signed at SSA in September 1985. Its ob-
jectives are to avoid the degradation of work,
to enhance the quality of working life, and to
create a three-tiered management-labor struc-
ture for future shared decisions. It explicitly
avoids trying to change the statutory require-
1985 1986 1987 1988 1989
ments of the relationship between manage-
Fiscal year

a ln 1964 there were 83,588 employees in SSA ‘They were: Irving Bluestone, retired Vice President of
SOURCE U S Congress, General Accounting Off Ice, Social Security Actions and UAW; and Dutch Landon, retired Quality of Work Life Direc-
Plans To Reduce Agency Staff, USGAO/HRD.8(-76BR, March 1986 tor at General Motors.
63

ment and labor, although the CSRA did not cially designed for comfortable support while
envisage labor-management cooperation at working), and related workplace issues. Each
this level. In other words, the agreement will of these projects was to be developed
permit both collective bargaining and cooper- project team with links to working teams of
ation of a new kind. management and labor at the operating level
and to make recommendations to the policy
As the union president described the
committee.
agreement:
Most observers feel that the success of this
. . . both sides put away “business as usual”
and go into a partnership, a joint action based agreement is essential to carrying out SM P.
on common interests and objectives, to look But in spite of the agreement, the union ex-
at what is going to happen to the workplace, pects “displacement and disruption to be the
the work, and the worker as this automation norm in the implementation of SM P. A union
is brought about. . . . [W’]here collective bar- official notes that:
gaining is the best remedy to the problem, we
shall do it. But we will seek in the main to sol~’e A rupture of the work force such as widespread
our problems together as co-equals and not as job loss or reassignment can be avoided. But
adversaries. only if a comprehensive program is adopted
tore-design field offices, one which starts with
The three-tiered structure consists of an ex- the premise that all workers will be given use-
ecutive committee level (which will include the ful jobs with similar skills or will be retrained
SSA Commissioner and the head of AFGE, and no one will be laid-off or downgraded. With
a project level, and a workplace level. such a program the phasing in of automation
will be conducted with the worker in mind, not
The Joint Policy Committee agreed on the as an after-thought. . . . This is the greatest
following guidelines: challenge to the union and management be-
cause it puts both into a new relationship at
the process (development, implementa- a time when neither trusts the other. 9 (Em-
tion, and oversight) will be joint and phasis added.)
co-equal,
employee participation at the workplace It is clear, however, that this objective, in-
level will be completely voluntary, terpeted literally, conflicts directly with the
innovations that result from the joint objective of reducing the work force to justify
process will not result in the loss of job investment in information technology. The
or pay of any employee, joint agreement could in theory provide a
the joint process is independent of the la- mechanism for compromise on this issue while
bor agreement and is not a replacement cooperatively working toward other goals such
for collective bargaining or the grievance as improved quality of the workplace. But by
procedure, May of 1986 the joint agreement appeared to
training and resources will be provided, be breaking down. According to workers, an-
the joint process will not be used as a bar- nouncement of appointment of a new commis-
gaining chip, and sioner weakened the influence of managers who
either part y may withdraw from the joint supported the mechanisms and thereafter
process. there were no meetings of the committees.
Union members believe that the appointment
The policy committee chose three projects signals a new determination by OMB to force
to work on immediately; including the effects drastic job eliminations, and they charge SSA
of the claims modernization process, issues re- managers with “passive acquiescence. ”
lated to use of visual display terminals (VDTS),
ergonomic furniture (i.e., desks and chairs espe-

“Letter of Kenneth Blaylock, President, to the union locals,


,June 3, 1985, quoted by permission. “Harris, letter, op. cit.
64

SSA AND ITS CLIENTS: ISSUES OF DEBT COLLECTION


AND FINANCIAL MANAGEMENT
SSA’S relationships with its clients, and its Under Public Law 96-265, Social Security
public image, have been adversely affected by Disability Amendments of 1980, the Secretary
its response to the government initiative for of HHS was required to review the status of
improved debt collection and financial man- all nonpermanently disabled DI beneficiaries
agement. As information technology allows the every 3 years, beginning in 1982. Until then
agency to become more efficient in this area, SSA had reviewed only a small percentage
more judicious management techniques will be (about 150,000) each year, primarily those ex-
necessary to avoid unnecessarily eroding the pected to recover from their disability and
trust that beneficiaries still have in the those voluntarily reporting either improve-
agency’s operations. ment or gainful employment. But GAO had
estimated that as many as 20 percent of those
In 1981, the President ordered tough enforce-
on the rolls might not meet the legal defini-
ment of the Disability Amendments Act of
tion of disability .10 The Administration there-
1980, which led to summary termination of
fore ordered stringent actions to purge the
over 1 million disability beneficiaries, causing
rolls.
a huge backlog of work for SSA. Rigorous en-
forcement, by the Administration, of this act In order to spread the workload on the States
and the later Debt Collection Act of 1982 sub- (which make the original disability determina-
jected SSA to bitter criticism in the press and tions), SSA began implementing the reviews
among its constituents and traditional sup- 9 months earlier than the statute required. Of
porters. Continuing and future efforts to im- 1.2 million cases reviewed, 500,000 benefici-
prove debt collection and financial manage- aries were summarily dropped from the rolls.
ment, and reduce fraud and waste, are likely This brought about a flood of protests and ap-
to be affected by the resentment that resulted peals, which only increased when 200,000 of
from this initiative. the 500,000 were reinstated by appeal to
administrative law judges, the first level of ap-
During this period the political climate for peal. Many congressional hearings were held
SSA was complicated by the fact that the two to consider these developments. ’ 1
Houses of Congress were controlled by differ-
ent parties, and thus oversight committees em- Those who had been dropped from the rolls
phasized somewhat different priorities and stopped receiving benefits, until Congress
directives. Members of some oversight com- passed stopgap legislation in 1982 (Public Law
mittees were pressing for greater assurance 97-455) to allow them to continue receiving ben-
that service levels would be improved as a efits while they appealed. About two-thirds of
justification for investment in systems mod- those who had been dropped from the DI rolls
ernization. Members of other committees were eventually reinstated. The courts, and the
wanted greater assurance that no effort was
being spared to reduce costs. Members of both “’l’or background, see Social Security Administration, office
of I,egislative and Regulatory PolicJr, Social 5’ecurit~~ Disabil-
parties and both Houses emphasized the need it~’ Amendments of 1980: I.egislati\’e liistor.~’ and Summar.\’
for better management, greater efficiency, and of Pro}risions, Social Security Bulletin, April 1981; and “Social
strict accountability. These pressures affected Security Disability Benefits Reform Act of 1984: I.egislative
llistor~’ and Summary of Pro\’ isions, ” Social Security Bulle-
SSA’S response to the President’s initiative, tin, April 1985 (both listed as SSA Pub. No. 13-1 1700); HHS
at a policy level; at the operational level, there News Release of Apr. 13, 1984, no title; and ,Vew }’ork Times,
were further difficulties. While the Disability Dec. 6, 1985.
11 Most recendy: U.S. Congress, Go\rernment Representa-
and Debt Collection Acts were increasing the ti~’es: Ad\zwates or .Ad}’ersaries:) I Iearing Before the Iiouse Se-
workload, a hiring freeze was imposed on SSA, lect Committee on Aging, 99th Cong., 1st sess., hlar. 18, 1985,
as well as other agencies, in 1982. Roston, MA.
65

States, raised serious concerns about the cri- mits termination of Disability Insurance ben-
teria used for “medical improvement, ” and efits only if there is ‘‘substantial evidence of
especially the criteria used in mental impair- medical improvement sufficient to allow the
ment cases and in evaluation of pain. The beneficiary ‘‘substantial gainful activity, ” or
Administration adopted a policy of “nonac- new medical evidence that vocational therapy
quiescence” in certain cases; in other words, or technology makes him or her able to work,
SSA would not apply court decisions about its or that the original impairment was not as dis-
criteria and procedures in other judicial clis- abling as it was originally considered, or the
tricts but defended its pradices district by dis- original determination of eligibility was in
trict, case by case. (This policy was rescinded error. 13
in early 1985. )
In 1983, similarly tough enforcement of the
By 1984, the disability review process had Debt Collection Act led to withholding all so-
all but collapsed, with half of the States either cial security payments to beneficiaries Who had
refusing to administer the reviews or under received overpayments, as opposecl to the ac-
court order not to do so. In April, HHS Secre- customed procedure of withholding no more
tary Margaret Heckler ordered suspension of than 25 percent of benefits until overpayments
the disability reviews ‘until new disability leg- were repaid. In addition, the U.S. Treasury
islation is enacted and can be effectively im- used direct electronic debiting of beneficiary
plemented. ” She also ordered SSA to resume bank accounts with no prior notice (’ ‘Treasury
benefit payments to those in the process of ap- recovery’ ‘). This could seriously jeopardize re-
pealing. cipients with no other resources.
SSA had suffered a severe blow to its esteem These actions kept telephones ringing in con-
with the public. An internal SSA memo ac- gressional offices as beneficiaries complained,
knowledged that “the agency’s credibility be- and the flood of inquiries and protests to SSA
fore the Federal courts is at an all-time low. ” district offices resulted in reduced attention
The official SSA position is that the harshness to servicing other clients. It also caused dis-
of its administration of the amendments was tortion in SSA management behavior, because
inadvertent and a startup problem; it says: local administrators were given pay raises or
promotions based on the amount of debt they
. . . a great many admii~istrative changes were
collected.
made beginning in 1982 to deal with these criti-
cisms. Thus the disability legislation as finally The controversy over these enforcement pro-
enacted, in 1984, reflects, in part, the evolu- cedures appears to have added to the fierce-
tion of the CDIl administrative process since ness of the controversy over systems modern-
1981.11 ization, even though there is little logical
The congressional response to the problem relationship between the two. Critics repeat-
was the Social Security Disability Benefits Re- edly point to these episodes as illustrating a
form Act of 1984 (Public Law 98-460). It per- commitment to efficiency at the cost of socially
desirable service to the public.

] ‘Social Securitjr Administration, office of IJegislati\’e and
Regulator I’olic>r, ‘‘Social SecuritJr I)isahilitJ Benefits Reform
Act of 1 9/+4: I.egislati\re I ii~torJ and SurnrnarJ of Pro\rision\,
Social Security Bulletin, April 1985, SS:\ I>utj, N(). 13-11 i’{l(J.
66

SSA AND THE ADMINISTRATION:


INDEPENDENT STATUS AND PRIVATIZATION
Possible Independent-Agency Status Hearings were held on H.R. 825, a bill to
for SSA make social security “off budget” and place
it within an independent agency, in Septem-
On July 22, 1986, the House of Representa- ber 1985.]’) (The Social Security Trust Fund
tives voted 401-0 to make SSA an independ- has since been moved off budget. ) Advocates
ent agency, as it was in its first few years of of an independent SSA argued that independ-
existence ( H.R. 5050). This bill was referred ence would help shield SSA from the full force
to the Senate Finance Committee 2 days later. of OMB demands for a cutback and help it
Making SSA an independent agency with resist demands for excessive contracting out
only the core functions of retirement, disabil- of work. Some hoped that the threat of such
ity, SSI, and possibly Medicare was recom- legislation would itself soften OMB pressure,
mended by the National Commission on Social since removing SSA from DHHS would take
Security in 1981. The National Commission on away about 60 percent of DHHS’S budget and
staff and leave some social programs related
Social Security Reform, in 1983, called for a
to core SSA functions in DHHS without co-
congressional study of how this could be ac-
herent administration.
complished, and a panel headed by former
Comptroller-General Elmer Staats conducted An “independent SSA” bill with 165 cospon-
a study of this recommendation for Congress. sors was reported out by the House Ways and
Means Committee and unanimously passed by
In June 1984, the panel outlined a design for
the House in late July of 1986. (The measure
a new Social Security agency, which would
is now before the Senate Finance Committee. )
have SSA headed by a single administrator ap-
This Budget and Administrative Reorganiza-
pointed for 4 years, with a nine-member bipar-
tion Act differs only slightly from the Staats
tisan advisory board, The administrator and
Panel recommendations. It would separate
board would have greatly strengthened man-
SSA from DHHS; the agency would be gov-
agement authority, including delegated au-
erned by a three-member Board, nominated by
thority over personnel, facilities, and computer
the President and confirmed by the Senate. The
systems.
boar-d would be responsible for the Trust Fund,
Hearings on the Staats Plan were held in make budget recommendations to Congress,
July 1984. Support for the plan came from and make policy recommendations to Congress
some influential members of Congress, AFGE and the President. The members of the board
union leaders, SSA local and regional office would serve staggered 6-year terms, and no
managers and pro-social-security interest more than two could be of the same political
groups. Opposition was registered by Acting part y.
Commissioner Martha McSteen’ and former
There would be a similarly appointed Com-
SSA Commissioner Ross. In late 1984 the in-
missioner as chief operating officer, who would
dependent agency proposal appeared unlikely
serve a 5-year term, and who would be specifi-
to pass. But unexpected political impetus for
cally charged with developing and implement-
the proposal arose in the House in the sum-
ing a long-range plan for advanced automated
mer of 1985, in reaction to the Administra-
data-processing systems. There would also be
tion’s proposed reduction in the SSA work
an Inspector General, and a Public Ombuds-
force and the closing of some local offices.
man to represent client/beneficiary interests.
-. ——
]‘[1. S, Congress, Back LO Basics: Social Securit.}. off-!lud~y~t
‘ ‘Former 11 I+;\fr S~~c’retar?r W’ilhur Cohen cx)n~n~ented that hc~r and Independent, 1 Iearing I]eforc tht’ I I(JUSC’ Select (’ommittee
position was “‘that of ()\l B, not necessaril}r her own. on ~lging. 99th (Tong., 1 st sess., Sc’pt. 9, 19H5,
As the Staats Panel recommended, the pro- The strongest motivation for some sup-
posed SSA would (initially for an 18-month porters of independent status for SSA is their
demonstration period) have broad delegated suspicion that information about SSA resource
authority over personnel management, facil- needs, progress in modernization, and abilit~~
ities management, and ADP contracting and to carry out congressional mandates, is filtered
management. SSA would carry out only its pri- through executive branch agencies that want
mary programs: old age, survivors and disabil- to justify budget cuts, possibly at the cost of
ity insurance, and supplemental security in- reduced services. They argue that independ-
come programs. ent status would make possible more effecti~~e
congressional oversight,
Opponents of independent-agency status for
SSA say that it is unnecessary since Congress Meanwhile, the whole concept of independ-
has now helped clarify SSA responsibilities and ent agencies has come into renewed dispute
provided solid appropriations for SMP. SSA, as an indirect result of the Gramm-Rudman-
they argue, needs a period to consolidate or- Hollings Act and the February 7, 1986, ruling
ganizational changes, provide personnel sta- by the U.S. Court of Appeals. The Court struck
bility, and restore the confidence of benefici- down key provisions of the act on the grounds
aries and account holders in SSA services, that:
Cutting Medicare and Medicaid loose while ac-
. . . the powers conferred upon the Comptrol-
cess to SSA records remains vital to determi- ler General . . . are executive powers, which
nations of eligibility, would be disruptive. Tak- cannot constitutionally be exercised by an of-
ing SSA out of DHHS, according to opponents, ficer removable by Congress. . . .
would:
This has been interpeted by some commenta-
1. remove policy coherence for the different tors as applying to independent agencies, par-
Federal social-welfare programs; ticularly since the court observed in passing
2. deprive SSA of representation and ad- that:
vocacy within the Cabinet; and
3. by removing essential oversight from It is not as obvious today as it seemed in
DHHS and the General Services Admin- the 1930’s that there can be such things as
istration (GSA), potentially allow SSA to genuinely “independent” regulatory agencies,
bodies of impartial experts whose independ-
drift back to its old “hardware orien- ence from the President does not entail cor-
tation. respondingly greater dependence upon the
Supporters see independent status as a committees of Congress to which they are then
means of recognizing social security’s special immediately accountable; or indeed that the
status as a trust program, and giving SSA decisions of such agencies so clearly involve
scientific judgment rather than political choice
management freedom from alleged DHHS in- that it is even theoretically desirable to insu-
terference, GSA neglect, and OMB constraints late them from the political process.16
that do not accord with congressional priori-
ties. With “extraneous” social welfare pro-
grams removed, SSA would be able to concen- 1[’[Jnited States District (’our-t for tht’ I )istrict of (’f)lumhia,
order filed Feb. ’7, 1986, in Ci\il Action \ () S,-)-l!j 15 ant] (~t h~~rs,
trate on its major programs the professional p. 40.
resources that have frequently been tapped to The first independent agencj’ was th[’ I nterstatt> (Tt)mmercw
support “non-Social Security programs. ” A Commission, in 1887; since then a number of agt~n[i(~~, primar-
ily of a regulatory nature, have been created with this status.
bipartisan board could concentrate on long- Although executive branch agencies, they report to I)oth the
range planning, policy development, and liai- President and Congress, and their heads serve fixed terms, and
son with Congress and the executive branch, can IN rcrno~ecf onl}! ‘‘for wrongdoing, Their constit ution:ll -
i t } wa~ uph[~ld in 19;1;1 (11 umphr(’~’ 1+; xe(.utor]. The ( )fficw of
while the Commissioner concentrates on ad- IA~g:Il Counsel, in thc~ I )[~part mtjnt of tJusticw, has se~era] t irne+
ministration and information systems. rai s[d t ht~ is SU[’ of L he cons L i tu t iorral i t J. of independent t ag(~n-
(contlnued on next page)
68

The District Court did not say that all in- Secondly, Congress would not necessarily be
dependent agencies whose heads have fixed assured of better information about SSA in-
terms and are not removable by the President formation technology management, since ex-
were unconstitutional. It based its ruling on ecutive branch constraints on SSA statements
the fact that the law governing the Comptrol- to Congress have not been the only factor in
ler General’s removal from office before the ex- oversight problems, as already discussed.
piration of his fixed term says that it may be
It is clear from SSA’S recent history that
done by joint resolution of Congress (for cer-
the extreme instability of leadership during the
tain listed causes), as well as by impeachment,
1970s contributed greatly to SSA’S difficul-
which applies to all U.S. officials. Neverthe-
ties in solving its internal problems; it is less
less, this ruling, if confirmed by the Supreme
clear that frequent changes in leadership (and
Court, will probably strengthen the opposition
the frequent reorganizations related to them)
to independent status for the Social Security
caused the problems. They could be viewed,
Administration, a nonregulatory agency which
alternatively, as unsuccessful efforts to solve
clearly performs executive (administrative)
those problems. At least as strong a case can
duties. (Under H.R. 5050 both the Social Secu- be made that the long stability, insularity, and
rity Board and the Commissioner are to serve
defensiveness of SSA’S middle and upper man-
for fixed terms and cannot be removed by the agement caused SSA to fall behind in meet-
President. )
ing the technological imperatives with which
Regardless of the outcome of this issue, the all large data-handling organizations were
questions will remain as to: struggling.
1. the justification for giving independent It is also clear from SSA recent history that
status to SSA, it has suffered from conflicts in priorities, if
2. whether this would make congressional not policies, set by the Administration on the
oversight more or less difficult, and one hand and Congress on the other; and to
3. whether it would solve basic management some extent from conflicts in priorities of the
problems within SSA. various oversight committees. This is however
a problem that is inherent in our form of gov-
Independent-agency status would not solve
ernment (indeed, was deliberately built into our
the problems associated with systems modern-
Constitution), and it becomes acute for nearly
ization and congressional oversight. First,
every agency at some time or other. Independ-
some factors or constraints would remain, or
ent status cannot be practical as a general so-
be only partly removed. OMB would still ex- lution, and in each specific instance it carries
ercise oversight on behalf of Presidential pol-
with it the risk of introducing unnecessary in-
icies. Recruitment of expert staff would still
coherence and irrational variation in policy for-
depend on improving the professional climate
mulation and administrative procedures. SSA
for programmers and systems staff, and as
may also have suffered from lack of a strong
civil servants they would still be subject to
direct voice in Administration policymaking
Federal pay scales. Legislation designed to as-
(since DHHS must speak for many disparate
sure competitiveness in procurement would
and quasi-independent components); independ-
still apply. ent status would not solve this problem but
——.—.—- —
(continued from previous page~ instead would worsen it.
Theodore olson, who headed this office during President
cies. Finally, at times, SSA communications to
Reagan ”s first term, has filed suit challenging the constitution- Congress about problems or resource needs
ality of the Federal Trade Commission. These challenges turn
on the point of the constitutional principle of separation of were constrained by considerations of Ad-
powers and the President’s inability to dismiss the heads of ministration policy and political initiatives.
independent regulatory agencies although they are within the
executive branch. 1 n 1976 the Supreme Court ruled that ~he
However, this has not been the sole source of
composition of the Federal Ijlections Commission was unc(Jn- oversight problems. SSA defensiveness and
stitutional for this reason. fragmented congressional oversight responsi-
69

bility have also played a part, as have the in- The size of the social security programs is
herent uncertainties in technological develop- also a concern. Relati\~ely few contractors
ment. Under these circumstances, it seems might be in a position to successfully deli~er
that independent status for SSA would not in systems/capabilities of this magnitude; anci
itself greatly facilitate the oversight process. therefore the level of competition might be low.
However, Congress may conclude that re- The union that represents SSA workers,
duction of SSA work force andior closing of AFGE, is of course bitterly opposed to the con-
field offices at this stage of systems modern- cept and is calling on labor organizations to
ization would degrade service to clients to an oppose any such ‘despoiling’ of the public so-
unacceptable degree, or would cause the re- cial security system. Adding further to the
appearance of the problems of the late 1970s, strong perception of job insecurity}’ would fur-
or would render SSA unable to respond satis- ther erode morale among SSA employees and
factorily to congressional mandates–any of increase management problems.
which outcomes would discredit the Systems
Modernization Plan and discourage further ef- There is a broader concern over whether the
forts to carry it to completion. In that case, competence and the commitment of SSA
independent status would be a more attractive workers can be matched by those of contrac-
option. tor organizations. The valuable experience that
SSA workers have built up over many years
Privatization of or Contracting Out has often been the saving grace that allowed
Major SSA Operations the agency to cope with a suddenl~ expanded
volume of data processing or repeated systems
The Administration is currently pressing ex- failures. In a crisis, SSA often calls on loyalty
ecutive agencies to implement OMB‘s policy and dedication over and above the call of duty
directive, Circular A-76, instructing agencies to get the work done, and dissal of these
to contract to private sector organizations Federal workers now would be unwelcome to
those Federal operations that could be done many in Congress, as well as to many S.SA
more cheaply outside of government. 1 n a managers.
memorandum dated ,July 25, 1985, DHHS di-
rected SSA to develop a plan to contract out There are major concerns about the wisdom
the equivalent of 8,600 full-time positions. and long-term effects of having an essential
Those under consideration include the process- and highly visible Federal function such as
ing of Annual Wage Reporting now done at administering the SSA database in private
SSA Data operation Centers; the filing and hands. Turning over sensitive and privacy act
mail work done in handling SSI folders: and protected records on 160 million Americans to
the operations of SSA’S National Computer a private contractor would probably be shar-
Center where the central beneficiary records ply resisted by bipartisan groups in Congress
are maintained. who see the social security system as a public
trust and would not trust these records in pri-
There are serious management issues to be vate hands. SSA is responsible for a signifi-
considered. It is likely that much time and cant fraction of all Federal expenditures—
money would be needed for a private firm to roughly 15 percent. The question of public
learn the operations and functional require- trust in the accountability of the Administra-
ments of the SSA system. Particularly with tion of these expenditures must be considered,
functional requirements still poorly defined, as well as the quality of services that can be
computer services firms would incur a signifi- assured, when the temptations of for-profit
cant risk in bidding without assured funding operations are combined with the possibilities
up front for startup operations. Government for fraud inherent in government contracting.
contracts must be recompleted regularly, and
any change in contractors would mean an ad- On the other hand, OMB and GAO have
ditional learning period. found that in many cases privatuzatition of gov-
70

ernment services results in significant savings would also probably cost less than alternatives
to the taxpayer and/or improved services. (either the current arrangements, or complete
OMB Circular A-76 requires that a function federalization), if the productivity levels of the
that is not inherently governmental must be 10 most productive State DDS organizations
put into a description capable of being bid on were assumed to apply. However if the aver-
by private companies. In some cases, the gov- age State productivity now is assumed to ap-
ernment agency is able to show that its costs ply, personnel costs would be $13 million
for providing the services is as low as, or lower higher than current costs.
than, those in the private sector, in which case
In addition, GAO pointed to some disadvan-
the services are not contracted out. This ne-
tages: the time necessary to get contracts
cessity has provided a new and powerful in-
planned, awarded, and operational, the possi-
centive for government agencies to make their
ble loss of expertise developed by (current)
operations cost-effective.
DDS examiners, and possible disruption of
SSA’S management maintains that apply- claims processing during the changeover. Fi-
ing Circular A-76 to SSA operations will not nally, GAO noted that there is a potential con-
necessarily result in contracting out these serv- flict of interest if a contractor also administers
ices, because the systems modernization has, private disability plans tied to SSA determi-
or will eventually, make the agency’s perform- nations; and that it may be difficult to find
ance so highly efficient that SSA could become competent contractors who are not already ad-
the lowest possible bidder. SSA officials pro- ministering such plans. If more than one con-
fess not to believe that they would lose a com- tractor were involved—for example, a differ-
petition for carrying out their data center, pro- ent one for each State—there would inevitably
gram service center or National Computer be disparities in costs and quality of perform-
Center operations, and thus do not see the re- ance. Further, the necessity of recompleting the
quirement of conducting an A-76 exercise as contract periodically would imply recurring
leading inevitably to contracting out. Some ob- periods of potential discontinuity, disruption,
servers, however, fear that some companies in changes in procedures and very likely in qual-
the private sector, paying low wages and anx- ity, and investment in contractor learning and
ious to get SSA operations as a high-visibility experience.
advertisement, would underbid SSA.
GAO did not, in this report, address the
It has recently been proposed that determi- question of whether the level of competition
nation of disability status, which is now done for such contracts would be adequate to as-
not by SSA directly but by the States, be priva- sure high performance and achievement of
tized. GAO found that privatization of Disabil- other congressional objectives, although the
it y Determination Services (DDS) would make GAO report did ask, but did not attempt to
the determination process less vulnerable to answer: “Are there enough private entities able
budgetary restrictions and hiring freezes, to process the disability cases?” GAO also
would improve Federal control and eliminate raised but did not discuss the significant pol-
State political and governmental influences, icy question: “Should a major federal program
and provide greater flexibility in selection of with a very complicated process and the obli-
location and size of offices.17 These functions gation to pay about $23.5 billion a year in ben-
—— —
efits, be operated by the private sector?’
‘Tsee U.S. Congress, General Accounting office, Current Sta-
tus of the Federal/State Arrangement for Administering the —
Social Securit.v Disabilit~’ Programs, Report to the Honorable of accuracy, medical consultative examination procedures, ph~’-
Jim Sasser, U.S. Senate, HRD-85-71, Sept. 30, 1985. In estab- sician participation, employee standards and salaries, etc. Dur-
lishing the Disability Insurance Program, in 1954, and the Sup- ing the initiative to purge disability rolls (described above, also
plemental Security Income Program, in 1972, Congress provided see ch.7), some States refused to cooperate. The Social Secu-
that the States should make determinations related to disabil- rity Act of 1980 strengthened SSA control and oversight of DDS.
ity (Disability Determination Ser\’ices, or DDS). There has al- There have subsequently been proposals both to fully federal-
ways been great variability among the States in terms of rates ize it and to pri~’atize it.
These issues argue that the question of seen as a matter of social policy, rather than
whether SSA operations should remain in the as a narrow question of competitive bids and
public service or he contracted out should be cost-effectiveness.

SSA AND THE PUBLIC: ISSUES OF


DATA PRIVACY AND SECURITY
SNP P has not had had a direct effect on You have a choice about signing t his form. But
privacy or on freedom of information, but it we must have accurate in format ion about -your
raises many issues for the immediate future, income and what you own to pay your Supple-
mental Security Income check. If you do not sign
and exacerbates some older issues. Congress
the form, your Supplemental Security Income
and the Administration are currently em-
Checks may be affected.
phasizing the efficient collection and sharing
of information to reduce fraud and waste. SSA The provision of this information, under the
accordingly is participating in many data-shar- implied duress, is of greater concern to civil
ing and computer-matching programs. It is an- libertarians because of the data-sharing and
ticipated that SMP, when implemented, will computer-matching activities described below’,
affect these programs by: 1 ) increasing their which means that the information (and errors
number, by making them easier or less costly; that it might include) can become widely dis-
2) encouraging their use for front-end verifi- seminated, through channels and to destina-
cation (that is, original determinations of eligi- tions that the citizen does not even know about.
bility for benefits programs); and 3) facilitat-
ing the electronic exchange of information, New Information Policy Directives
including “hits’ or successful matches, over for the SSA
long-distance wires, cables, or satellite trans-
missions. Civil livertarians are concerned be- During the 1970s three major themes gov-
cause data-sharing and computer-matching erned Federal information policy: defining the
capabilities increase the opportunities for in- privacy rights of individuals, defining rights
advertent or deliberate violations of privacy, to government information, and defining the
and could be misused for governrnent surveil- rights of individuals to access to and partici-
lance of individuals. pation in government decisionmaking through
Throughout its history, SSA has had an ex- eight major piece. - of legislation. 18’
cellent record of respect and care for the The information policy’ legislation of th~~
privacy of its clients. Recently. however, the 1980s is concerned with different concerns and
increased emphasis on reduction of fraud and subjects: reduction of fraud in Federal pro-
improved debt collection sometimes comes into grams, efficient management of information
conflict with the letter or the intent of legisla- resources, and reducion of debt owed to gov-
tion designed to protect the privacy of citizens. ernment by releasing information to debt col-
For example, the privacy of tax information lection agencies.
is protected by the Internal Revenue Code, Sec-
The computer-matching activities of SS.4,
tion 6103, 26 [J. S, C., which permits disclosure
and the continual sharing of SSA data with
only by consent of the individual, and clearly
other Federal agencies and with State agen-
spells out the meaning of consent as ‘ ‘volun-
tary action. ‘‘ The following notice, taken from
an SSI application form, peremptorily de-
mands from the client tax information to be
used in making Supplemental Security Income
benefit determinations:
72

cies, reflects these new priorities. These pro- use modern business methods to reduce costs.
cedures are a departure in spirit if not in law PCIE was, in particular, designed to increase
or administrative procedure from SSA’S tradi- the use of government computer matching
tional policies regarding personal data, set out programs.
first in 1935, as described in chapter 5. While
These new initiatives have put pressure on
these traditional rules allowed data-sharing
SSA to engage in aggressive debt collection
under some circumstances at the agency’s ‘dis-
practices, and caused SSA to move strongly
cretion, ” SSA historically did so only rarely
to establish a Federal parent locator system
and with reluctance until recent years. 19
and a series of data exchanges with other Fed-
The Privacy Act included an ambiguous pro- eral and State programs.
vision that agencies should share information
Concerns about privacy, confidentiality, and
only for a purpose compatible with the pur- freedom of information are likely to grow in
pose for which it was originally collected–the the next 5 years, although at present they are
‘‘routine use clause. The implied limitation
overshadowed by concerns about efficiency
against sharing data was never seriously en-
and productivity, with a resulting emphasis
forced by OMB.
on sharing of Federal data. Legislative pro-
In the 1970s, GAO reports tended to reflect posals to protect due process rights of individ-
congressional concerns with invasion of pri- uals who are the subject of Federal computer-
vacy; by the mid- 1980s GAO reports encour- matching programs and related programs are
age the sharing of information among govern- nevertheless a possibility, within the next 5
ment programs at Federal, State, and local years. The need for security, data quality con-
levels in order to reduce fraud, waste, and trol, and system integrity will continue to
costs. Six major GAO reports recommended grow, and may well be made more acute by
use of computer-matching and tax return in- the threat of political terrorism. New legisla-
formation to reduce fraud and abuse in Fed- tion in this area is possible, especially if there
eral entitlement and benefits programs (these are significant lapses in security or discoveries
reports did not focus exclusively or directly of fraud.
on SSA).20 The sharing of information among
The major thrust of information policy in the
agencies was encouraged by OMB interpre-
near future, however, may be additional require-
tation of the routine use clause as covering any
ments for SSA to share information with other
use published in the Federal Register. The
Federal, State, and local agencies. Active po-
President Council on Integrity and Efficiency
litical support in both parties for maximizing
(PCIE) and the Grace Commission were estab-
government use of information will put addi-
lished in 1980 to assure that Federal agencies
tional demands on SSA information systems.
-—......—. — SSA’S Data-Sharing Programs
“hlany E’ederal agencies ha~e felt the pressure for increased
sharing of dat,a and lessened emphasis on prit’ac~’ and security
concerns; but for a contrasting ~riewpoint, see Sherry Court- SSA has important reporting and data ex-
Iand, “Census Confidentiality’. Then and Now, ” Go~rernnlent change relationships with States and localities,
information Quarterljr 2:4, 1985, pp. 407-418. other Federal programs and institutions, and
‘f’U. S. Congress, General Accounting office, Federal infor-
mation h’.vstems Remain Highlj’ \’ulnerable to Fraudulent, private insurers (through its continued admin-
I$rasteful, .4 bu.~i\re, and Illegal Practices, NASA D-82- 18, April istration of Medicaid/Medicare). These rela-
1982; .1 ction.q $reeded To lieduce, Account for, and Collect O\’er- tionships are a function of policy and stat-
pa~”n]ent.~ to Federal lietirees , 4FNID-8;?-29, June 1983; Com-
put~r ,lla[ches ldcntift Potential Uncnlplqjrnwnl Benefit (h’er- utorily defined programs. SSA’S major data-
pa~’ment, (;(; 1)-/+3-99, August 1983: [;.30 obser~”ations on the sharing relationships are:
[ [s~, of ‘f’a<y ]i(~turn lnforl]l:~tion for \’edification in Entitlement
l’rograms, t{ 1{1)-84-72. ,June 1984; Better ll”age-,$lat(>hing S~rs- ● The Beneficiary and Earning Data Ex-
tem.~ and Procedures 11’ould l+;nhance Food Stamp Program 1n- change (BENDEX), created in 1968 to pro-
(egrit~’, R(’E 1)-84-1 12, September 1984: and 1+,’iigibilit.kr \’erifi-
cat ion and Pri~’ac.\’ in Federal Benefit Programs: .4 Deli(’a te vide Title 2 information to States for ad-
ilalan(.e. 11 1/1)-/+5-22, hlarch 1985. ministration of the AFDC programs. This
73

is a monthly batch system with transmis- SSA’S Computer-Matching Activities


sions occurring at the request of specific
States (not all States are members of the In modern society, most persons leave a trail
system). There are on average 3 million of transactions with various institutions—
inquiries per month. governmental, retail, financial, educational,
The State Data Exchange (SDX), devel- professional, criminal justice, and others- as
oped in 1974 at the time of implementa- discussed in a recent OTA report on surveil-
tion of Title 16 (the Supplemental Secu- lance. :] Before the widespread use of conl-
rity Income Program), to advise States of puter-communication systems, linking various
the amount of SSI payments, eligibility kinds of transactions was very difficult, if not
for Medicaid, and other information to as- impossible, since transactions were paper
sist in administration of income, health, based and the cost of matching or linking pa-
and food programs. Data is exchanged per records was prohibitive. In addition, the
(usually by magnetic tape) weekly or time delay inherent in paper linkages would
monthly depending on agreements with negate much of the potential surveillance
a State. There are about 2 million ex- value. The advent of large fully’ computerized,
changes per month. easily accessible databases, and the ability to
The State Income and Eligibility Verifi- exchange and compare data between them, cre-
cation System (SIEVS). The Deficit Re- ates a much larger risk of violations of privacy.
duction Act of 1984 required the States At present, some government uses of data for
to develop a verification system for ad- purposes other than those for which they were
ministering federally assisted programs collected, albeit for legitimate governrnent
such as unemployment insurance, fod functions of law enforcement and investiga-
stamps, Medicaid, and AFDC. SSA will tion, are being challenged.
provide data to the SIEVS from SDX and Because SSA collects, stores, and uses a
BENTDEX and will respond to State re- large amount of data about individuals—earn-
quests for assistance. SIEVS will also be ings and income, employment records, depen-
used in social security number \’erifica- dents, home and work addresses, etc.–and
tion. SSA in turn will be able to receive matches these data with data about the same
information from the States to aid in ad- individuals from o{ her sources (e. g., State
ministration of SSA programs and avoid prison systems and welfare agencies), its pol-
overpayments. icies and procedures with regard to individual
The Tennessee Data Exchange (Model privacy are of special concern.
Program). This is a pilot on-line data ex-
change between a Tennessee State welfare Computer-matching is a technique whereby
agency and SSA; it was designed to speed a computer compares two databases to iden-
the provision of SSA data to the State for tify overlaps, e.g., individuals for whom both
eligibility determinations. databases have records. The rolls of recipients
under one public assistance program, for ex-
The upgrade of the SSA systems so far has
substantially increased the ability to respond
‘] Pri~acy, security, and surveillance issues ar(’ a primarj’ ftJ-
to batch requests from State agencies. Whether cus of a series of OTA assessments known collect i~elj’ as t,he
in the future SSA capacity will be sufficient Federal Government Information Technology’ Assessments.
to support on-line response to State agency in- Three reports from this study hate been published: F;lectroni(
Sur\’eiilance and Civil Liberties, October 1985; Tfanagement,
quiries is still uncertain. The SIEV program Securit.v, and Congressiomd O\ersight, February 1986: and fi;fec-
in particular will place an additional workload tronic Record S~’stems and indi~’idual Pri\ac>’, (June ] 986. ‘1’he
on SSA; when fully implemented, SM P will in- next few paragraphs draw liberally on these reports, and on
responses to a Federal Agen(~’ [)ata Request which was sent
crease SSA efficiency in meeting the require- (Jut to assist the OTA staff in collecting information for use
ments of this system. in these assessments
74

ample, may be matched with the rolls of the files that are used remain the property of
another such benefits program to identify peo- SSA and must be returned or destroyed, as
ple who are getting multiple benefits. Both appropriate, after use; they may not be dupli-
databases may include several kinds of infor- cated or disseminated without written permis-
mation about the person; the match will or can sion; they must not be used to extract infor-
aggregate this information, thus potentially mation about ‘‘nonhit individuals’ (i.e., those
allowing the user to know or deduce a great who appear only on SSA records); they must
deal about the subject person. Although the be used only by authorized employees under
purposes of computer-matching are generally supervision, and those users must be explicitly
legitimate and justifiable, it also opens the door informed about Privacy Act requirements and
for misuses of such personal information by OMB guidelines as to protection of privacy.
government, or by persons who have access
As can be seen in table 2, SSA generally uses
to the information and may use it for unautho-
computer-matching to verify the status of clai-
rized purposes.
mants or their dependents with regard to ben-
There are several questions to be asked about efits programs or to determine whether an in-
such programs, in addition to the broad issue dividual is collecting a paycheck or another
of whether they are inherently an unjustifia- form of assistance. For example, is a bene-
ble intrusion on privacy or an unacceptable risk ficiary’s surviving dependent in full-time at-
to civil liberties. These include: tendance at a legitimate school or university,
in order to qualify for students’ benefits’? Is
How are computer-matching programs au-
a recipient of disability benefits in prison (in
thorized and who is responsible for their
which case benefits are suspended)? Under a
use? pilot program, SSA is matching data with
Is the data used strictly and solely for the
State agencies about interest payments from
purpose for which it was collected, as re- financial institutions, to assess individuals’ in-
quired under the Privacy Act?
come and resources, for use in Supplemental
Are these activities cost-effective?
Security Income determinations. SSA data,
What assurance is there that the matches,
conversely, is shared with several Federal and
or ‘‘hits, are valid, that is, accurate and
State agencies, including IRS, the Veterans
verifiable? Administration, and the State of California.
What safeguards does the individual ha~~e
against incorrect “matches” that penal- SSA does not, as yet, use conlputer-match-
ize him or her in some way’? ing in the original determination of eligibility
for a program for new applicants (“front-end
SSA makes liberal use of computer-matching
verification”). It may do so in the future; no
techniques. These matching programs are not
decision on this point has been reached as yet.
specifically mandated by law, but are often rec-
ommended to SSA by GAO to increase the ac- The computer-matching programs, it can be
curacy of its determinations of eligibility and argued, are a significant departure from the
benefits amounts. In other cases SSA allows spirit if not the letter of SSA’S famed Regula-
its data to be used by other agencies—Federal tion No. 1, issued in 1935, which expressed the
or State—for their own purposes. Table 2 agency’s commitment to safeguarding the con-
shows the major computer-matching programs. fidentiality of personal data (see ch. 5). There
have, however, been no court challenges to
SSA computer-matching is undertaken un- SSA on the grounds of privacy in computer-
der OMB guidelines and the conditions are matching.
spelled out in written agreements with the co-
operating (matching) Federal or State agency. SSA has not done any formal cost-benefit
When SSA allows other Federal or State agen- analysis of the computer-matching programs,
cies to use its data for matching, these agree- either before or after the matching is run. How-
ments typically contain a set of safeguards: ever, there is usually a pilot run, which gives

75

Table 2.—SSA Computer-Matching Activities

SSA data “matched for SSA purposes:


(continuing, annually, unless otherwise indicated)
Agency matched with Type of data For SSA determination
U.S. Department of Education Full-time attendance status Eligibility for student benefits
U.S. Department of Education Student marital status Continuing benefits
U.S. Department of Defense Military payments SSI overpay mentsa
U.S. Office of Personnel OPM payments SSI computation
Management
Railroad Retirement Board RRB payments SSI computation
U.S. Veterans Administration VA payments SSI computation
Various State and Federal Workers’ compensation, Benefit computational
agencies State pensions, AFDC,
general assistance
benefits
Federal and State prison List of felons Precluding payments of
systems benefits (Public Law
96-473)’
U.S. Office of Personnel OPM payments Prevention of overpay mentsb
Management
U.S. Department of Labor Black Lung reports Overpayment of Part C Black
Lung benefitsa
Various State agencies Annual interest income Overpayments from under-
from financial institutions reporting of income/
resources (pilot)a
State and Federal agencies Workers’ compensation Overpayments (pilot)a
U S. Internal Revenue Service Income data Overpayments because of
unreported income/
resources
SSA data matched by other institutions for their purposes:
User agency Data Agency’s purpose

U.S. General Services SSA master files Social Security number
Administration validation e
U.S. Internal Revenue Service SSA data Administration of Elderly Tax
Credit a
State of California SSA data Eligibility for Medicaid
benefits e
U.S. Veterans Administration SSA data VA offset of SSA Black Lung
payments f
a
As needed
b
Twice a year
c
Three times a year
‘Quarterly
‘One time
‘Monthly
SOURCE Off Ice of Technology Assessment, 1986

some indication of whether the matching will Some individuals do lose their benefits, or
be productive, and there is a calculation of sys- have them reduced, as a result of computer-
tems costs in running a program. The agency matching; otherwise there would be no bene-
says that in preparation for expanded comput- fit to the agency (and OMB) in using the tech-
er-matching when its systems modernization nique. SSA goes to some length to verify
has progressed further, procedures are being “hits”; they are checked against the original
developed for systematic prerun and postrun data on SSA’S tapes or disks, and the subject
cost-benefit analysis. individual, who has not of course given con-
76

sent to or been notified of the matching proce- law enforcement and various investigative
dure, is given an opportunity to challenge and functions of government.
refute the results.
The technologies that OTA considered in-
Benefits that an individual may “lose” as clude for example, satellite communication sys-
a result of these computer-matching activities tems, digital switching and transmission tech-
are (assuming that the information is accurate) nology, computer databases, electronic mail,
unlawful benefits, that is, benefits to which he and integrated services digitil networks, many
or she was not entitled. The real concern of civil of which SSA uses or will be using. Others are
libertarians is the possibility that such tech- less likely to be of use to SSA, although their
niques, and the databases on which they oper- use at some time in the future is possible. 23
ate, might be used for other purposes, such as SSA expects to use, but does not now use,
surveillance. teleconferencing, expert systems, voice mail,
and optical disks. There are pilot projects now
Future Information Systems and underway to explore some of these techniques.
Possibilities for Abuse SMP does offer the potential, in the future, of
giving people access to their own SSA records
In 1985, OTA issued a report on the use of through home computers. SSA is not planning
computer and telecommunication technology for this but several States are considering such
by the Federal Government for surveillance plans with selected State record systems.
and monitoring of individual behavior. zz The
report said that many new and emerging elec- SSA is not an enforcement or investigative
agency, but it is responsible for certain func-
tronic technologies can be used for monitor-
tions such as entitlement determination and
ing individual behavior, and the use of other
debt collection, that could involve surveillance,
electronic technology, such as telecommuni-
as well as for safeguarding its data and its
cation systems, can be easily monitored or
transactions, which involves monitoring the
recorded for investigative, competitive, or per-
use of its equipment and the behavior of its
sonal reasons. The existing statutory frame-
employees. Much sensitive SSA data will flow
work and judicial interpetations, OTA pointed
over leased lines between headquarters and in-
out, do not adequately cover new electronic sur-
teractive terminals in field offices when the new
veillance applications; the law has not kept
claims modernization project becomes oper-
pace with technological change.
able. Satellite communication links are also
The basic public law for protection of oral possible. The new systems that SSA plans to
and wire communications is Title III of the develop to assure the integrity and confiden-
Omnibus Crime Control and Safe Streets Act tiality of its data are not yet fully developed.
of 1968, which predates most of these technol-
ogies. Digital communications between com- Security of SSA Systems
puters is not covered by existing statutes, and
policy on database surveillance—the monitor- Data in computers and telecommunication
ing of transactions on computerized record sys- systems are vulnerable not only to misuse but
tems and data communication linkages is not to inadvertent loss through systems failure,
clear. The courts have on several occasions to theft, or to manipulation or destruction
noted that the law has not kept pace with these through sabotage or terrorism. The security
technological changes. Congress in legislating of information systems against internal or ex-
in this area attempts to strike a balance be- ternal violations is of primary importance.
tween civil liberties and the needs of domestic
——.—
J ‘Electronic eavesdropping technology, optical/imaging tech-
‘<U.S. Congress, office of Technology Assessment, Elec- nology for visual surveillance, sensor technology, civilian band
tronic Sur\eillanre and CiviI I.iberties (Washington, DC: U.S. radios and vehicle location systems, polygraphs, ~’oice stress
(;c)lt’rnment Printing office, octoher 1 985). analyzers, t’oice recognition, laser interception, and cellular radio.
77

The security procedures at SSA’S National As interactive terminals and personal com-
Computer Center are those common at most puters are added to field offices, these concerns
large ADP centers. Physical security for the will become pressing. Access controls are be-
facility and for separate rooms and floors is ing reviewed and revised as part of SMP, but
thorough; and data security is safeguarded this work is not complete.
with standard techniques of personnel screen-
There have been a number of cases of inter-
ing, restrictions on dial-up access, passwords,
nal sabotage and computer-related crime at
and audit trails. Backup battery power and
SSA, as is perhaps inevitable.’:) SSA says,
generators are available to keep the computers
however, that no known instances of computer
going for 3 days in case of power outages. SSA
crime involved data processors; they occurred
says that all records are backed up.
earlier in the work process. A typical case is
The DHHS Inspector General warned SSA a field office employee inventing a fictitious
in early 1984 that: claimant, or altering information about a ben-
eficiary or a payment amount.
SSA is not prepared for a disaster in the
NCC. . . . SSA’S ADP systems are highly cen- SSA has long been criticized for having in-
tralized in its NCC and operate without ade- adequate safeguards against unauthorized
quate backup in the event of critical damage, access to its data. Specifically, it has lacked
or worse—a catastrophe. Although there have programmer security controls, internal access
been attempts made to plan for contingencies,
controls, and audit trails. Though no computer
efforts to date have been inadequate. Further,
progr ammer at SSA has ever been found guilty
off site backup of data and software is incom-
plete and untested.24 of fraud against the agency, it has been quite
possible for programmers to make changes to
The audit report said that responsibility for pay themselves benefits; unauthorized people
contingency planning had not been focused at could log onto systems; data review techni-
a high level, SSA had not performed necessary cians in District Offices could enter claims for
risk analyses, and SSA components ‘‘whose themselves without leaving an audit trail.
expertise is necessary to develop a workable
plan” had not contributed to the effort. Sub- The SMP will: create an audit trail for com-
sequently SSA agreed to establish a new secu- puter program changes, assign personal iden-
rity planning work group and assign greater tification numbers to claims representatives
importance to contingency planning. A risk and local workers, create an audit trail for all
analysis study for the National Computer Cen- transactions, and employ a central security
ter had been done in 1982; but subsequently systems package like those used by the mili-
there have been several additional contractor tary to handle log-on commands and records.
studies of risk as well as top secret access pro- However the very rationalization of SSA pro-
cedures and audit controls. cedures and the existence of schematics and
diagrams mapping the system, pose a threat
None of these security measures apply to the to security that does not exist now, in that
use of microcomputers, outside of the National more people may be able to discover how to
Computer Center, for example in headquarters get into the databases.
and operations buildings, and there are no
established security policy or procedures for
microcomputer users. While the integrity of ~“’In the 4 years before the SMP beg-an, there were at least
data is fairly well assured, privacy may not be. 46 known cases of vandalism insidesecure data-processing areas,
and former SSA officials told Congress of other threats of sab-
otage that had been received. See U.S. Congress. illisrna.nage-
ment of SS’A Computer S’>’sterns Threatens Social Securit>r
~ ‘[; .S. I)epartment of I Iea]th and I]uman Ser~rices, office of Programs, 33d Report b> the l~ouse Committee on Go~’ernment
I nspect~)r (~ener:il, Nlemorandum to Martha A. NlcSteen, Act- operations. Sept. 30, 1982, p. 9. The SSA response to OTA
ing (“f)m missi{)nc~r of .Sc)cial Securitjri ‘‘ Audit Ilpport — .$s.~!s F’ederal Agencj Data Request acknowledged some (presuma-
l’l:inning for (’ontingt’n~”ie> in t h~> N(’w Computer (’enter, ,4(’N bly recent ) ‘‘known instances of crime and abuse’ but specified
1 ,-)-lY62i, \l :ir. 22, 1 9k 1, that they did not in~ol~e data-processing people.
Chapter 4

The Future: The Necessity of


Long-Range Technology Planning
Contents
Page
Elements in SSA’S Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
SSA’s Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
The Placement and Structure of the Agency . . . . . . . . . . . . . . . . . . . . . . . 82
Staff Changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Clients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Trends in Information Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Software Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Hardware Trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Management Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
SSA’s Continuing Uncertainties: Decisions are Necessary . . . . . . . . . . . . . 8’7

Table
Table No. Page
3. New and Potential Technology forSSA Functions . . . . . . . . . . . . . . . . . 83
Chapter 4

The Future: The Necessity of


Long-Range Technology Planning

The Social Security Administration badly assessment, even if greatly improved at SSA,
needs well-developed and effective long-range will not constitute a long-range planning ca-
planning capability. The Systems Moderniza- pability. Factors other than technology will in-
tion Plan deals primarily with technology, and fluence the organization’s future, such as
secondarily with technology management. It changes in its mission, resources, and relation-
will be most effective if it is implemented and ships with other institutions; and these fac-
extended in the context of an institutional long- tors shape the purposes and goals for the tech-
range plan that provides insight about how the nological systems. Some of these factors are
agency mission will change over time, alter- within SSA’S control, and some are not; but
native ways of performing that mission, the continuing monitoring and analyzing of such
resources that will be available, and the capa- factors can allow the agency to be prepared
bilities that will be needed. for changes and make a smooth transition to
new ways of fulfilling its mission.
Technology is an essential element in SSA’S
future. But technological forecasting and

ELEMENTS IN SSA’S FUTURE


SSA’S Mission to respond with credibility to either the Office
of Management and Budget, Congress, its em-
SSA’S basic mission is unlikely to change ployees, or its critics on questions related to
significantly over the next decades. There will the realization of productivity gains, the most
almost certainly be congressionally mandated manageable rate of work force reduction, and
changes in coverage, entitlement provisions, the appropriate timing of further procure-
benefits, etc. Some of the programs now admin- ments, largely because it has not explicitly de-
istered by SSA could be removed, or other fined its goals and milestones in a way that
responsibilities added, with additional require- can be credibly justified. Looking beyond the
ments for data collection and handling. New nexts years, a long-range plan should provide
social programs that might be assigned to alternative technological and institutional
SSA, however, usually have long gestation mechanisms for service delivery in the context
periods, and the assignment can be anticipated of evolving needs of the clients and future ca-
by agency planners. pabilities and costs of information systems.
SSA’S mission therefore is, and is likely to re- The functions essential to carrying out
main, more stable, coherent, and routinized than SSA’S mission involve:
that of many government agencies. This is a
significant advantage for long-range planning. data collection;
data processing;
The objective for planners, in this situation, data protection (privacy and security);
is to help the agency define its goals and pri- service delivery;
orities in carrying out its mission, and to help accountability and information dissemi-
it set reasonable standards of performance in nation (providing information needed by
terms of quality, timeliness, and costs of serv- the Administration, Congress, and the
ice delivery. At present, the agency is unable public); and

81
82

● coordination with other Federal and State Staff Changes


agencies.
This may be the most immediately challeng-
Each of these is subject to change. New ing and least adequately considered element
kinds of data may have to be collected, new in SSA’S future. Technological change, con-
services may be mandated, law and public pol- gressional budgetary decisions, and Adminis-
icy may set new standards for data protection tration policies are pressing toward significant
and accountability, and new technology will changes in SSA’S work force, but the agency
change both the available techniques for and is responding in a largely reactive rather than
the costs of performing all of these functions. proactive mode. A rational and persuasive plan
for reconciling and mediating these pressures
The Placement and Structure could allow the agency to shape and influence
of the Agency decisions that will finally be made or sanc-
tioned outside the agency.
SSA might at some time be separated from
the Department of Health and Human Serv- The groundwork was laid for an innovative
ices and be given the status of an independent approach toward cooperative labor-manage-
agency. The most important changes in man- ment adjustments to change, but there are in-
agement and reporting responsibility would in dications that this promising start is being al-
this case be mandated by law. The most likely lowed to wither. Unresolved questions involve
structure of agency leadership is already appar- recruitment, training, job classification, com-
ent from legislative proposals before Congress, pensation adjustments, promotion, relocation,
and should therefore not take the agency by working conditions, labor-management rela-
surprise. tions, and retirement policies. All apply to
managers, professional staff, and clerical staff,
Internal reorganizations are a more imme-
but there are different needs and constraints
diate and more likely possibility, and have sig-
for each group. These needs and constraints
nificant implications for long-range planning.
are intimately related to changes in technol-
Information systems development and plan-
ogy, and to Administration policies, the two
ning to be most effective must reflect and sup-
factors with which long-range planning will be
port the flow of work–i.e., the movement of
primarily concerned.
information and the sequencing of steps in its
processing —through the organization. A long-
range planning process that enabled the agency Clients
to define its goals and priorities could provide
a valuable guide to organizing the agency for While SSA’S mission is basically simple and
greatest effectiveness. It would allow the phas- stable, there will be changes in the demo-
ing in of desirable changes in a logical and or- graphic makeup of its clients over the next two
derly manner, providing at the same time a decades, which will or should affect the way
rationale and justification for the changes to in which the mission is performed and the cri-
managers and workers. Conversely, the rela- teria for excellence in performance. For exam-
tionship between programs (OASI, SSI, etc.), ple, the age distribution, educational level, eth-
between operations and systems development nicity, language problems, technological so-
components, and between field offices, Re- phistication, and family resources of benefici-
gional Commissioners Offices, Program Serv- aries and their survivors may change signifi-
ices Centers, the National Computer Center, cantly. Some of the problems which occurred
and headquarters staff, should be considered with implementation of the SS1 program illus-
in establishing a planning unit and determin- trated the way in which changes in client needs
ing its responsibility, location, role, and report- and expectations determine the effectiveness
ing processes. of traditional SSA procedures.

83

TRENDS IN INFORMATION TECHNOLOGY


New advances in information technologies Table 3.—New and Potential Technology
and related management tools, beyond those for SSA Functions
envisaged in SSA’S 5-year Systems Modern- Function Technology
ization Plan, are available now or are reason- Communications . . . . . . . . . . Local area networks
ably certain of becoming available in the near Electronic mail
term, and can be rationally anticipated and Private branch exchanges
Digital switching and
planned for (see table 3). They will be the stand- transmlssion
ards by which experts will evaluate SSA sys- Fiber optics
tems and management in the 1990s. Unless Communications satellites
Cellular mobile radio
SSA begins now to systematically prepare for Data encryption
modernization beyond a 5-year horizon, it may Integrated Services Digital
again find itself falling far behind the “state- Networks (ISDN)
Voice mall
of-the-art” at which SMP is aimed. Teleconferencing
Two-way cable
SSA has historically emphasized heavy-duty Data collection ., . . . . . Home computers
computing needs of the core administrative Client-operated devices
functions-the processing of enumeration, (similar to ATMs)
Data input ., ... . . . . . . . . Optical character recognition
earnings, and the master beneficiary file. Other Voice recognition
uses for computing will be increasingly im- Data output and
portant. presentation ., ... ... Computer graphics
Voice synthesizers
Some new developments in information tech- Data processing . . . . . . . . Minicomputers
Relational databases
nology will be useful chiefly for advanced sci- Query-by-Example
entific research, at least in the near future, but Multi-use super-micros,
such capabilities are generally soon adapted application S O f t W a r e
packages
to the more prosaic operations of corporate and Supercomputers: multiple
government institutions. Monitoring these de- instruction multiple data
velopments and trends as they emerge will be (MI MD) processors, vector
processors, data driven
necessary if SSA is to plan to use new techni- processors, FORTRAN
cal capabilities when they become reliable and programs
cost-effective. Parallel processing
Associative processors
Other technical capabilities are already avail- Entitlement evaluation,
adjudication . . . Spreadsheet applications
able and in use in leading private sector orga- Expert systems
nizations and in some government agencies. Storage ., . . . . . . . . . . . . . Magnetic bubble devices
Some of these do not, however, yet appear in Wafer-scale semiconductors
Optical disks
SSA’S Systems Modernization Plan, or are Smart Card
only incorporated as eventual enhancements SOURCE Off Ice of Technology Assessment 1986
-

rather than as pivotal points of leverage for


making optimum use of information systems.
pointed out above, they could also be used, un-
For example, to use information technology fortunately, to select and present only favora-
to the fullest in improving its management of ble indicators and benchmarks.
operations, SSA will need to develop more
powerful administrative processing and man- Software Trends
agement information systems, designed to ac-
cess transactions information and manipulate Software development has historically
it to answer managers’ questions. These man- lagged behind hardware development, but it
agement information systems could also aid is now a major focus of information science re-
Congress in oversight of SSA, although as search and development in the United States
84

and in other nations, and can be expected to order-of-magnitude increases in productiv-


move rapidly over the next decade. Major ity; it is estimated that most fourth-gen-
trends in software that could be better utilized eration languages result in 10 times greater
by SSA are: productivity than COBOL. Most of these
languages are not yet capable of being
Enhancements to Systems Analysis and
used for sophisticated applications but
Design: These tools, called “interactive re-
they are likely to be so in the future. In
quirements analysis tools, ” are designed
the area of management information sys-
to codify data, reduce errors, and improve
tems, they are now capable of providing
the documentation required when estab-
responses to most kinds of questions
lishing the requirements for a large infor-
needed for reports. Examples are FOCUS,
mation system. They fall into the larger
RAMIS, SQL.
family of tools called “structured analy- ● Relational Databases: These database
sis, design, and programming techniques. ”
structures (ways of organizing data) per-
New ones use forms-based systems to
mit great flexibility yin the ways of asking
achieve uniformity of documentation and
for information, and they use English-like
design, by prompting the designer with
fourth-generation languages in addition
questions and blanks. They improve com-
to being compatible with COBOL and
pliance with programming and design
other third-generation languages. Exam-
standards, which has been one of SSA’S
ples are Univac’s Mapper, Culinet’s
problems. Others help the systems ana-
IDMS, and IBM’s DBase-II. Relational
lyst to put the systems design in a par-
databases are now a small but rapidly
ticular format or notation that can be fed
growing portion of the data storage and
into a code generator for key parts of a
retrieval market. For high-volume applica-
computer program. These software design
tions they are still too slow; as of yet they
tools can in the future be incorporated into
cannot be used with very large, hetero-
larger automated design tools.
geneous databases with many thousands
Code Generators: These are computer pro-
of records, but very powerful computers
grams that assist progr ammers in quickly may in the future enable SSA to take
producing third-generation language pro-
advantage of them.
grams, for example, COBOL programs.
One is Quick-Code, which generates Developing and maintaining software is now
DBase II programs; others are Quick-Pro, the dominant cost of creating and operating
the Producer, and Genasys. Some reports large computer systems. Advances in micro-
have shown productivity gains approach- electronics have steadily reduced the cost and
ing ’75 percent with the use of these code improved the performance of hardware, but im-
generators. provements in the productivity of program-
Development of Fourth-Generation Lan- mers has been much more difficult to achieve.
guages: These are computer languages Software engineering, techniques for improv-
that use English-like vocabulary and syn- ing the productivity of programmers and de-
tax, and are useful for the development signers, is increasingly important. Research
of administrative and management infor- in this field includes highly theoretical work
mation systems. The distinction between directed at fundamental understanding of the
fourth-generation languages and database nature of programs and “proof,” in some math-
management (software) systems is in fact ematical sense, that they will work as intended.
becoming ambiguous, and often depends At the other end of the spectrum, it includes
in part more on who will be using them behavioral science research on the ways in
(systems developers or systems end-users), which people interact with computers (the
than what they can do. For programmers, ‘‘man-machine interface’ and techniques for
these computer languages can m-educe management of programming tasks.
85

Artificial intelligence is a field of research 10 to 20 times, reducing the cost of these


concerned with extending the ability of the operations.
computer to more nearly match human men- ●Increased Power: It is likely that in the
tal capabilities, such as recognizing and un- next toyears the cost per unit of comput-
derstanding speech and visual images, reason- ing power will continue to decrease and
ing, choosing among options, or deciding, and the speed increase, as they have in the re-
spontaneously communicating. The first sig- cent past. Supercomputers—or advanced
nificant commercial applications of artificial architecture and parallel processing-may
intelligence, after at least 25 years of research, become available to very large organiza-
are in the area of expert systems. These are tions like SSA; allowing processing at
‘‘intelligent information retrieval systems de- much higher speed and much lower unit
signed for use in tasks requiring expert knowl- cost. The need for sophisticated cost-
edge, such as insurance underwriting, medi- effectiveness comparisons of systems in
cal diagnosis, weapons control, or business the future will be of increasing importance.
decisionmaking. Expert systems store not only ● Mass Storage Technology: Magnetic disk

data, but rules of inference that describe how storage technology continues to improve
an expert would use the data to make decisions. incapacity and cost, but laser optical disk
The expert systems’ rules of inference are de- technology offers storage of up to 2.5 bil-
rived from analysis of the decisions of many lion bytes on one 12-inch disk, several
experts, and can therefore make a user’s deci- orders of magnitude more density than
sions more comprehensive, more rigorous, and possible on a magnetic disk. There are still
more consistent than those of a typical user a number of technical problems related to
who is not in the top echelon of experts in the the use of optical disks for organizations
field, and at least in theory better than the de- with immense databases like SSA’S; for
cisions of any one expert might be. Since in- example, improved computer-controlled
surance underwriters are already using expert indexing systems are needed, and there
systems, applications for SSA claims repre- are unanswered questions about their ar-
sentatives, or State determiners of disability, chival ‘shelf life. Until very recently, op-
are an obvious possibility to be explored. tical disks could not be used in a read-write
mode; since they could not be changed
once data was recorded, they were useful
Hardware Trends only for permanent storage. Several firms
are now working on erasable optical disks.
In order to run sophisticated languages and Some commercial applications are avail-
relational databases, faster and cheaper ma-
able and others are nearly ready for mar-
chines with new capabilities are needed. The ket. Access time for erasable optical disk
faster and cheaper the machines become, the
systems may continue to be a problem be-
easier and more economical it is to use lan- cause of the enormous volume of data they
guages which are less efficient but more suit-
can hold, but they clearly promise superior
able for use by nonspecialists. The following
performance in comparison to existing se-
developments should be helpful for SSA:
quential-access magnetic tape systems.
● Database Machines: In order to operate ● Communications: SSA does not have
relational databases at reasonable speed, much in the area of local area networks
it is necessary to use parallel processing, (LANs), PBXS, and wide-area networks
so that several operations can be done at (WANS) to link computers in buildings,
once. A database machine is several orders cities, and counties. Failure to take advan-
of magnitude faster than other machines; tage of these technologies will prevent it
some are on the market that can execute from developing efficient office automa-
10 million instructions per second and tion and enhancing productivity in local
speed up database transaction processing district offices.
86

● Optical Character Recognition Technol- design are being approached, and further in-
ogy (OCR): OCR is now undergoing a creases in computer performance may require
spurt of rapid development. Typed or parallel processing architectures. That is, oper-
printed material can be read into the com- ations performed on the data would be decom-
puter 40 to 50 times faster than it can be posed into tasks that can be simultaneously
keyboarded. State-of-the-art OCR equip- carried out by many computational units work-
ment can read up to 23 different fonts with ing in parallel.
an error rate smaller than that of an ac-
Very large-scale integrated (VLSI) circuit de-
complished data-entry clerk, and some of
the devices can also handle simple hand- sign facilities, using the most powerful com-
printed notations, such as numerals. With- puters, are being used to develop and test new
in the next few years, OCR should have architectural designs. Three U.S. companies
the capability to recognize, isolate, and (CRAY, Control Data Corp., and Denelcor) are
read or copy specific bits of information developing “next generation’ supercomputers,
as are several Japanese companies. Japan has
within a larger volume of data.
embarked on major supercomputer projects
New concepts of computer architecture now that may challenge U.S. leadership in super-
being explored in the laboratory will provide computers. One, called the Fifth Generation
two waves of innovation: highly specialized, Computer Project, is aimed at producing a
low-cost computer modules that do specific computer using artificial intelligence, or rea-
types of tasks at extremely high speeds, and soning functions similar to human thinking
future generations of supercomputers. ’ processes.
Because computer hardware has in the past
been expensive, users have tried to allocate its Management Tools
cost over a variety of applications and uses, SSA has adopted new software engineering
with general-purpose mainframe computers. techniques such as project management sys-
But as the costs of hardware drops, some com- tems, automated documentation systems, re-
putational tasks, involving high-volume stand- port generators, screen editors, database man-
ardized data, may in the future be done on in- agement systems, etc. Already, however, some
expensive, special-purpose hardware with the private sector organizations are moving from
general-purpose computer used as a routing this software engineering technology approach
switch, sending a computing task to one of sev- to methodologies such as prototyping the end-
eral different specialized processors. user development (some of these new tools were
The term supercomputers is used for the described above under software trends). SSA
most powerful computers available at any one should test and evaluate these newer technol-
time. The next generation of supercomputers ogies as soon as it has a database on which
is likely to have a radically different “architec- they can operate.
ture. ” Computer architecture is the internal Federal agencies are required by the Paper-
structure of a computer, the arrangement of work Reduction Act to introduce information
the functional elements that carry out calcu-
resource management (IRM) as a concept and
lations and manipulation of data. Since their
organizational tool. This includes the creation
invention, computers have basically followed
of a data administration function and a data
one model, the von Neumann sequential proc-
administrator’s position at a high level in the
essing architecture. The limits of computa- organization. The concept of IRM is still con-
tional speed that can be achieved with this
troversial and poorly implemented in most
Federal agencies and in private organizations.
‘Much of the material in this chapter is based on an earlier Nevertheless, it is important that SSA take
OTA report, Information 7’echrzolo~ R&D: Critical Trends and
Issues, OTA-CIT-268 (Washington, DC: U.S. Government Print- a fresh look at the implications of this concept.
ing Office, February 1985). It will need, for example, to put greater em-
87

phasis on development of organizational de- structured analysts, prototypes, data comnlu-


vices such as Information Centers, to help and nication network specialists, and decision sup-
support end-users. port specialists. A large number of profes-
sionals will have to be recruited or retrained
If SSA is able to adopt new technologies,
to produce the required labor force by the
a large shift in data-processing skills will be
1990s. Since many experts estimate that the
required, from traditional complete reliance on stock of useful knowledge which a program~-
COBOI. programming toward greater capabil- mer has is depleted by one-half every 5 to 7
ity to work with fourth-generation languages.
years, SSA’S technical training programs will
There are vast differences in the skills required.
be increasingly critical in terms of upgrading
There will be an increased need for database
technical knowledge on a continuing basis.
designers and administrators, code optimizers,

SSA’S CONTINUIN~ UNCERTAINTIES:


DECISIONS ARE NECESSARY
SSA’S response to its 1982 problems must making key decisions about protocols, orga-
be understood in terms of the technologies that nization, and maintenance, which heretofore
were available. Much of the criticism of SSA’S the telephone companies had assumed.
Systems Modernization Plan rests on the per- By the 1980s, database technology has al-
ception that SSA did not, in 1982, pay enough
lowed a shift in thinking about data from ‘‘in-
attention to the wide diversity of options avail-
formation as a cost” to ‘*information as a re-
able to it; and in the intervening 4 years it is
source. ” In the 1970s, information was usually
not clear that SSA has developed an under-
highly fragmented among different levels of
standing of the rapidly developing and diver-
the organization, different specialized func-
sifying technological options that it will have
tions, and different computer programs. It was
in the future.
difficult for specialists in one division to share
By 1982, it was no longer necessary for an data with those in another division. Database
organization to arrange the entire computing management in the mid- 1980s means:
function in a large, centralized, information codification of data elements to define
systems department; machine intelligence common meanings and to catalog origins
could be distributed throughout the organiza- and uses (a data dictionary);
tion wherever needed. In addition, microcom- reorganization of data elements from thou-
puters, mainframes, minicomputers, and pe- sands of computer files into a single pool
ripheral devices can be tied together with or “database”;
PBXS or LANs. separation of application programs from
The 1980s is also a period of tumultuous the data elements, by use of a database
change in the organization and marketing of application development language; and
telecommunication equipment and services. reorganization of data elements to permit
Many large organizations are bypassing local greater flexibility in responding to in-
telephone companies by using their own fiber quiries.
optic or microwave intracity telecommunica- It must be recognized, however, as has been
tion networks. The deregulation of the telecom- stressed throughout this report, that SSA’S
munication market has led to an explosion of performance and problems are not greatly
offerings of new telecommunication services, different from those of other organizations at-
and rapid reduction in the price of both long- tempting to catch up with and stay abreast
haul and local telecommunication costs. of broadening technological capabilities. The
Managers must take more responsibility for implementation of true corporate and govern-
88

mental databases is 5 to 10 years behind the by the limitations of their data-processing staff
idea, because existing software has to be amor- and can develop systems faster and in a more
tized, there are significant organizational costs rational, planned manner. There is increasing
involved, hardware must be upgraded, and top recognition of the close connection between
management is often not convinced that the systems and organizational structure, and of
large costs are actually justifiable. the fact that relatively large changes in or-
ganizational structure may be necessary to
The principle themes in software today are
take advantage of new technologies.
higher level languages, user-friendly lan-
guages, automatic code generation and soft- In 1985, debate about social security issues
ware engineering. Information now plays a has receded, and congressional attention has
strategic role in the operation of many organi- been focused primarily on SSA’S continuing
zations. The nonspecialized general user has attempt to modernize its technology and its
become more important; computer specialists management. This attempt is making prog-
no longer have a monopoly on data process- ress, but its promise is limited by SSA insis-
ing. There are more hardware, software, and tence on looking only 5 years ahead.
services options. Organizations are less bound
Part III

The Case History of Information


Technology Management at the
Social Security Administration

Created in 1935 to provide retirement insurance for American workers, the


Social Security Administration (SSA) grew through four decades to serve an in-
creasing number of beneficiaries in a variety of social programs. SSA’S data-col-
lection and management responsibilities from the beginning dwarfed those of any
private sector insurance company or of other government agencies. This agency
was from 1935 through the 1960s a pioneer in the adoption, utilization, and man-
agement of information technologies. Yet by the end of the 1970s, SSA’S data-
processing systems could no longer meet the requirements of SSA policies and
programs. How did a leader in one era of technological change become threatened
with obsolescence and failure in the next phase? How adequate is SSA’S current
response to the overwhelming problems that became obvious in the late 1970s
Chapters 5 and 6 describe SSA history between 1935 and 1981. Chapter 7 de-
scribes how the SMP was developed and initiated. It discusses in detail the fac-
tors that led to serious problems with the Paradyne procurement of 1982 and their
effects on SMP management. Chapter 8 lays out the structure of monitoring and
oversight mechanisms for SSA in both the executive and legislative branches of
government, reviews major oversight actions related to SMP, and identifies some
oversight problems likely to affect the monitoring of all complex agencies using
very large advanced technology systems.
Chapter 5

Years of Service
and Satisfaction,
1935-71
Contents
Page
The Birth of the Social Security Administration . . . . . . . . . . . . . . . . . . . . . 93
Tools of the Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Management in the Early Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Healthy Growth, 1940-71. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Programs and Resources (1940-71) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Management (1940-71) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Labor-Management Relations (1940-71) . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Technology and Procurement Policy (1940-71). . . . . . . . . . . . . . . . . . . . . . 98
Emerging Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Figure
Figure No. Page
7. The Organization of the Social Security Administration
in the Early Years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
--

Chapter 5
1
Years of Service and Satisfaction, 1935-71

BIRTH OF THE SOCIAL SECURITY ADMINISTRATION, 1935-39


When the Social Security Act was passed Staffing the agency was one of the first and
in 1935, a three-member board was created and most important tasks of the board and senior
given a modest lead time to set up the program. staff, and the way it was done was to have a
Old age insurance account numbers were to profound and long-term effect. The act required
be issued by January of 1937, and the first ben- employees to be chosen through Civil Service
efit payments were to begin in 1942. A first “except for experts and lawyers. ’ Top man-
year’s budget of $1 million was proposed by agement made liberal use of the ‘expert des-
the board and accepted by Congress. ignation to choose highly qualified persons not
then available through the Civil Service, and
The autonomy that SSA senior management
in that Depression era they had no trouble
had in their first 3 years of organization and
recruiting well-educated and highly qualified
operations was considerably more than most
workers. At the clerical levels management
Federal agencies enjoyed. Although social se-
was also able to pick the cream of the crop,
curity was considered a New Deal program, the
including clerks who had worked in record-
first Chairman of SSA was a Republican. The
keeping operations at the FBI and Census
strong support of the President, the fact that
Bureau.
SSA did not have to report through a Cabinet
officer, and the absence of any interagency ri- As a result SSA started with an unusually
valries over the mission or program boundaries well-qualified work force, imbued with a mis-
gave the first two chairmen very broad discre- sionary spirit of dedication to social insurance
tion in setting up the agency (see figure 7). concepts and a “client service” outlook.2 These
concepts were reinforced in the training pro-
1 Information i this chapter, where not otherwise Cited is
n
grams set up for all new employees, especially
drawn from a contractor report, prepared for OTA by the Educ~- training in courteous dealing with the public.
tional Funds for Individual Rights, Inc., New York City, Au-
gust 1985. The contractors conducted many interviews with
former and current SSA executives, congressional staff mem-
bers, Federal executive branch officials, union officials, tech- ‘Located in Baltimore, which had a large black population,
nology vendors, and scholars of SSA. They also drew exten- the SSA had from the beginning a policy of hiring substantial
sively on published materials. Many references and citations numbers of black clerical employees, a markedl~’ different pol-
not included in this chapter are provided in the contractor re- icv from that of the Federal Government at the time. Marvland
port and can be supplied if needed. then had legal racial segregation.

Figure 7.—The Organization of the Social Security Administration in the Early Years

-
I
Division of Bureau of Hearings
Division of the Actuary Program Research & Appeals
+ +

I v I I I
Bureau of Bureau of Old Age & Bureau of Federal
Children’s Bureau
Public Assistance Survivors Insurance Credit Unions
SOURCE Social Security Admlnlstratlon 1986

93
94
.— —

By 1939, SSA had a competent and highly were no significant procurement constraints
motivated work force, led by shrewd and dedi- on SSA in seeking out manufacturers to de-
cated headquarters executives; a large staff of sign new products or adapt existing machines
technical experts in actuarial, accounting, and for the agency’s special needs.
social-welfare operations; and afield corps com-
mitted to the practical delivery of helpful serv-
Management in the Early Years
ice to “entitled” clients.
The Social Security Board found no models
Tools of the Trade in the private sector in setting up its proce-
dures, since the insurance industry did not
SSA soon developed into the largest insur- have the enormous database, the need for fre-
ance organization in the world, in terms of num- quent updating, and the history-based entitle-
bers of persons covered, the number of persons ment process that characterized the old age
receiving benefits, the amount of benefits paid, insurance program. While Census, FBI, and
and the character of hazards insured.3 The the military had large recordkeeping and ac-
technology then available to SSA was largely counting operations, none of them had devel-
manual, mechanical, or electromechanical. oped procedures that could be applied to SSA’S
Data was stored in hard copy—ledger sheets, needs. SSA brought in outside consultants and
punched cards, carbon-paper forms, and file also began hiring and educating experts of its
drawers. Data processing depended on man- own, building up an in-house expertise that
ual operations and some early electric account- was, down to the early 1960s, at the leading
ing machines, such as the Hollerith system edge of recordkeeping and data-processing
first developed for the Census Bureau in 1890. science.
Data communications depended on trucks, the
mails, and sparing use of the telephone. The 1935-39 period saw several traditions
established that would persist at least until
SSA had to develop specifications for new
the late 1960s. The top managers were per-
types of recordkeeping and information-
sonally interested in and spent a great deal of
handling technology, and to call on leading
time on information management. Prompt en-
manufacturers to design new machinery or
rollment of new beneficiaries and getting pay-
adapt existing machines to new tasks. Three
ments out on time were given top priority. SSA
examples of such specification and innovative
adopted a deliberately incremental approach
responses were the development of the collat-
to technological innovation; at the same time,
ing machine by IBM for SSA use; adaptation
however, a cadre of experts was always at work
of the “Soundex” system for phonetic arrange
looking for new machines and new techniques,
ment and retrieval of names to large-file man-
and such activity was valued by the top leader-
agement; and application of early microfilm
ship. There was tension between operations
processes to SSA recordkeeping and data proc-
people, who generally wanted to continue to
essing.4 By 1940 SSA technical staff and ex-
use the machinery they had, and the systems
pert consultants were stretching the state of
people who wanted to push new approaches,
the art in information technology. They could
but this tension was usually mediated in the
do so by foreseeing future technological needs
Commissioner’s office.
and motivating manufacturers to meet those
functional requirements. In 1935 to 1939 there SSA was in these years an example of strong
—] — —
‘ Arthur J. Altmeyer, The Formative Years of Social Secu- administrative efficiency and program effec-
ritv (Madison, WI: University of Wisconsin, 1968), p. 6. tiveness, and the agency gave that highly
~Ibid.,p 71.
. Jack
, S. Futterma, The Future SSA pmCess:
An Apprm”sal for the AFGE of Its Impacts Upon Social Secu- favorable picture wide publicity. Its cost of
rity Administration Employees and Employment, May 1979 maintaining a worker’s account was then 20
(unpublished), pp. 4-5. cents a year, and the administration of the
95

trust fund and programs was done for slightly ployee or beneficiary information but left dis-
over 2 percent of each dollar collected.:) closure up to the discretion of the agency. ’
Disclosures were however approved very spar-
There was one public controversy over ad-
ingly. Legislative amendments in 1939 gave
ministrative matters, when reporters Drew
statutory weight to the board’s own confiden-
Pearson and Robert Allen reported that there
tiality rule.
were millions of unidentified ‘John Doe’
records. () SSA figures showed that these were The board instituted physical security pro-
less than 1 percent of total wage reports and cedures from the outset. Published literature
the agency had an active program to investi- records no instances of outside penetration or
gate and post them. In those days, attacks on inside misconduct in the 1935-39 period.
SSA’S recordkeeping usually reflected politi-
It became the agency’s policy not only to
cal conflicts rather than administrative ineffi-
allow old-age insurance account holders to ex-
ciency. In this era, SSA had a reputation in
amine their records but to actively solicit such
Washington for administrative agility and im-
inspections. Rights of inspection for account
aginativeness, and enjoyed significant auton-
holders were publicized and a sizable volume
omy. It had a reputation with the public for
of inspections took place each year. The agency
excellent service to clients. Within the agency,
saw this as a useful way to increase file ac-
information management was seen as a cen-
curacy, identify problems in their procedures,
tral, high-priority concern.
and to enhance public confidence in the system.
Concerns about potential misuse of personal
The 1939 amendments also provided a full
information by the Federal Government sur-
set of due process rights for retirees, widows,
faced as soon as social security was proposed.
and dependents; findings of fact and rulings
The President had responded with public as-
of the board could be challenged by the claim-
surances that all personal information would
ants, ‘‘reasonable notice and opportunity for
be confidential and used only for program
a hearing’ had to be provided by the board,
administration.~ This guarantee was not writ-
and the board’s decisions could then be ap-
ten into the 1935 law, but the first regulation
pealed to Federal district courts. These were
issued by the SSA Board dealt with confiden-
not onerous requirements in an era of low claim
tiality. It did not forbid all disclosures of em-
levels and “entitlement” relations with clients,
F’~r ~~mp~rison, ~dministr~ti~’e expenses for the major SS,I\ as well as high judicial deference to adminis-
programs in 1984 as percent of henefits paid, were 1,1 percc~nt trative expertise.
for old-. Age and Surti~ors Insurance, 3,3 percent for I)isahil-
ity Insurance, and 9,4 percent for Supplemental Securit~ 1 n-
come. — ——
“Altmeyer, op. cit., 196X, pp. 123-12 I. ‘Alan Wrestin and hlichael A. Baker, Data barIks in a }’”rt’t’ .%)-
‘I bid., pp. 58,70. ciet~’ (New York: Quadrangle ]Iooks. 19’721. pp. :16-:18.

HEALTHY GROWTH, 1940-71


Between 1940 and 1972, SSA enjoyed a com- The economy expanded and with it came
bination of favorable external and internal fac- steady expansion of social security. Political
tors. These 32 years were marked by sweep- elites, financial experts, and the public were
ing social change, and included three wars, generally confident that the social security sys-
cultural and ideological changes, the first tern was fiscally sound. Challenges to this be-
Republican Administration (1952-60) in SSA’S lief in the late 1940s and early 1950s were dis-
history, and organization of a union at SSA, tinctly minor dissents.
also for the first time. This was one of the first
unions for Federal workers. Collective bargain- Social security moved from a program cre-
ing was legalized in the Federal service after ated by the Democratic party opposed at its
1962. creation by the mainstream of the Republican
party, to a very broadly supported bipartisan appropriations and personnel authorizations
program. It acquired a large constituency of that it needed to keep up with the expanding
beneficiary recipients; over 25 million retired workload, and was therefore able to handle
workers, dependents and survivors, and their these changes effectively.
families, were receiving social security pay- Beginning in 1953, however, there were some
ments by 1971.
early indications of what could happen when
Congress made program changes that in-
Programs and Resources (1940-71) creased the workload, with a highly com-
After 1950, major changes were made in the pressed deadline and without additional per-
scope and nature of SSA programs: sonnel and material resources to carry out
these mandates. As a result of amendments
● The Social Security Amendments of 1950 to the Social Security law in 1950, many new
extended compulsory coverage and added claimants waited until July 1, 1952, to file their
optional coverage; benefits were increased claims, in order to take advantage of more lib-
by an average 77 percent, the wage base eral benefits computations. The workload for
was authorized to rise, and the tax rate new claims increased by 39 percent. In addi-
was allowed to rise to 1.5 percent. tion, because of further amendments in 1952,
● Expansions of the old age system became changes had to be made in the benefits amounts
a regular practice, occurring seven times for 4.6 million people already on the roles, and
between 1951 and 1965, including four in- these changes had to be made between July
creases in benefits. 18 and the issuance of September checks. In
● A new Disability Insurance cash benefits spite of this workload, the Eisenhower Admin-
program was enacted in 1956 and the age istration taking office in January 1953 sharply
limitation on disability benefits was re- curtailed the budget for the last half of the fis-
moved in 1960. cal year that had begun in July 1952, making
● Medicare was added in 1965. it impossible to add staff to catch up with the
● In 1969 Congress gave SSA adminis- backlog. In 1953 this resulted in a temporary
tration of claims related to Black Lung decline in the quality of administration and re-
disease. duction in service to the public.’
● In 1972 there was a 20-percent increase
in benefits, the wage base was increased, No such crunch took place after the Eisen-
and an automatic cost of living (COLA) hower Administration concluded that SSA was
system was added. The Supplemental Se- a well-run operation not requiring further bud-
curity Income (SSI) program was also get cuts, until 1968. In that year the Revenue
enacted but did not start until 1974. and Expenditure Control Act resulted in SSA’S
full-time work force declining by more than
These program changes called for substan- 2,000 persons in 2 years, while the workload
tial increases in SSA workloads: in opening went up by 10 percent. President Nixon then
files for newly covered workers, calculating re- ordered total Federal employment to be re-
vised benefits, and administering payments, duced by 5 percent, and all Federal agencies
and in the case of Medicare, dealing with third- to reduce the average grade level of their em-
party providers. But at least until 1968 there ployees.
was general expansion of the Federal work
force and a continuing supply of good employ- During this crunch, computers allowed SSA
ees. As a result SSA leaders saw no real prob- to cope with rising workloads; in 1971 systems
lems for the agency’s administrative respon- improvement “saved” the equivalent of 2,022
siveness in continuing growth of the social employees and $19.9 million for SSA. However
security program. the resource limitations of 1969 to 1972 were
to leave the agency in what turned out to be
SSA remained quite successful in obtaining
from Congress and the executive branch the ‘) Altrneyer, op. cit., 1968, p. 201.
97

a seriously weakened position for the expanded sions were created, not as line managers, but
operating demands and the reduced ADP to serve as “the Commissioner’s eyes and
(automatic data processing) support that were ears. 1
to unfold in the middle to late 1970s.
Labor-Management Relations (1940-71)
Management (1940-71)
Most of the successful elements of agency
A series of broad organizational changes had administration remained largely intact from
taken place in these years. The three-member 1940 to 1971. SSA remained a lifetime career
board had been abolished in 1946 and its func- service for most of its employees; at headquar-
tions transferred to a single commissioner ters, SSA had a lower turnover rate than in
under the Federal Security Administration. In any other Federal civilian agency and much
1953 SSA was incorporated into the Depart- lower than in private industry. Staff quality
ment of Health, Education, and Welfare (HEW). remained high, and mission dedication strong.
Public assistance and Children’s Aid programs Field operations maintained smooth and pleas-
were removed from SSA in 1963, leaving it ant client relations. Disability determinations
with the Old Age and Survivors Insurance Pro- were done by State agencies, and disappointed
gram and the Disability Insurance Program. claimants did not therefore generally see SSA
Two years later SSA was reorganized, when employees as their adversaries. Through these
it was given responsibility for the new Medi- years, customers were always right, and the
care Program. customers and the taxpayers were considered
to be the same people.
Between 1946 and 1972 there was steady
growth in executive branch oversight of SSA By the late 1960s, however, there were some
operations and plans. The relocation of SSA signs that the “pioneer period” of dedicated
first to the Federal Security Administration employees was shifting into a new, more com-
and then to HEW began to limit SSA’S previ- plex phase of employeeemployer relations. The
ous degree of autonomy. The saving condition permanent work force almost tripled from 1959
—secretary-level satisfaction with the agency’s to 1972. For a time at least many old-line SSA
administration-was only as good a shield as employees feared that the new recruits would
SSA performance was strong. When that fal- not share the agency’s deep-seated public serv-
tered, after 1973, secretarial protection could ice ethic. The influx of new employees made
evaporate swiftly. it harder to give intensive, personalized train-
ing. New social values, including suspicion of
After the 1965 reorganization, SSA still had
large organizations, were widespread in soci-
a mixture of program and functional units.
ety. Employees were becoming more assertive
Four bureaus operated the four major pro-
about their rights and more demanding in
grams: Retirement and Survivors Insurance,
terms of working conditions. In the long run,
Disability Insurance, Health Insurance, and
however, SSA has tended to have a high de-
Federal Credit Unions. A single centralized
gree of employee loyalty and commitment com-
recordkeeping organization handled databases
pared to other public and private sector orga-
and data processing for all programs, and had
nizations.
both systems analysis and operations compo-
nents. ’” Specifications for new systems came A 1962 Kennedy Executive Order author-
from the program bureaus, and systems coordi- ized collective bargaining in the Federal serv-
nation and advanced planning were in the Of- ice. Previously the American Federation of
fice of Administration. Ten Regional Commis- Government Employees (AFGE) had repre-
sented fewer than 5 percent of SSA employ-
1f)
Jack S. Futterman, The Social Securit.v Administration “s ees. After 1962 AFGE had 2,500 members out
Recent Reorganizations and Related Administrative Problems,
report to the National Commission on Social Security, July 28,
1980 (unpublished), pp. 9-13. “Ibid, p. 13.
98
— —

of 11,000 headquarters employees, and by 1971 ware that was needed to program and operate
40,000 of SSA’S 52,000 employees were cov- the new EDP systems. Data communication
ered by union contracts, although probably capabilities also expanded, as teletype systems
only a quarter of them were union members. came on the market, and then on-line input and
The union began raising issues of adverse retrieval of data through telecommunications.
working conditions, sex and race discrimina- Finally, microfilm printing became available,
tion, and technology impacts. with major possibilities for a massive manual-
records-based account-number operation.
SSA’S top management saw itself as pro-
union, based on SSA’S strong alliance on so- By the end of the 1960s and early 1970s man-
cial security policies with the labor movement. agement of all large organizations were pre-
However stresses in labor relations surfaced sented with a group of key decisions:
by 1965 that were harbingers of later fissures.
for which files was it cost-effective and or-
A comprehensive article in the Baltimore ganizationally important to automate;
SUZZ12 in 1966 identified these problems: whether to go from batch processing to
interactive, on-line systems for high-
c bad working conditions, especially over-
volume operations;
crowding;
whether to concentrate mainframe com-
● changes in work force dedication, and loss
puters in one data center or create regional
of missionary spirit about social security;
data centers; and
c disaffection among clerks, who consti- whether to create a communication net-
tuted half of the 11,000 headquarters
work or stay with mixtures of telephone,
staff, and particularly among women and
teletype, and physical transportation.
the 21 percent of clerks that were black
(a fact that SSA, which had been a leader SSA had a number of technological choices
in hiring blacks for office work in the and decisions to make:
1930s, had difficulty in realizing); and ● to stay with the dominant IBM system,
● concern over automation—many clerical
or adopt competing systems, which could
and production employees felt they were
mean extensive reprogramming;
“economic units” who served the ma-
to retain SSA early tape media or move
chines. to new higher density and higher speed
storage, which required new tape drives
Technology and Procurement Policy and some changes in job control language,
(1940-71) but was not an enormous task;
● when to move from early software pro-
From 1940 to 1954 there were only modest
gramming such as COBOL, to higher or-
enhancements of electrical accounting machin-
der languages, which had advantages but
ery and microfilm capabilities for SSA to con-
would be expensive to reprogram; and
sider. Then came the EDP (electronic data ● how to keep state-of-the-art systems and
processing) revolution, beginning in the early
programming staff.
1950s with first-generation computers, mov-
ing into second-generation computing in the What needs stressing is how much such de-
late 1950s, and reaching third-generation ma- cisions were a matter of art rather than sci-
chines in the 1960s. With the revolution in cen- ence. In the 1950s and 1960s many Federal
tral data storage and processing capacities agencies mastered that art and were at the fore
went major related changes in data collection front of successful information technology ap-
and input-output mechanisms, and in the soft- plications: the military, the FBI, the Census
Bureau, the Internal Revenue Service, and
‘ ~A&m r
Spiegel, ‘*The Giant in Woodlawn,” l)arts I-I\ , l;al- SSA, which was still among the leaders in EDP
timom %n, Apr. 25-28, 1966. applications.
99

In this period there were major changes in In 1965, Congress enacted Public Law 89-
procedures for procurement of Federal com- 306, usually called the Brooks Act. Because
puters. The securing of budget authorization of concern about the overwhelming dominance
for large EDP acquisitions had come under of IBM in Federal computer purchases, the act
HEW, the Office of Management and Budget, required full competitiveness in hardware ac-
and congressional scrutiny by the early 1960s, quisitions and attempted to limit sole-source
as the costs of equipment became substantial. purchasing. The General Services Administra-
But these reviews generally extended only to tion (GSA) was designated to supervise and
determining the need for and timing of expend- monitor EDP acquisitions. Under certain con-
itures. SSA was able to define its needs and ditions, GSA could give an agency Delegated
then enter into special relationships with lead- Procurement Authority for large procure-
ing vendors in the accounting machine, com- ments. SSA was then almost completely an
puter, and microfilm industries. The vendors IBM shop, although there was one RCA-301
were not only anxious to get the high-volume in the Central Office and one in each of the six
business, but also to have the prestige and pub- program service centers; thus SSA would soon
licit y that came from having their equipment have to justify continued acquisition of IBM
selected by SSA. computers to skeptical scrutiny.
A special relationship had developed be- SSA was still a user of leading edge infor-
tween IBM and SSA in the first years, 1935 mation technologies throughout most of this
to 1939, and became even more important from period. Successful innovation was possible be-
1940 to 1965. IBM was the leading vendor of cause management placed high priority on ac-
punch card systems, and worked to provide curate recordkeeping, advance planning for
special applications for SSA. From 1950 to new technology was well institutionalized,
1965 IBM was the dominant vendor of first-, there was an effective technical staff, and there
second-, and third-generation computers. Fed- was a generally sound balance between pursuit
eral agencies were often “90 percent IBM. of new technology and attention to operational
For SSA, IBM provided first-class briefings performance. The agency was sensitive to the
and plans and justifications with which to ap- human-factor impacts of new systems, and
proach Congress on expenditures. generally had employee, and union, support.
Some examples of SSA adoption of new in-
As more and more IBM computers were in-
formation systems during this period were
stalled at SSA, assuring the compatibility of
first-generation EDP equipment in 1955 for
new computer acquisitions with existing oper-
posting, benefit computation, reinstating in-
ating systems became a key procurement need,
correctly reported earnings items, and statis-
leading to the adoption of still more IBM com-
tical work; the development of the microfilm
puters. IBM’s interest was not in conflict with
printer (linking computer and microfilm tech-
SSA’S independent technical judgment. The
nology) in 1959; and automatic card punching
custom software programs written in SSA to
equipment, in 1963.
handle their specialized operations were still
adequate. The concept of large data-processing The need for systems integration was rec-
facilities centralized in one location was the ognized by the late 1950s. SSA was able, based
prevailing wisdom as the best way to maximize on its good service performance and popular-
the utilization of expensive hardware. SSA’S ity in Congress, to have its case for continued
approach was paralleled by what leading in- acquisition or upgrading of its IBM computer
surance companies and banks were doing. As stock accepted by GSA and the Brooks Com-
of 1965, then, SSA centralized, batch-proc- mittee. It did move into purchases of several
essed data operation both met SSA’S needs UNIVAC computers for administration, and
well and also suited IBM marketing strategy. to General Telephone & Telegraph for a very
100

large communication acquisition, which mod- number of SSA field offices increased by about
erated its total reliance on IBM. 50 percent during these years.
SSA profited from making its systems oper- When the workload rose heavily and stead-
ations highly visible to the public. This tradi- ily, in the late 1960s, advance planning often
tion of good work, well advertised, served SSA became a casualty of the need to keep opera-
well with three major constituencies: the pub- tions from falling behind. The timetables for
lic, as taxpayers and program participants; starting new efforts could never be kept.
Congress and the White House; and its own The reliance on “homegrown” programmers
work force.
and systems experts also began to have costs
As of 1972, SSA did not yet have what would in this period. Because of constant operational
today be called a computer system. It was still demands there had been no substantial repro-
a paper operation assisted by EAM (electronic gramming of software. In the early and mid-
accounting machines) and EDP machines. File dle 1970s, in private firms and some Federal
folders and microfilm records of account ap- agencies, substantial resources were devoted
plications were the primary source of determi- to revising software as new techniques of soft-
nations and responding to inquiries. SSA, in ware engineering emerged. SSA did not do this.
its Golden Age, was still a well-organized, well- By 1972 SSA was well behind the leaders in
staffed, and well-led machine-assisted people both the private and public sectors in that in-
system. creasingly critical aspect of total EDP man-
agement.
Emerging Problems (1965-72) This growing weakness was not yet appar-
ent outside the agency. Through the heroic use
During the “Great Society” years of the of accumulated people, and organizational and
1960s, the Johnson Administration relied systems resources, SSA’S service delivery still
heavily on SSA to implement social welfare met program demands and client expectations.
programs. Strong emphasis was put on estab- However, SSA was falling steadily behind in
lishing “an SSA presence” close to the client, anticipation of systems overload, people-
to make it easier for the aged and disabled to machine balances, technical procurement work,
talk with SSA representatives face-to-face. The and top management actions.
Chapter 6

Deepening Problems, 1972-81


Contents
Page
The SSI Crisis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Expanding Programs and Congressional Oversight . . . . . . . . . . . . . . . . . 103
The Supplemental Security Income Crisis, 1973 . . . . . . . . . . . . . . . . . . . . 104
The Political Environment and SSA Resources . . . . . . . . . . . . . . . . . . . . . . 106
Work Force Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Technological Choices. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11O
Privacy and Security Concern s....... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Disruptive Reorganizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
The 1975 Reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
The 1977 Reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
The 1978-79 Reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Deficiencies of Information Technology Management . . . . . . . . . . . . . . . . . 115

Figures
- N ~ Page
8. Growth inthe Social Security Administration’s Workload From 19’75
to 1983 As Defined by the Beneficiaries Per Staff Year . . . . . . . . . . . . 108
9. The Organization of the social security Administration Circa 1972 . . 112
IO. The Organization of the Social Security Administration
Following the1975 Reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
11. The Organization of the social security Administration
Following the1977 Reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
12. The Organization of the social security Administration
Following the1975 Reorganization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Chapter 6

Deepening Problems, 1972-81

Between 1972 and 1981, the Social Security get (OMB); the General Services Administra-
Administration reached a state of crisis. Thi_s tion (GSA); Presidential commissions; the Na-
term was used, and flat statements that pri- tional Research Council; and a multitude of
mary operations were faltering, were voiced experts, consultants, and clients. What they
publicly at the end of the decade by: SSA man- shared was a common judgment that SSA was
agement, union leaders, overseers in Congress; in near collapse as an effective government
the General Accounting Office (GAO); the De- agency, and that the disheveled state of the
partment of Health, Education, and Welfare ADP (automated data processing) systems was
(DHE W); the Office of Management and Bud- at the heart of that perilous condition.

THE SS1 CRISIS


Expanding Programs and 1980,1 the Reagan debt collection initiative of
Congressional Oversight 1981,’ and the Omnibus Reconciliation Act of
1981, all of which meant that reprogramming
There were three major streams of action by was necessary for calculation of benefits. To
Congress pertaining to SSA between 1972 and implement the Cost of Living increase in 1981
1981. First, there were 15 new laws making required changes in 600 software programs,
changes in the I?etirement and Survivors In- because as written they could not accept four
surance Program and Disability Income Pro- digits (that is, any benefit amount of over
gram; four of them made significant alterations $999). The adjustments required by the 1980
in determination of entitlements and benefits. Disability Amendments meant that changes
Secondly, Congress gave SSA a major new pro- had to be made in over 880 programs.”
gram to administer: the Supplemental Secu-
rity Income Program, which took three Fed- When it was impossible to do the calcula-
eral/State programs formerly administered by tions through EDP (electronic data process-
ing) procedures, SSA was forced to do them
the States (payments for the aged, blind, and
manually, at heavy costs. There was, accord-
disabled poor) and converted them into a fed-
ing to a Senate report, “constant shifting of
erally administered program. Finally, SSA was
management priorities and the coming and go-
given additional support and assistance pro-
ing of new policy initiatives. $
grams to administer (such as the energy and
refugee programs).
‘The 1980 amendments mandated disahil~t> reint’t~~t ig:i-
Following the tradition begun in 1935 to tions, producing large-scale rernok’als of disahlllt!’ clit,nts and
1939, Presidents and Congress continued to a flood of judicial appeals, which e~entu; ill}’ nullified most of
reject the concept of universal flat benefit pay- the exclusions.
~The Administration insisted on aggressi~e collt~c.ti{)n of c’r-
ments such as many other nations used, with roneously made payments, as descritwd in ch. :1.
minimum administrative complexity, in favor ‘U.S. Congress, The Social Securit~’ .4 utomated l)ata Ih-ot-
of a mixed insurance and welfare system, with essing S~rstern Crisis, a report prepared h~’ the staff of t hc~ Sul)-
committee on Sociaf Security, House Commit& on li’a)s and
highly complex entitlement and benefit for- Means, 97th Cong., 1st sess., May 22, 1981, p. 6; hereafter cited
mulas. After 1972 benefit levels embodied both as I louse Subcommittee on Social Securit~ (title, datel.
automatic cost of living adjustments (COLA) ‘U.S. Conh7ess, .Sociaf Securit~r: How U’eli 1s It $+r~ing t ho
Public? I I earing Before the Semite Special Committee on ?i~ting,
and periodic adjustments and readjustments, 98th Cong., 1st sess., h’o~r. ’29, 1983, p. 138; hereafter cited as
such as the Social Security Amendments of Senate Special Committee on Aging (title, date).

103
104

The time provided by Congress for SSA to ance administered by the States, although
make changes in programs or institute major partially funded by the Federal Government.
new programs proved again and again to be Federal social security benefits were to be de-
inadequate. Sometimes SSA commissioners termined by past earnings; the State-admin-
were following stem Administrative directives istered programs were to distribute public
when they told Congress that they could take assistance on the basis of need. Other insur-
on new programs or changes; sometimes they ance and assistance programs for the blind and
had underestimated what it would take to ac- for the disabled were created by the 1935 act
complish the new work on time, while main- and later amendments. The assistance received
taining basic services and accurate perform- by the needy varied considerably from State
ance. Partly this tendency to accept unrealistic to State, in spite of Federal contributions, and
deadlines without demur was a function of the in some States their income remained far be-
commitment of SSA leadership to social secu- low poverty levels.
rity programs and to meeting what they saw
In 1972, amendments to the Social Security
as acute needs, and the SSA tradition of get-
Act repealed these State-administered assis-
ting nearly impossible jobs done through he-
tance programs for the aged, blind, and dis-
roic manual efforts. Partly, it represented a
abled and replaced them with a new Federal
weakness in advance assessment of work re-
program, Supplemental Security Income (SS1),
quirements.
which became effective on January 1, 1974,
But the situation also reflects two generic, to be administered by SSA. SS1 was intended
or structural, problems in congressional over- to be a basic national income maintenance
sight. The statements that executive agencies system, administered in a manner compara-
can make to Congress about their resource ble to the way in which the Retirement and
needs or their management problems must al- Survivors Insurance, now called Old-Age, Sur-
ways be vetted by the Administration and pass vivors, and Disability Income Program, was
through the filter of Presidential policy and administered. 5
OMB directives. In addition, some congres-
Under the States, monthly payments to an
sional committees and their staffs may lack
individual with no other income varied widely,
the knowledge and experience to understand
from $75 to $250; the new SS1 program was
the limitations of and the resource demands
to provide a flat minimum income, originally
posed by highly complex operations and highly
set at $130. Eligibility requirements had also
complex technological systems.
varied; SS1 was to have minimum barriers to
The Medicare Program, added to SSA in eligibility except for lack of capability to earn
1965, had been handled successfully. Most of other income, and to have fairly generous pro-
those who had to be enrolled were already visions for disregarding other forms of income
beneficiaries of the retirement program; the such as help from one’s children. This “flat
biggest task was working out procedures for grant” approach encouraged Congress to sup-
deducting the Medicare Program from their pose that the new program could be adminis-
benefits and for delivering payments to a serv- tered much like existing SSA programs.
ice provider. Although these were complicated
But since the Federal grant would be less
tasks the agency adjusted relatively smoothly.
than some recipients were getting in some
This was not the case, in 1973, with the Sup-
States, States were allowed to continue (or to
plemental Security Income Program.
choose to give) supplements to the basic grant.
SSA would administer and deliver the State
The Supplemental Security Income
Crisis, 1973 5SW u s
. ConWeSS,
. The Supplemental Security Income Pro-
gram, a report of the staff to the Senate Committee on Finance,
The same act that in 1935 established the 95th Cong., 1st sess., April 1977; hereafter cited as Senate Fi-
SSA also created a program of old-age insur- nance Committee (title, date).
105

supplements, since they had to have the same A Senate Finance Committee reportG con-
criteria for eligibility as the Federal basic cluded that:
grant, and would therefore be only add-ens.
(The) initial problems far exceeded the nor-
Taking over this program turned out to be mal concept of start-up difficulties. . . . The ca-
a traumatic shock for SSA, and a dozen years pability (of SSA) to adapt its existing mecha-
later some employees and some outside ob- nisms and procedures to the new program was
servers think that morale at the agency never greatly overestimated. As a result, the re-
fully recovered. There were two kinds of closely sources that were provided—both human and
material-proved inadequate to the task. The
related problems—systems problems and pub-
time allotted between enactment and imple-
lic relations problems, and together they shook mentation proved insufficient. . . .
the confidence of, and in, the agency.
Why did this happen? The Senate report said
SSA had 14 months to set up the SS1 pro- that at the time of the legislation,
gram after the 1972 legislation, although the
grandfather clause (assuring that no one lost . . . it did not . . . appear to be an unreasonable bur-
eligibility for assistance because of the change- den. Representatives of the Social Security Admin-
istration . , . indicated no doubt about their ability
over) and other amendments were added al- to do the job.7
most at the end of this time. The agency had
decided that the new program could not sim- SSA leaders had wanted for some time to
ply be integrated into its existing processes, see federalization of this program for the needy
but required a more highly automated commu- aged, blind, and disabled. SSA district offices
nication system to link district offices, which in hundreds of communities and SSA’S ad-
would deal with clients, to headquarters, where vanced computer operations were arguments
their participation in other Federal insurance for federalizing the administration of the pro-
programs would have to be checked. The new gram. SSA had, well before the legislation, cre-
system (SSA Data Acquisition and Retrieval ated two staff units to plan for and facilitate
System, or SSADARS) was inaugurated at the such a transfer from the States.s The plan-
same time as the new program, which was ning units developed a concept of “assistance
probably a mistake. Before, field offices had centers” to be located throughout the coun-
not used interactive terminals at all; claims try. Another option was for an interactive com-
data were sent to headquarters by teletype. munication system which would allow exist-
With SSADARS there was on-line query and ing field offices to function as ‘‘assistance
response, but the one to four terminals per of- centers’ by having fast access to claimants’
fice were operated in the “back room” by data- or applicants’ social security records. This
entry technicians, and their machines were planning was however almost completely in-
often down for several hours, or for a day at effective because SSA did not have the re-
a time. The communication terminals quickly sources, nor provide the authority, to develop
became a bottleneck in processing the claims. or test either option until the legislation was
There were severe startup problems, and in passed, and in fact, there was considerable
addition the new system was quickly over- doubt that the legislation would pass until the
loaded. This resulted in long waiting lines at very end of the congressional session. Suddenly
district offices, massive backlogs, and high it did, and SSA had 14 months to get ready.
error rates. Claimants often waited for hours
only to be sent home at closing time, to return ‘Ibid., p. 27.
another day. The need for highly trained per- ‘Ibid., p. 26.
‘Ibid., p. 26. According to SSA there was an ABDA (Aging,
sonnel for the system had been grossly under- Blind, and Disabled) Planning Staff in Baltimore, and a Yt’el-
estimated. Staff overtime skyrocketed. fare Reform Planning Staff in Washington.
106

The implementation of SS1 was in any case subject to challenge and controversy. A quote
a massive undertaking, made more difficult by from a high-level SSA officialg is illuminating
factors beyond SSA’S control. As late as 2 here:
weeks before the program was to begin, nearly
People came in, sat down, and negotiated
a dozen States had not decided whether to pro- how much they were going to get. And that
vide State supplemental payments, which SSA really wasn’t what we were about. Our motto
would be obliged to administer. As it was, even had been. . . “you get every penny that com-
with its backlog growing and long lines of wait- ing to you, not one cent more, not one cent
ing claimants, SSA was criticized for inade- less. ” But the clients-they were coming out
quate outreach because the number of appli- of an environment. . . where they had a nego-
cants was smaller than earlier estimates. tiated benefit. And in January 1974 they
would come into an office that has a suppos-
The public relations problem, and the em- edly fixed benefit structure . . . but it could
ployee morale problem that resulted from it, vary on forty-five different variables, plus
were perhaps as predictable but more unavoid- mandatory State supplement. . . .
able than the systems problems. The expecta-
tion that SS1 administration would be like that So SSA representatives found themselves, in
of other SSA programs and could be handled effect, negotiating. SS1 claimants by defini-
tion had no other source of income, and were
with traditional efficiency was unrealistic. The
program was very different from other SSA often in desperate straits, needing and demand-
ing emergency funds, and in no mood or con-
programs in the demands it placed on the agency.
dition to be patient with bureaucratic delays,
Retirement and survivors benefits were mat-
however inevitable.
ters of earned right and were set by formulas
based on lifetime earnings. SS1 benefits were The authority for granting benefits had to
set through individual determinations and re- be left almost entirely in the hands of field of-
quired SSA to ask a number of personal ques- fice employees who interviewed the applicants,
tions. The assumption had been that claimants with quality assurance resting on review of a
would be predominantly needy elderly, much small number of the cases. There was a rash
like SSA’S other clients. But the proportion of lawsuits challenging SSA procedures. Some
of assistance beneficiaries made up of the dis- observers believe that SSA was so traumatized
abled had been growing rapidly before the shift by the introduction of SS1 in 1974, under in-
to a Federal program; so that 80 percent of ap- adequate staff resources, that its operations
plications came from (and 70 percent of new would have been badly shaken even if com-
awards went to) the disabled, who then made puter and systems failures had not also taken
up nearly half of the total beneficiary popula- place.
tion. Claims processing for the disabled is
much more complex than that for the aged, ‘From the proceedings of a workshop held by OTA during
requires a higher level of expertise, and is more the course of this study, Mar. 5, 1986.

THE POLITICAL ENVIRONMENT AND SSA RESOURCES


As one congressional committee put it in external factors forced SSA into a corner. For
1981, the key questions were what had caused example, did OMB or cabinet-level policies con-
the SS1 crisis, and why nothing had been done tribute to the debacle? Were congressional
about it by SSA over the years that the likeli- directives or oversight procedures at fault ei-
hood of such a situation was developing. One ther in contributing to the problems, or fail-
also needs to ask whether the problems and ing to bring them to light before they became
the failure to attack them effectively, were severe? Answers to these questions could dis-
solely failures of SSA management, or whether close generic problems in the management of
107
— —

government agencies in a period of continu- back the government work force was seen as
ing technological change. The answers neces- a necessity for sound fiscal policy and effec-
sarily involve political, social, and resource tive government administration.
factors.
Further, in the 1970s, emphasis on humaniz-
From 1973 to 1981 periods of “stagflation” ing and enriching work began to collide with
and a series of recessions produced cutbacks the efficiency thrust of many automation ef-
in basic industries, significant blue-collar un- forts. After 1979, this was to become a power-
employment, and mounting national budget ful concern of the union representing SSA’S
deficits, which reduced resources for financing employees, an issue about which union leaders
social programs at the same time that there would increasingly seek to become involved.
were rising needs for such services. Increased
Shortage of resources was a key factor in
utilization of benefits and a growing imbalance
both the operational weaknesses and the poor
between current users and paying supporters
ADP performance between 1974 and 1982.
had created fundamental questions, by the late
1970s, about the financial soundness of the SSA was already weakened by the 1969 to 1973
cuts in personnel levels and budget, with field
Social Security Trust Fund system and the ca-
staff and headquarters staffs strained to the
pacity of the Social Security System to con-
limit.
tinue paying its own way. The bipartisan con-
sensus under which SSA had operated since When it was given the SS1 program in 1972,
1937 came under serious challenge. SSA received approval to increase its field per-
sonnel, but these resources proved to be wholly
Under Presidents Carter and Reagan, Fed-
eral agency requests for appropriations and insufficient. It was estimated, SSA officials
remember, that the States had together 32,000
staff authorizations were cut back, ways were
people employed, whose work was to be shifted
sought to curtail the expansion of program ben-
to SSA. It was assumed that 10,000 temporary
efits, agency operations were monitored more
hires would suffice for SSA, since about 70 per-
closely, and campaigns were initiated to cur-
tail “fraud, waste, and abuse” in Federal oper- cent of the claimants would already be on the
social security rolls. 1’) The results were delays,
ations.
gross overpayments and other high error rates,
In spite of this, SSA programs continued to confusion in operations, and general disarray
expand in coverage and benefit levels, and new in the field offices. Both employees and outside
programs were assigned to SSA. Even when critics maintained that SSA had completely
changes were made limiting SSA programs, misestimated the amount of labor required to
in 1980 to 1981, these further increased admin- work the system. But SSA requests for more
istrative demands on SSA. American society people had been repeatedly refused. ’
had become accustomed to swift and sophis-
On top of this came two successive high-
ticated information-handling capacities and
demand assignments from Congress: the 1977
SSA as an “advanced user” of information
Social Security Amendments and the 1980 to
technology was expected to achieve a high level
1981 legislation. In between, Congress, in a
of service.
1978 attempt to reduce paperwork for em-
It was widely believed in the 1970s that orga-
nizations applying the new office technologies 1OT~e~e ~iWre~ ~e]y on the memory of the responsible SSA
would not have to layoff large numbers of officials, but they are at least approximately correct.
1 I For ~xmp]e, one SSA employ~ told OTA~ ‘‘It (OMB) to-
workers, but could direct them into other ex- tally underestimated the labor intensiveness of the SS1 pro-
panding operations. But by the early 1980s for- gram-how much work that it would really involve. And I would
eign competition began to force business man- submit that still to this day they do not understand and do not
estimate correctly the labor-intensity of delivering personal serv-
agers to use automation to shrink work force ices to clients. ” A DHHS high-level official remembers that:
size as sharply as possible. There was a paral- .,
. . . OMI? kept coming back and saying, ‘Cut personnel, ’ and
lel approach by government leaders. Cutting . . . ‘Drop personnel and we’ll worry about that later, “
108

ployers, mandated a change from quarterly re- Figure 8.—Growth in the Social Security
Administration’s Workload From 1975 to 1983
port of earnings to annual reporting. For SSA As Defined by the Beneficiaries Per Staff Year
this meant a change from a quarterly cycle in
its workload to an annual peak early in the
year, which was harder to manage. The earn- 28 - - 830
ings reports are central to the computation of 26 - \\ - 810
benefits, and if they are not posted promptly, 24 -
\\
~ - . . - 790
%% ~ e
other work tends to back up. Eventually SSA g 22 - - 770 u
had a 3-year backlog of unposted earnings. (u
g 20 - - 750 :
8
SSA again failed in a series of key efforts 3 18 -
- 7303
to obtain adequate resources. ]2 A request for 16 . - 710
12,000 new permanent positions resulted in ap- (right workload)
14 - - 690
proval for only 10,000 temporary jobs. In 1977 --- Disability
12 - (left workload) . 670
Congress voted on the personnel resources
SSA sought, a 2 percent rise in total staff, but 1 1 I I 1 1 1 1 i 1 I
1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984
the Civil Service Reform Act just then levied
Year
a complete personnel freeze in the Federal serv-
SOURCE: Kenneth Laudon, “SSA Management and Use of Information
ice. SSA’S work force shrank by 7 percent from Technology the Current Period. 1982 -1988,” contractor report for
1977 through 1980 and the proportion that OTA, 1985.

were part-time and temporary workers rose


slightly. ]3 legislative changes; SSA veterans claim that
In spite of the governmentwide personnel they consistently begged for more people and
cuts and freezes of 1981 to 1982, SSA’S work were refused. It appears that throughout this
force was, by 1983, 5 percent larger than in period OMB applied heavy pressure to agen-
1980. But the ratio of beneficiaries to staff- cies to reduce their work forces. There were
years had grown by 15 percent (figure 8). Con- however serious weaknesses in SSA’S top man-
gress consistently authorized higher staff lev- agement between 1973 and 1981, as discussed
els than OMB and the Department of Health in the next section.14 Whatever the reasons,
and Human Services permitted. If there had SSA was always running hard to get its work
been a marked improvement in ADP and com- done. . . and falling. The agency pushed its peo-
munications support, the increased workloads ple in ultimately self-defeating ways to make
would not have resulted in heavy “burnout” up the difference,
. and lost the quality staff it
pressures for staff or in degraded service to once enjoyed.
clients. But the combination of inadequate per- Congressional and executive branch confi-
sonnel and inadequate or even counterproduc- dence in SSA’S management clearly eroded.
tive ADP systems were compromising basic Weak program delivery, poor quality, doubts
delivery of services. about fraud and waste, and bungling of ADP
There is considerable disagreement as to activities brought efforts in the executive
where the blame for this situation lies. Congres- branch and congressional committees to
sional staff tend to assert that SSA consist- remedy these problems. ]5
ently misestimated or inadequately projected ADP facilities were still another troubled
the resource requirements of new programs or area; the computer facilities in SSA’S Opera-
tions Building suffered from inadequate-elec-
lzs~~~te speci~ committee on Aging, swid *curity: How
Well 1s It Serving the PubZic? 1983, p. 131. “1bid.
‘“At the end of 1977 SSA’S total work force was 87,500, of ‘5U. S. Congress, Mismanagement of SSA Computer Syst-
whom 7,300 were part-time/temporary workers, At the end of ems Threatens Social Security Programs, 33d report by the
1980, the total was 81,700, with 7,200 part-time/temporary. Sen- House Committee on Government Operations, 97th Cong,, 2d
ate Special Committee on Aging, Social Security, How Well Is sess., 1982; hereafter cited as House Committee on Government
It Serving the Public? 1983, p. 131. Operations (titIe, date).
.

109

tricity and air-conditioning, limited fire pro- Means Committee that the move was a year
tection, and overcrowding. A new computer behind schedule due to construction prob-
center in Baltimore was authorized by the Ford lems. lG However, the move was completed
Administration on SSA’S promise to formu- during 1981 with no serious disrution of day-
late and implement a plan for ADP develop- to-day operations.
ment. No such plan was implemented, but be-
tween 1976 and 1980 a new computer facility
was constructed. In 1978a move into the build-
ing under construction was approved on con- ‘6Written response to questions, from Richard S. Schweiker,
dition that a plan to facilitate competitive Secretary of Health and Human Services, May 28, 1981, to Con-
gressman Pickle, Chairman of the Subcommittee on Social Secu-
procurement had been developed. From 1979 rity, following Hearings, U.S. Congress, Automated Data Proc-
to 1980 the work to move old computer hard- essing Systems, Hearing Before the Subcommittee on Social
ware into the new building caused implemen- Security and the Subcommittee on Oversight of the House Com-
mittee on Ways and Means, 97th Cong., 1st sess., May 22, 1981,
tation of new ADP systems to be tabled, and p. 51; hereafter cited as House Committee on Ways and Means
in May 1981, SSA told the House Ways and (title, date),

WORK FORCE PROBLEMS


Personnel problems became troublesome in Office of Systems Development, no division
these years. For 15 years, SSA had promoted chief had a college degree, and of 400 profes-
into computer and systems jobs former claims sionals in the division, only two dozen had ad-
clerks and computer operators who were given vanced degrees, none in relevant subject
minimal training and lacked the fundamental areas. ’g A former Associate Commissioner for
knowledge and skills needed to stay abreast Systems told Congress that in this situation,
of changing technologies. 17 Then, for reasons “retraining is not the answer. 20
to be discussed later, SSA was unable to at-
Many of those who had only on-the-job train-
tract sufficient newly educated programmers
ing were highly competent at their jobs, but
and systems experts to upgrade its staff, and
this did not necessarily equip them for concep-
suffered heavy attrition from the most talented
tualizing new approaches to highly complex
of those it did hire, as they encountered ad-
technological problems, or give them the
verse working environments, heavy overtime,
knowledge necessary to foresee emerging tech-
low pay scales, and assignments on antiquated
nological possibilities and ways of pushing for-
systems that offered no possibility of profes-
ward the state of the art. SSA had developed,
sional growth or satisfaction.
or fallen into, a policy of giving promotions
By the late 1970s, middle managers in the strictly on the basis of seniority, rather than
Office of Systems were typically former claims training, credentials, or merit. 21 This policy
clerks who had learned on the job but had no had, and probably still has, the effect of build-
formal training in advanced systems. ’8 In the ing in those who rise through the ranks to deci-
— — sionmaking positions, a fierce loyalty to the
‘TU. S. Congress, Viabih”ty of the Social Security Admim”stra- agency. However, it tended to frustrate the at-
tion Computer Systems, Hearings Before a Subcommittee of
the House Committee on Government Operations, 97th Cong.,
tempt to attract and hold bright and ambitious
1st sess., Sept. 23, 1981. (This was the Subcommittee on Legis-
lation and National Security, chaired by Rep. Jack Brooks.) Here- lgTestimony of Rhoda M~chur, former Director Of the Of-
after cited as “Brooks Committee” (title, date). fice of Systems Development, SSA, in app. 111, of the House
1~House Committee on Government Operations, Mismanage- Committee on Government Operations, 33d report, cited in foot-
ment of SSA Computer Systems Threatens Social Security note 18.
Programs, 33d report, 1982, p. 9, which quotes Dr. Jan Prokop, ‘Dr. Jan Prokop, in testimony quoted in the House Committ-
former SSA Associate Commissioner for Systems. His testi- ee on Government Operations, 33d report, cited in footnote 18.
mony appears in Brooks CornmI“ttee, Viabti”tyof the Social Secu- *lAccording to congressional testimony by Dr. Prokop and
rity Administration Computer Systems, 1981, pp. 127 ff. Ms. Manchur, cited in footnote 18.
110

people more recently trained in computer sci- sonnel to work on its systems because that
ence and eagerly sought by industry, where would entail a huge displacement problem and
they got not only higher salaries but the op- consequently would be unacceptable. “23
portunity to work on state-of-the-art systems, The fear of this displacement, or of being
to continue to build their skills, and to advance
downgraded, was pervasive among SSA staff
rapidly in their profession.
after the reorganization of the Office of Sys-
Commissioner Svahn testified before a con- tems in 1979. With the Reagan Administra-
gressional committee that he was “under no tion’s budget-cutting initiatives in 1981, the
significant artificial impediments to hiring, ” Office of Personnel Management directed that
but had serious problems in recruiting and re- many ADP positions be reviewed for possible
taining professionals. Svahn blamed this on reclassification-that is, for reduction in grade
serious morale problems arising from “six-day level and salary. The possibility of adverse
work weeks for six months at a time, ” rather personnel actions magnified the already seri-
than on SSA’S promotion policies or its ob- ous problem of job uncertainty and low mo-
solescent systems. zz rale, as acknowledged by another congres-
sional report.24
A congressional report noted another factor,
that SSA: “cannot hire enough qualified per-
%:lHouse CommittW on Government operations, 33d rePort~
————. . . . 1982, p. 9.
~~p~~p~ed statement by John A. Svahn, Commissioner of “U.S. Congress, The Social Security Automated Data Proc-
Social Security, for the Subcommittee on Legislation and Na- essing Crisis, a report prepared by the staff of the Subcommit-
tional Security of the House Committee on Government Oper- tee on Social Security, House Committee on Ways and Means,
ations, Sept. 23, 1981. See Brooks Committee, Viability of the 97th Cong., 1st sess., May 22, 1981, p. 9; hereafter cited as House
Social Securit.v Administration’s Computer Svstems, 1981. Subcommittee on Social Security (title, date).

TECHNOLOGICAL CHOICES
Most large organizations during this period 1969 to 1973 insurance companies were invest-
grew to depend heavily on ADP systems for ing heavily in technology, and their work forces
their basic daily operations, and the capabil- were also growing. Employment in the insur-
ity of resorting to manual backup grew weaker. ance industry showed strong growth during
Aging computers from the 1960s, less efficient the 1970s; insurance companies were diversify-
than newer systems, were a common problem, ing their products and expanding their mar-
and when they were replaced, it was necessary kets, while at the same time they were just
to undertake the software conversion of data- learning how to use the technology to increase
bases and instructions. Organizations had productivity. Beginning about 1979, these
more options in designing their information productivity gains began to show up in lower
systems; but this was also a period of rising unit costs of service delivery, in constrained
expectations as to what information systems work force growth, and more recently in work
should be able to do in the near future. force reductions.
Private sector insurance companies were au- Researchers agree, however, that the com-
tomating their procedures during this period. panies that were most successful in using
Studies of this industry” indicate that from advanced systems tended to be: 1) relatively
———. .—
‘bEileen Appelbaum, Technology and the Redesign of Work of City and Regional Planning, 1983; Barbara Baran, Techno-
in the Insurance Industry, Institute for Research and Educa- lo~”ctd Innovation and Deregulation: The Transformation of the
tional Finance and Governance, Stanford University School of Labor Process in the Insurance Compa.v, Berkeley Roundtable
Education, Project Report No. 84-A22, November 1984; Bar- on the International Economy, Contract No. 433-3610.0, pre-
bara Baran and Suzanne Teegarden, Women Labor in the In- pared for the Technology and Economic Transition Project, Of-
surance Office, University of California, Berkeley, Department fice of Technology Assessment, U.S. Congress, January 1985.
111

small companies, and 2) those that took a “bot- In communications, the late 1970s and early
tom-up” approach to planning and implemen- 1980s saw the arrival of free-standing packet
tation. SSA, with its mammoth size and work- switching networks with their own host com-
load, compounded its problems by holding to puters, separate from the database processors.
a thoroughly “top-down” approach to planning Processing capabilities could be distributed
and decisionmaking. according to varying loads and priorities. The
networks constituted a utility by which trans-
Organizations had to be increasingly adept,
actions and messages could be shipped around.
anticipative, and technically well-staffed and
well-led to stay abreast, and the costs in dol- Managers in most organizations had to be
lars and performance of falling behind were convinced by technical experts that it was nec-
growing heavier. The choices involved hard- essary to hire systems and programming staff
ware, data storage, software, and communi- with the new software engineering knowledge,
cations. upgrade staffs, retrain supervisors, bring in
IBM mainframes dominated the large sys- consultants, and spend substantial amounts
of money to apply these resources to software
tems market, but IBM began to stop main-
conversions. For Federal ADP shops operat-
taining older systems. More IBM-compatible
ing under a combination of civil service and
mainframes became available to organizations
personnel classification controls and budget
that had IBM software. In the mid to late
limitations on large-scale software projects, the
1970s organizations could move from tape to
decision to modernize older computer systems
new disk storage, but changes had to be made
did not come easily.
in job control language and in applications pro-
grams. Software was the critical element; the Throughout this period, SSA was falling be-
development of modern database management hind. The extent of this slippage will be illus-
depended on separating programs from data, trated later, but SSA failed to keep up with
that is, making the database independent and the private sector in hardware, and more im-
usable by multiple programs. portantly, in software development.

PRIVACY AND SECURITY CONCERNS


By the mid-1970s Americans wanted and Integrity: Internal quality control and audit
gradually got regulation over the way infor- requirements; computer-matching proj-
mation about individuals was handled by pri- ects (since the late 1970s) to deal with
vate and public organizations. From 1973 to fraud or waste in benefit programs.
1981, SSA faced a growing set of requirements Due Process: Federal court decisions set-
for protecting data from misuse: ting information and procedural require-
● Privacy and Confidentiality: The Federal ments for SSA determinations, particu-
Privacy Act of 1974, amendments in 1975
larly in the disability area.
Information Management: Paperwork
and 1982; OMB circular A-108; the Tax
Reduction Act of 1980; Information Re-
Reform Act of 1976; the Paperwork Re-
duction Act of 1980.
source Management requirements under
● Freedom of Information and Public Ac- OMB supervision.
cess: The Freedom of Information Act To meet these requirements an organization
amendments of 1974, 1976, 1978, had to be in effective command of its ADP sys-

Security: OMB Circular A-71 and Federal tems in terms of both operations and advanced
Information Resources Management Reg- planning; such management command of ADP
ulations; GSA regulations. was simply not present at SSA in this period.
112

As will be seen, there were repeated occur- complaints of violations of privacy related to
rences of computer-related fraud and sabotage social security data.
at SSA. However, there were no significant

DISRUPTIVE REORGANIZATIONS
In 1972 to 1981 frequent changes took place Figure 9.—The Organization of the Social Security
in top leadership and unsuccessful agency re- Administration Circa 1972
organizations. In the first 38 years SSA had k t # 4 b
six commissioners, with an average tenure of
6.5 years; and two men led the agency for 27
of the 38 years. From 1973 to 1981, SSA had
seven commissioners or acting commissioners,
for an average tenure of 1.1 years. None of the
r 1
confirmed commissioners had experience with-
in SSA or was directly knowledgeable about Office of the Actuary
it. The senior staff was also shaken up re-
peatedly as many of the new commissioners
i
Bureau of
Administration I
brought in their own senior people. As former
Associate Commissioner for Administration
Jack Futterman noted in a report for the Na-
i
Office of Research
& Statistics
I i Office of Public Affairs I
tional Commission on Social Security in
1980, 26 the direction of SSA by its Commis-
sioner could never be the same as in earlier eras.
No new Commissioner could, from personal ex-
perience within the agency, know the whole
organization and its “enormous range of pro-
grams, administration, management [and]
technology. ” All Commissioners would be
“more dependent on key subordinates” and
i
Bureau of Hearings
& Insurance I
“would need to make large delegations of au- Bureau of
thority. “ The sheer increase in size had taken
I
Disability Insurance Bureau of Retirement
a toll. & & Survivorship
i
Two commissioners in the mid to late 1970s
decided that fundamental reorganization of the
agency was the way to gain control (see figure
9). SSA had in fact three major reorganiza-
i
SOURCE”
Bureau of
Health Insurance
Social Security
I
Administration 1966

tions: in 1975, by Commissioner James Card-


well; in 1977, as part of a general HEW de-
partmental reorganization; and in 1979, by The 1975 Reorganization
Commissioner Stanford Ross. Every major Commissioner Cardwell, who had no SSA ex-
analysis of SSA’S performance in this period perience, concluded that there was insufficient
stresses the disruption and adverse effects accountability for program operations, that the
that these reorganizations had on agency oper- Commissioner was forced to resolve too many
ations. conflicts between programs, and that diffusion
of responsibility was a major source of trou-
ble. The 1975 reorganization therefore elimi-
26 Futtermm , Op. cit., 1980, P. 13
nated the separate line organizations for the
113

Figure 10. —The Organization of the Social Security Administration Following the 1975 Reorganization
#
Advanced Operating 4 Office of the
Systems Staff Commissioner
*
I
I
Associate
7 ? I I 1 r I I
Associate Associate Associate Bureau of
Commissioner Commissioner Commissioner Bureau of
Commissioner Hearings
Management and Program Policy Program Health Insurance
and Planning External Affairs Operations and Appeals
Administration
I
T I I 1 1
Bureau of Bureau of Bureau of Bureau of
Retirement and Field Supplemental Bureau of
Survivors Disability
Operations Data Processing
Insurance Insurance Security Income
b
SOURCE Social Security Admlnlstratlon, 1986

Retirement Security Income, Disability In- The 1977 Reorganization


come, and Supplemental Security Income pro-
grams and merged these, along with the staffs In 1977 the Department of Health, Educa-
from the former Bureau of District Office Oper- tion, and Welfare, which included SSA, was
ations and the offices of the 10 Regional Com- reorganized. Medicare and Medicaid were
missioners, into one large Office of Program merged and put under anew HEW Health Care
Operations (OPO). This reduced the number Financing Administration for which SSA took
of senior staff reporting to the Commissioner. on important recordkeeping functions. SSA
A special Office of Advanced Systems (OAS) was now to administer the Aid to Families with
was created to develop better computer sys- Dependent Children (AFDC) and the Refugee
tems; this unit reported directly to the Com- Assistance Programs, and the Commissioner
missioner (see figure 10). A Policy Council of SSA was designated as director of Child
made up of the heads of first-line units was cre- Support Enforcement. AFDC was assigned to
ated to recommend new policies. a new SSA Associate Commissioner for Fam-
ily Assistance (OFA), which meant that SSA
Fundamental problems arose with this orga- field offices now reported both to him and to
nization between 1975 and 1979. The Futter- the Office of Program Operations. The Com-
man report cited above, based on extensive in- missioner now had to resolve boundary dis-
terviewing of SSA people, concluded that the putes and resource issues between the two
reorganization was never completed; large offices 27 (see figure 11).
numbers of employees were never reassigned,
or were left in jobs that no longer existed, and The 1978-79 Reorganization
issues about the jurisdiction of senior officials
were never resolved. The new Office of Pro- In 1978 Commissioner Ross was appointed
gram Operations (O PO) established a large new with instructions to tie SSA more closely to
level of staff superimposed on and duplicat- HEW policy direction. There was another
ing the staff of the three former program bu- sweeping change in the agency organization.
reaus. Neither the OPO staff nor the bureau The Commissioner’s Office was reorganized
staff could be effectively held accountable for
results and performance. ‘-Ibid.
114

Figure 11 .–The Organization of the Sociai Security Figure 12.—The Organization of the Social Security
Administration Foiiowing the 1977 Reorganization Administration Foiiowing the 1978.79 Reorganization

Office of Office of
Office of the Chiid support Office of the Commissioner Office of Chiid
Advanced
Systems Commissioner Enforcement Support Enforcement
Deputy Deputy
Commissioner Commissioner
8I for Programs for Operations

Associate Commissioner for ?


Program Operations - t
Bureau of Data Processing
I
m
Associate Commissioner,
t Public Affairs
Associate Commissioner for
■ Management & Administration Bureau of Retirement and
Survivors insurance Associate Commissioner,
Government Affairs I

Associate Commissioner for
Program Poiicy & Pianning Bureau of
m
Disability insurance Associate Commissioner,
Policy
I

i
Associate Commissioner for
Externai Affairs I - Bureau of Suppiementai
Security income

Director
“ Bureau of Hearings & Appeals
Regionai Commissioners
I 1
. Associate Commissioner,
4 Assessment

Associate Commissioner for


i} Famiiy Assistance I
J
SOURCE Social Security Administration, 19SS
i
Associate Commissioner,
Systems I
and two Deputy Commissioners (for Opera-
tions and for Program Policy Issues) were in-
stalled. The rest of SSA was rearranged into
a new “functional structure, ” with 10 offices,
i
Associate Commissioner,
Hearings and Appeais
I
each headed by an Associate Commissioner.28 Associate Commissioner,
The 10 Regional Commissioners were retained, i Family Assistance I
reporting directly to the Commissioner. The
Office of Advanced Systems was abolished,
leaving SSA with no independent systems Regional Commissioners
(lo)
planning effort (see figure 12).
SSA operations were thus grouped around
general administrative functions rather than
around major programs, so that all of the same
Associate Commissioner,
Central Operations
I
.—
~HTh~Y ~e~e: Policy, Management, Budget, and personnel;
Public Affairs, Governmental Affairs, Assessment, Systems,
Operational Policy and Procedures; Central Operations, Hear-
ings and Appeals, Family Assistance. SOURCE. Social Security Administration
115
— — -———

kinds of administrative procedures would be Now it was up to the Commissioner to co-


conducted by a specialized unit for all SSA pro- ordinate a dozen Associate Commissioners and
grams. This ran counter to 40 years of SSA 10 Regional Commissioners who reported di-
experience, by dividing up program segments rectly to him. An additional feature of this re-
even more than had the 1975 reorganization organization was that Commissioner Ross de-
and scattering them through functional offices. liberately overrode internal career-promotion
The Futterman report said, “It became almost lines in selecting top managers, reaching down
an impossibility . . . to render a current to promote staff and bringing in outsiders.
accounting of the status of RSI, DI, or SS1.
. . . “ However, it paved the way for agencywide
automation and system redesign in the 1980s.

DEFICIENCIES OF INFORMATION
TECHNOLOGY MANAGEMENT
Information management is not different in modernization plans, and relied repeatedly on
kind from general administration; it is still fun- a form of organizational blackmail: “only we
damentally dependent on the overall direction know how to run old programs, ” and “give us
of ideas, people, material resources, and or- what we want or we can’t get the checks out
ganizational structures and processes. Be- next month. ”
tween 1972 and 1981, SSA had experienced
While this criticism may be slightly exag-
a profound change of mission and operating
gerated, vestiges of these attitudes are still
culture with the onset of the SS1 program.
clearly discernible; many long-time SSA man-
Then came an on-line system (SSADARS),
agers still react with strong emotion to offi-
with high-pressure, fast-turnaround require-
cial assessments of SSA performance that were
ments, which was a dramatic and often re-
presented to Congress by the Commissioner
sented change in the basic work system. As
and his management team in 1981-82, saying
already noted, SSADARS did not work well.
heatedly that “things weren’t that bad” and
New performance measurement pressures on
that backlogs and error rates were overstated
field staff further worsened morale, by most
and exaggerated.zg Whether or not this is
accounts.
true (and all evidence indicates that the situa-
The internal awareness of SSA’S deepening tion was indeed very bad and worsening, re-
problems, and the strong sense of comitment gardless of the accuracy of certain indicators
and loyalty to the agency that SSA had long presented to congressional committees), the
enjoyed, unfortunately combined to produce dispute points again to the increasing difficulty
an extreme defensiveness on the part of many
SSA people toward any outside criticism. To
those in oversight roles and to other external ‘Whe descriptions of the state of affairs in the late 1970s and
early 1980s are based in large part on SSA documents, espe-
observers, it often appeared that SSA people cially the 1982 Systems Modernization Plan, and on testimony
“circled the wagons” and fended off any sug- by SSA officials at congressional hearings during that period.
gestions for basic changes, maintaining that These descriptions were confirmed by many people inside and
outside of SSA who were consulted by OTA. But in written
glacial incrementalism was the only feasible comments to OTA on an early draft of this case study and in
way to improve patchwork systems. many discussions, SSA officials repeatedly disputed statements
taken from those documents. In explanation, some pointed out
The charge was and frequently is made by that the documentary statements in question were assembled
SSA’S critics that the operations staff in the and used in 1982 ‘‘by the new management team’ in defending
requests for appropriations for systems modernization, ‘‘over
late 1970s and early 1980s consistently were the bitter protests” of those at SSA who had been “satisfac-
hostile toward outsiders brought into develop torily coping with the problems, ”
116

and complexity of effective oversight of very ADP system supported by staff professionals
large data-handling operations. could make SSA work. Yet budgeting and plan-
ning within SSA treated hardware, software,
Accepting the fact that SSA was having se- and telecommunications not as the core need,
vere problems in carrying out its mission, the the structural necessity for doing the work, but
tasks of top management in this situation as a peripheral service supporting “oper-
were: ations.
● to improve the existing systems, and to Sustained management interventions would
get or hold on to efficient equipment and be needed to regain top management control
effective personnel; of an organization in which bureaucratic pa-
● to carry out the planning of major new sys- thologies had taken hold and were dominat-
tems, developing a rationale for reorganiz- ing all reform efforts.
ing jobs, people, and structures;
● to institutionalize this planning and sys- Though it never failed in these “crisis years”
tems development in such a way that to get the monthly beneficiary checks out—
would not be frustrated by, and would not which was accomplished by heroic efforts by
on the other hand interfere with, the heavy SSA staff, given the disarray of manual and
daily requirements of carrying on opera- computer systems— serious problems had de-
tions; and, therefore, veloped with the quality and timeliness of SSA
● judiciously to allocate resources between services. This had produced areas of signifi-
operation~ needs and new system devel- cant client dissatisfaction. Privacy Act require-
opment and resolve conflicts over that al- ments for “accuracy, timeliness, and complete
location. ness . . . to assure fairness, ” were not being
met. Court-defined requirements for due proc-
Top management did not accomplish these ess in hearings and appeals were often not
four tasks. According to people within SSA forthcoming. Key information needed on a
at the time, Commissioners were frequently timely basis for disability hearings was often
told by senior staff that changes “just weren’t not available. Security and integrity procedures
possible. ” With frequent changes and short were found by executive and congressional au-
tenure, commissioners lacked the depth of dits to be weak or nonexistent. Procurement
knowledge of operations to challenge those policies and compliance with procurement
statements. Teams of outside specialists were monitoring were seriously weak and key
hired and then defeated by insiders. Plans were procurements had gone awry. Morale in the
made but not implemented. No effective sys- field, district offices and service centers had
tem was developed for specification of user re- fallen seriously, and key units at central head-
quirements. System development groups could quarters felt similarly demoralized by the suc-
not discover the basic functional requirements cessive reorganizations, leadership shifts, and
they needed to work with. Budgeting for ADP project failures or abandonments.
was not done in a way that specified the rela-
tionship of expenditures to operations and mis- As a result, SSA by the end of 1981 had lost
sions or to meeting specific information pol- the reputation for excellence in performance
icy requirements. that had been its hallmark from 1935 to 1972.
With its well-publicized problems came a loss
SSA’S mission had greatly expanded in the in confidence in SSA at DHHS, OMB, GSA,
1970s and its staff had grown from 50,000 to the White House, GAO, and key congressional
75,000 people between 1970 and 1975. It had committees. By having failed to use informa-
reached a cross-over point, at which it could tion technology effectively to cope with seri-
no longer be run effectively with manual proc- ous problems in its external and institutional
esses, even aided by computers and older elec- environments between 1973 and 1981, SSA’S
tromechanical equipment. By 1975, and cer- basic ability to carry out its assigned missions
tainly by 1981, only an effective and integrated was now in jeopardy.
Chapter 7

The Beginning of the Systems


Modernization Plan, 1982
Contents
Page
The Dimensions of the Crisis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Developing the Plan . . 121
The Data Processing “Environment:: :::::::::::::: :::::::::::::: : 121
The Paradyne Affair.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
The Paradyne Contract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Failures inthe Paradyne Implementation . . . . . . . . . . . . . . . . . . . . . . . . . 127
SSAFailures in Managing the Paradyne Project . . . . . . . . . . . . . . . . . . . 129
Other Factorsin the Paradyne Contract Problems . . . . . . . . . . . . . . . . . . 129
An Unfinished Story. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Aftereffects and Implications forSMP . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

Table
Table No. Page
4. Si.ze ofWork Forceof Social Securitv Administration, 1975-84 ..,.... .124
w
Chapter 7

The Beginning of the


Systems Modernization, Plan 1982
By fiscal year 1982, the Social Security Ad- –Income Survey for the Department of
ministration (SSA) had 260 million names in Health and Human Services on Federal
its account number files, and was maintain- program participants,
ing 240 million earnings records. It was pay- —Recordkeeping of vested rights in pri-
ing $170 billion annually to 50 million benefi- vate pension benefits,
ciaries. It had 88,000 full-time, part-time, and —Information for the Internal Revenue
temporary employees, 1,344 field offices, 10 Service on employer annual reports for
regional offices, 32 teleservice centers, 6 pro- income tax enforcement, and
gram service centers, 3 data operations centers, –Other minor responsibilities.
the Baltimore headquarters complex, and a
The magnitude of SSA operations was im-
new computer center under construction. ]
pressive. SSA was in 1982’:
SSA programs included:
maintaining 240 million records on per-
● Income Support Programs: sons with an active social security ac-
—Retirement and Survivors Insurance count, or their survivors;
(RSI), paying monthly benefits to over 50 mil-
–Disability Insurance (DI), lion people;
–Supplemental Security Income (SS1), issuing 10 million new Social Security
and cards annually;
–Aid to Families with Dependent Chil- posting annually 380 million wage items
dren (AFDC). reported by employers;
● Other Social Service Programs: receiving 7.5 million new claims applica-
–Black Lung Disease Claims (BL); tions each year;
–Health Insurance (Medicare), shared processing 19 million postadjudicative
now with the Health Care Financing Ad- transactions annually, including 2.5 mil-
ministration; lion benefit recomputations; and
–Food Stamps (for SS1 participants); handling more than 120 million bills and
—Low Income House Energy Assistance; queries from private health insurance in-
—Refugee Assistancez; and termediaries, carriers, and providers.
–Child Support Enforcement.”
SSA was, however, by its own admission in
● Administrative Services for Other Federal
1982, only “marginally capable of performing
Agencies’:
critical program functions. In nearly all
–Assistance to Selective Service for draft
areas there were serious problems. Both SSA
registration,
and Congress now realized that action must
‘Social Security Administration, Annuai Report to the Con- be taken, and soon.
gress for Fiscal Year 1981; also Social Security Administration,
Office of Systems, Systems Modernization Plan: From Sum”val
to State of the Art, SSA Pubs. No. 41-002, Baltimore, MD, 1982,
hereafter cited as SSA: 1982 SMP. Another source, the Social
Security Bufletin, 1984-85, tables 14, 69, 174, 175, says that
in cafendar year 1981 SSA paid $124 billion annually to 36 mil-
lion OASI beneficiaries plus $6.5 billion SS1 payments to 4 mil-
lion beneficiaries. fU, S. Congress, Social Securit-y: How Well Is It Serving the
‘This program reimburses State and local governments for Public? Hearings Before the Senate Special Committee on Aging,
refugee programs. 98th Cong., 1st sess., Nov. 29, 1983; hereafter cited as Senate
,?A ~ro=m t. collWt payrnent9 due for child suPPort. Special Committee (title, date). Testimony by the General
‘These programs constituted about 10 percent of total SSA Accounting Office, pp. 38-39.
workload. ‘SSA: 1982 SMP, p. I-4.

119
120

THE DIMENSIONS OF THE CRISIS


There were major problems in service deliv- SSA operations were no longer cost-effective:
ery and in making operations cost-effective. ● it was having to meet most legislative
In terms of service delivery7:
changes in programs through manual
● issuance of new numbers and cards now processing, often overtime, and at serious
took 4 to 6 weeks; costs to other operations;
● SSA was 3 years behind in recomputing ● to implement Cost of Living Adjustment
retiree’s benefits to credit them with ad- (COLA) increases required 20,000 hours
ditional earnings, and backlogs had grown of computer processing, day and night
to half a million items; over a period of 4 monthsa;
● claims processing operations were behind ● SSA itself argued, using GAO estimates,
schedule 50 percent of the time and pay- that using programmable terminals in
ments and notices to beneficiaries were only 4 of the 10 labor-intensive functions
delayed; that it was hoping to automate would re-
● SSA was 3 years behind in posting the 380 sult in “savings of over 1,000 years, rep-
million annual wage items reported by em- resenting $133 million in savings, after
ployers, and over $69 billion in unposted taking into account the costs of adding
items had accumulated by 1982; these additional processing capabilities.”9
● checks totaling $60 million were mailed Problem elements in the data-processing sys-
to 8,000 people who had been dead for at
tems in 1981 involved hardware, data storage,
least 2 years;
software, data communications, personnel, and
● there was a 3 month backlog of data
facilities; in short, all elements of the system
needed to notify employers about incor-
were in trouble, as will be described in the fol-
rectly reported employee earnings;
lowing section. Procurement practices were,
● annual cost-of-living increases processing
at best, inept. SSA’S practice had been to ex-
forced suspension of all other processing press its mission requirements in terms that,
for 1 week each year;
in effect, made IBM the only competitor. GAO
● large backlogs in processing Medicare
advised Congress that SSA did not have the
claims caused payments for services to be
expertise to develop sound procurement strat-
badly delayed; egies based on mission requirements. In 1978,
● systems security failed to meet minimum
at the request of the Brooks Committee, GSA
standards for Federal agencies;
put a hold on SSA’S computer acquisitions un-
● SSA was over 2 years behind in enforce- til they could be reviewed; subsequently 300
ment operations to detect overpayments;
out of 500 were canceled.10
● computer procedures to detect potential
fraud were not able to be done regularly;
and
● overwork and alienation of workers was
high, tapes were deliberately destroyed 8SSA: 1982 SMP.
and equipment sabotaged, with 46 acts of ‘Ibid., p. 1-19.
1°House Committee on Government Operations, 33rd report,
willfull vandalism reported between 1977 1982, p. 11. See also U.S. Congress, General Accounting Of-
and 1981. fice, Solving Soa”al Security’s Computer Problems: Comprehen-
sive Corrective Action Planning and Better Management 1s
7HOUSe Cofitt= on Government Operations, Mism~age-
JJeeded. A report by the U.S. Comptroller-General to the Chair-
ment of SSA Computer Systems Threatens Social Security man of the House Committee on Government Operations, HRD-
Programs, 33rd report, 1982, p. 4. 82-19, Dec. 10, 1981. Hereafter cited as GAO (title, date).
.—

121

DEVELOPING THE PLAN


1 n anticipation of asking Congress for nearly
$150 million to rebuild SSA’S information
systems, SSA’S new Commissioner, John V.
Svahn, painted a dire, bleak public picture of
its situation. Some who had been SSA man-
agers for a long time now say that the situa-
tion was never as bad as it was portrayed, but
that in order to build support for a large mod-
ernization program it was necessary to go
along with the public posture that disaster was
near. To some extent, the extraordinary defen-
siveness of SSA since 1982 to outside criticism
can be attributed to these tensions.
Those who were struggling with the prob-
lems on a day-to-day basis understandably
want to emphasize that SSA continued to cope.
Those who were determined to make a new
start may even have misrepresented some de-
tails; from the outside, it is not possible to pin
all of these down, In some sense, these details
are now unimportant; the situation was clearl}
bad, and the critical questions for government,
and particularly for Congress, were why did
it become so bad? and how can this situation
be prevented in the future for SSA and for
other governmmt agencies’?
In the rest of this section, therefore the em-
phasis is on three questions: 1 ) why was SSA
in a crisis’? 21 how did it get in that situation?
and 3) who was m a position to know- was
Congress warned that the situation was de-
veloping”?

The Data-Processing Environment


122

in SSA expending two-thirds of its computer of these did not have internal standard labels
resources (230 work-years annually) on soft- to allow the computer to check on whether the
ware maintenance-not redesign but changes proper tape was being run. This increased the
in old codes in order to fulfill new information level of errors.
requirements.
Many of the major production jobs were de-
Only a handful of SSA people knew how a signed to operate on only one specific computer
large number of the computer programs oper- or were too large to run on other computers.
ated; as these people retired or left SSA a sig- The lack of adequate hardware meant that very
nificant amount of the code was no longer little computer time was available for testing
maintainable by the remaining staff. and development of new programs.
These problems also bedeviled many large SSA failed to meet its computing require-
private sector organizations in the mid-1970s, ments 45 to 75 percent of the time, each month
but SSA was about 5 years behind private in- in 1982. According to Svahn, SSA estimated
dustry in making important technological tran- that its gross computing capacity require-
sitions. ments in 1982 approached 5,000 central proc-
essing unit (CPU) hours per month. The maxi-
Hardware mum capacity of the computers was 3,000 CPU
In 1982, SSA was operating outdated, un- hours per month, and staffing levels would sup-
reliable, and inadequate hardware. Of the 26 port only 2,000 CPU hours. Program analysts,
large-scale computers, 23 were supporting operators, and managers operated systems on
program-related operations and 3 processed an overtime basis to process critical workloads,
administrative workloads. SSA operated 11 while backlogs continued to mount.12
IBM 360/65 systems in its Program Service
Centers (PSCS) and central offices, and two Telecommunications
UNIVAC 1108 systems in Baltimore. The Field offices need timely access to data
UNIVACS had not been manufactured or mar- stored and processed at the central computer
keted for 10 years; their operating costs were facility to take claims for benefits and to proc-
more than $3 million, compared to $1 million ess changes. The telecommunications system
for more modern equipment. The IBM 360/65 had evolved over the previous 15 years, since
systems were first produced in the 1960s. SSA SSA entered into an interagency agreement
also operated an IBM 370/165 and an IBM with the General Services Administration
370/168, which were 10 years out of date and (GSA) in 1966 to be a prime user of its Ad-
no longer manufactured or marketed.ll vanced Records System (ARS), a teletype net-
work. The SSA telecommunications system of
Since this hardware was no longer supported
1982 included:
by the manufacturers, SSA had to contract for
costly third-party maintenance. This hardware ● three types of data-entry terminals: ARS
contributed to about 25 percent of the produc- teletypewriter equipment, SSA Data Acqui-
tion jobs having to be done over, wasting ap- sition and Response System (SSADARS),
proximately 30 percent of the available com- and interactive video display units in lo-
puter processing power. cal offices, plus other key-to-disk record-
ing equipment in the program service
A great deal of labor was required to load,
centers;
unload, and catalog the magnetic tapes. Each ● concentrators (telecommunications mini-
month, 30,000 production jobs required man-
computers which receive data and query
ual handling of 150,000 tapes. About one-third
messages and send them to a main host
computer);
1 ITe~timony of Commissioner Svahn, in 1981, in Brooks
Committee, Viabih’ty of the Social Security Administration
Computer Systems, 1981. “SSA: 1982 SMP, p. 1-3.
123

● modems and local communication lines on 500,000 reels of magnetic tape stored in a
connecting SSADARS terminals to the vault on portable pushcarts; tight scheduling
concentrators; and a great deal of labor (200 people, or more
● high-speed trunk lines connecting the com- than a third of the operations staff) were dedi-
municators and front-end processors; cated just to handling the tapes and getting
● front-end processors that interface be- the reels into use. Physical disintegration of
tween trunk lines and host computers and the tapes, plus human error, caused a high
translate between them; number of failures and subsequent reruns.
● the host computers, already described; About 24 percent of CPU hours were lost in
and this way each month.
● SSADARS software (communications
Data was stored on tapes at 1,600 bytes per
and applications programs).
inch (bpi), a very low density compared to com-
When built in 1974, SSADARS consisted of mercially available 6,250 bpi drives. It was
two IBM 370/165s, and was designed to han- organized by programs, with many data ele-
dle 20,000 inquiry-response transactions and ments repeated from one program to the next,
80,000 data transactions per day. It was satu- and there were more than 1,500 separate pro-
rated a year later and required updating to grams. There was no formal data dictionary
370/168 computers. Since then teleprocessing with standard definitions of all data elements
has grown by 500 percent. comprising the SSA databases. The same data
elements (e.g., earnings) were labeled and de-
By 1982, SSADARS had old, inadequate
scribed differently indifferent programs, which
concentrators, insufficient communication cir-
made for confusion.
cuits, and obsolete front-end processors. It
suffered overload, frequent failure, absence of These transaction processing systems are
manufacturer support, unavailability of re- the foundation for higher level systems, which
placement parts, and extended outages. Dur- in many large organizations include manage-
ing the first half of 1981, the system was down ment information systems and decision sup-
11 percent of working hours and 88 percent port systems. The former are systems designed
of the downtime was due to hardware fail- to support middle and senior-level manage-
u r e . Field office staff had to come in on ment by providing routine reports on opera-
weekends to key in data that SSADARS was tions. In modern organizations, information
too overloaded to accept during the week. By needed for management is often routinely co-
1982 there was little capacity remaining in off- pied from transaction files into a management
peak periods to handle current workloads. In information system file that allows managers
other words, workload could not be shifted to to access it through personal computers or
off-peak hours; high traffic peaks occurring in some network arrangement. In SSA, the trans-
peak load time had to be backlogged, and en- action data was not generally available to
tire streams of communication were frequently managers because it was on magnetic tape, and
lost, requiring rekeying, which meant that all requests for reports had to be funnelled
transmission time was lost while messages through central processing. There might be de
were rekeyed. This resulted in printing back- lays of up to several years in the production
logs ranging from 10,000 to 100,000 messages of reports needed to manage decisionmaking
at a time. and control. There was no management infor-
mation system and no plans to develop such
Database a capability.
Methods used for the storage and organiza-
Personnel
tion of fundamental SSA data were about a
decade behind the times, in 1982. Data was There was constant pressure from OMB
under several Administrations to constrain or
IJSSA : 1982 SMP, p. 2-16. reduce the size of the work force (see table 4).
124

Table 4.—Size of Work Force of Social Security with them much of the knowledge of the patch-
Administration, 1975.84 work software. Only 21 of them were replaced.
— —
F u l l - t i m e p e r m a n e n t ‘- ‘ - SSA’S 1982 System Modernization Plan (SMP)
Staff on duty Temp;rary and noted that:
Fiscal year end of yeara part-time Total
b
The full impact of ADP staffing losses is
1975 . . . . . . . 78,400’ 7,300 85,600–
1976 b . . . . . . . . 78,400 7,300 85,600
more serious because the knowledge of patch-
b
1977 ... . . . . 80,300 7,300 87,500 work software is lost due to the Iack of docu-
1978 b . . . . . . 78,600 7,100 85,700 mentation. New recruits cannot be prepared
1979b . . . . . . . . 76,300 7,600 83,900 adequately for the maintenance of undocu-
1980 . ., . . . . 74,500 7,200 81,700 mented programs and systems using archaic
1981 . . . . . . . . 74,600 9,700 84,200
1982 . . . . . . . . . 74,800 11,300 86,100 programming techniques.
1983 d . . . . . . . . 76,000 9,900 85,900
1984 d . . . . . . . 75,800 8,000 83,800 SSA says that in 1981 entry-level program-
aFlgur~~ ma; not acjcj across due tO rounding mers got 6 weeks of training; some remember
b Re p r e s e n t t e Ievels of employment Derived from subsequent Year’s
approxima

appropriations’ justfftcatton
that it took about a year for them to learn
clncludes 6000 employees who had 2-year term appointments at that time
dThese are ~stlmates derived from 1984 appropriatlons’ justification
enough to perform adequately.
NOTE This table provides actual and estimated levels of employment for the So-
c Ial Security Admln lstration, and does adjust for various reorgan Izations Computers had also changed the work of the
and shifts In agency responslbllity, e g , the transfer of Medicare respon.
stbilttles to the Health Care Ftnanclng Admintstratlon and adoption of rest of SSA’S staff. Over one or two decades,
AFDC and child support enforcement program functions the amount of material a claims representa-
SOURCE Alan Westtn
tive had to master had enormously increased.
As one employee said:
Both Congress and OMB reasoned that invest- Now I am (expected to be) not only an ex-
ment in automation should be justified in pert with respect to retirement and survivors’
terms of increased productivity, defined as a benefits, and disability benefits, but how to
saving in labor costs. By about 1980, private make all those work in a computer system.
industry (e.g., the insurance industry) had be- From a Claims Manual of three volumes, that
gun to realize these gains in lower labor costs I started from, now (I have) no less than 20
per unit, but these gains showed up only some volumes, half of which are systems instruc-
years-at least a decade—after the companies tions. . . . Claims reps have long since given
first began to build a modern data manage- up trying to keep track of rules and regula-
tions and law. Now you are only dealing with
ment infrastructure. That infrastructure was instructions.
not yet in place at SSA.
Labor relations were, according to both man-
Perhaps even more important was the fail-
agement and labor, at an all time low. In 1979,
ure of SSA to maintain and upgrade the skills
the American Federation of Government Em-
of its computer specialists relative to the rap- ployees (AFGE) proposed a consolidated bar-
idly advancing state of the art of computer sci-
gaining unit and SSA agreed. The parties bar-
ence, or to attract the best of the crop of new gained for 18 months over a contract which
graduates in this field. There was no adequate
finally went to arbitration. After 23 days, an
program in SSA for replacing experienced pro
agreement was signed, in 1981. According to
w arnmers who were about to retire, or for train- management, labor was using charges of un-
ing new staff. In 1981, the agency lost 112 of
fair labor practices to stall improvement in
its 560 experienced programmers1 4; they took
operations —in 1 year, AFGE filed over 800
ld&.cording t. the SSA: 1982 SMP, p. 1-15. However, SSA charges of unfair labor practices. According
now disputes this (in written communication to OTA) saying to labor, management failed to take into ac-
that: “In 1981 total losses in the 334 series (which includes com- count the interests of the workers when design-
puter prograrnm ers and systems analysts) was 71, not 112. This ing and implementing new systems, especially
is one below the average yearly loss for the period 1981-1985.
In 1981, new progr ammer trainees totalled 155, higher than in quality of worklife issues and employment im-
any subsequent year. ” It is possible that Mr. Svahn exagger- pacts. Both management and labor agreed that
ated, but SSA was certainly feeling the scarcity of competent unless there were drastic changes in the cli-
programmers in 1981-82.
mate of distrust that prevailed in 1982, the de-
125
-—

velopment of new information technology The 1982 SMP also warned that because of
would intensify the strife. deficiencies in controlling access to records and
to the telecommunication network, SSA was
Security vulnerable to fraud, abuse, and sabotage. It
Privacy protection, physical security, ac- noted that there had already been “limited
countability, prevention of abuse and fraud, instances of fraud and abuse perpetrated by
and backup and recovery capability had also its employees . . ." and that ‘‘some instances
suffered from lack of coherent management. of sabotage causing the destruction of equip-
SSA had poor physical control at its facilities ment and tape files have occurred in the past,
and few audit trails to determine who in the and could be repeated by disgruntled em-
agency initiated actions, either on paper or by ployees working under increasing workload
computers. There was no systematic method pressures. . . .”
for communication among various programs,
Planning and Management
so that an individual could obtain multiple ben-
efits under multiple programs without over- SSA’S most critical weakness was its in-
payments being detected. The 1982 SMP doc- ability to gain management control over infor-
ument noted that due to computer processing mation resources and systems. SSA itself rec-
backlogs and faulty programs, duplicate pay- ognized that it:
ments were often made, and “the computer
. . . had not yet undertaken the management
backlog has reached the point where SSA can-
initiatives necessary to insure adequate con-
not carry out its earnings enforcement opera- trols over the development, operation, and
tion (a primary overpayment detection mech- maintenance of its systems. ...17
anism) nor employ automated means to detect
conditions indicating potential fraud. “*5 SSA had an explicit and well-institutional-
ized advanced systems planning group in the
Another sign of poor management control 1940s. But by the mid-19S0s, each program
was the inadequacy of systems backup and re- bureau was independently working on plans
covery plans, which were limited to storing co- and development of its own systems, without
pies of master files in an offsite storage area. regard to agencywide considerations. In the
An SSA document in 1982 warned: late 1950s, another central systems planning
. . . SSA’S systems operate without any unit was formed, with a broad charter to de-
backup in the event of critical damage, or velop concepts for advanced system and inves-
worse—a catastrophe. . . . Although backup tigate the technology to move the agency
files are available to some extent, they are not toward that system. This appeared to be work-
duplicates. The destruction of a large number ing fairly well until the mid- 1960s. 18 But dur-
of key tapes would probably result in an in- ing the late 1960s advance planning was usu-
ability to produce payment tapes. . . . Should ally sacrificed to the need to deal with recurring
a major disaster occur, untold billions of dol- crises. A former SSA official recalls that:
lars could be lost as a result of SSA’S computer
and communications systems being out of The heads of the two main program bureaus
commission for up to a year. 16 would withdraw people from systems planning
and put them into current operations work. , .
since those jobs just had to get done, I tried
to keep the advance planning staff working
1~SSA: 1982 SMP, p. 1-7. SSA says that it did have an an- ahead as much as possible, but there really was
nual operation called MAFDUP which identified potential dupli- a kind of blackmail at work—Operations needed
cate Title I I payment situations and alerted processing center
personnel to review the affected folders.
‘hSSA: 1982 SMP, p. 1-18. SSA, however, now says that it ‘TSSA: 1982 SMP, p. 1-14.
maintained backup copies of all master files in a secure storage ‘hJack S. Futterman, The Social Security Administration
area; these backup files were not in fact duplicates; and restor- Recent Reorganizations and Related Adxninistratit’e Problems,
ing master files would ha~’e been difficult, expensive, and time- report to the National Commission on Social Security, July 28,
consuming. 1980, unpublished.
126

people to get the changes done and the checks zation this office was decimated. GAO
out, and we couldn’t deny them the re- recommended 20 that the planning for infor-
sourcesolg mation systems be assigned to a separate, in-
The planners, in any case, had no resources dependent component reporting directly to the
to begin to implement any of their concepts; Commissioner (as the Office of Advanced Sys-
those resources would have to come from oper- tems had done).
ations budgets, and the operations people were Shortly after Commissioner Svahn was ap-
never willing to make this contribution. There pointed in 1980 he began to try to reintroduce
has always been an inherent dilemma in sys- a strategic information systems planning group
tems planning and implementation in very apart from the operational systems personnel;
large and complex organizations such as SSA. this became the origin of the SMP. Multiple
Bottom-up planning and implementation gives reorganizations had failed to separate system
a better fit to the needs of users, and is more operation from system planning and devel-
likely to succeed than a top-down approach be- opment.
cause the users have a vested interest in it.
But bottom-up planning is also likely to result SSA then undertook two major initiatives
in a lack of integration and a failure to address to address its systems problems: the Paradyne
the long-term needs of the organization as a project and SMP. The Paradyne project was
whole, esecially if that implies a significant and initiated to replace the old GTE equipment
fundamental change in the way the organiza- that was then beyond its estimated system’s
tion conducts its day-to-day business. life and was failure prone and expensive to
maintain. It is usually said to predate SMP,
The Office of Advanced Systems was cre- since planning for it began in 1979, but because
ated in 1975 in an attempt to gain manage- the two initiatives are closely related, and be-
ment control over the planning and develop- cause the outcome of the Paradyne project has
ment of information systems, and buffer it had significant effects on the way SMP is be-
from the demands and assumptions of the oper- ing conducted, it will be described here.
ations side of SSA. But in the 1979 reorgani-
-
‘gJack S. Futterman, “Administrative Developments in the mu .s. Conwess, Gener~ Accounting Office, The smj~ *u
Social Security Program Since 1965, Social Security Bulletin, rity Adrm”nistration Needs To Continue Comprehensive Long-
April 1972, pp. 3.9. Range Planning, HRD-79-118, Sept. 10, 1979.

THE PARADYNE AFFAIR


The Paradyne project was one of the largest computer systems with an anticipated life of
single government civilian information sys- 8 years, plus related software.
tems upgrades ever undertaken. The original Initially this was to be a one-for-one re-
contract was for $115 million, the largest ever
placement of SSA’S deteriorating and obsolete
let for information technology by SSA. It be-
SSADARS data communications terminal
came a management disaster, even though in
equipment, located in District Offices. Before
some technical respects the effort worked.
SSA issued its terminal solicitation in June
1980, GAO and GSA had reviewed the plans.
The Paradyne Contract Both objected to the simple, original plan for
purchasing dumb terminals that were not pro-
On March 27, 1981, SSA awarded a commu- grammable and could not easily be adopted to
nications terminal replacement contract to the future changes in requirements, and that re-
Paradyne Corp. of Largo, Florida. Paradyne stricted the network architecture to the cur-
was to supply the agency and its field offices rent method of operation, precluding local of-
with approximately 1,850 programmable micro- fice data processing. SSA had simply thought
127

about the existing SSADAR system and how offices must have speedy access to data to is-
to make it more efficient, rather than reconcep- sue social security numbers, maintain earnings
tualizing the entire information processing sy~- records, accept claims, and process changes.
tern. This was to be a fatal weakness through- Before the Paradyne purchase, the network
out the Paradyne affair. was composed of a variet y of incompatible and
outdated equipment going back to the 1960s
When GAO recommended (with strong con-
and early 1970s. The primary components were
gressional support) that the terminals be ex- three types of data-entry terminals (including
panded to allow distributed processing, SSA the SSADARS, as described earlier), a collec-
agreed in concept that eventually the agency tion of modems,24 and local communication
would require programmable terminals in lo- lines to connect the terminals to concentrators
cal offices. But they argued that obtaining such (minicomputers). The modems and local com-
equipment would have to be deferred. The munication lines operated at low speeds of
memory capacity of the terminals would be en-
about 1,200 bits per second (bps). The concen-
hanced after they were installed. In January
trators combined, condensed, edited, and refor-
1980, GAO agreed to this approach.” This matted messages and sent them on to front-
project was now envisaged as a major part of end processors, which are communication com-
SMP’S proposed Data Communications Util- puters attached to the mainframe computer
ity Program.22 in Baltimore by high-speed trunk lines.
The equipment was simple in concept. Each
SSA also wanted the Paradyne network to
installation was to include a programmable
eliminate the key-to-disk terminal equipment
controller.’ s Access to SSA’S main computers
in the Program Service Center, which could not
would now be distributed, by a series of add- handle on-line inquiries or editing. Instead of
ons to the existing telecommunication net- operating three expensive, out-of-date, and in-
work. The Paradyne terminals would later be efficient telecommunication terminal subsys-
enhanced from dumb terminals to something tems, SSA would then have a single terminal
very much like a microcomputer, having local system.
storage and data-processing capability and the
ability to produce reports, draw graphs, make
lists, and store high peak load data for trans- Failures in the Paradyne
mission later. This would be an early and ma- Implementation
jor component in the multiyear SMP. SSA As already noted, SSA initiated the Para-
planned a phased installation of the equipment dyne procurement before SMP was imple-
between June 1981 and July 1983. mented, but later made it an integral part of
SSA depends heavily on its data communi- SMP. SSA planned to have installed the Para-
cation network to perform its mission. Field dyne terminals by September 1983, and to
have completed the hardware and software en-
2’U.S, Congress, General Accounting Office, Social Security
hancements for local processing, and also to
Administration Data Communication Contracts With Para- have begun designing user applications (such
d.vne Corporation Demonstrate the Need for Improved Man- as benefit payment computation or prepara-
agement Controls, IMTEC-84-15, vol. 33, July 9, 1984.
ZZA data ~ommunications utility iS a communications ‘et- tion of claims applications) to be automated
work in which all remote terminals and a central host computer locally using the enhanced equipment. SSA
are connected by a common “back bone’ capable of supporting hoped by September 1984 to begin using these
a large variety of data communication requirements and
equipment. applications so that operations could be com-
~~A control device through which terminals and other periph- pleted at the local level and public service
eral equipment such as printers, card readers, and off-line stor- would be improved. By March 1986, accord-
age devices are comected to a single communications line; hence,
a single programmable controller could control several printers
for outputting data, and also be connected to a card reader or
terminal for inputting data, and could handle these loads ‘(on- ‘~Devices that interface and translate between a digital com-
line, ” i.e., simultaneously. puter and an analog telephone line.
128

ing to the plan, SSA would have installed its and to submit ‘written analysis for actual
new data communications utility, providing tests” if certain hardware components
a high-speed communication network that were not available.z~
would integrate the Paradyne terminals and
By December of 1982, SSA had installed
other local office equipment into the central-
1,600 of the 1,800 Paradyne terminals. It had
ized national databases and computer systems.
given Paradyne a contract for software to en-
But from the very beginning, the Paradyne hance the transmission capabilities by changes
equipment had severe operational problems in the operating system. It had issued a com-
and breakdowns. SSA began acceptance test- petitive solicitation for applications software
ing of the first 16 systems on April 30, 1981. to begin automating field office operations.
All 16 failed to successfully complete 10 days
of continuous testing.
SSA awarded a sole source software contract
of more than $2.5 million to Paradyne on Sep-
SSA made a major contract modification and tember 8, 1982, to enhance the data transmis-
changed key operating standards so that the sion capabilities of its terminals by modifying
terminals would pass the test. Significant per- the terminal software. More than $1.8 million
formance problems continued. Acceptance of this was for documentation of all terminal
testing was suspended and the requirements software and developing a workplan for con-
were modified. During the first 16 months, structing the software modification and doc-
Paradyne made repeated changes to the ter- umentation.
minal controller in attempts to solve system
performance problems.” Paradyne did not But given the performance problems, SSA
begin to consistently meet contract perform-
began in April 1983 to rethink the role of Para-
ance requirements until April 1983.
dyne terminals. The SMP was more and more
focused on a strategy of centralized process-
GAO later found that: ing, which would eliminate the need for local
SSA did not use benchmarking techniques intelligence in the terminals. The sole source
in an effort to minimize costs to vendors software contract was canceled (April 29). By
in qualifying for contract considera- then SSA had paid Paradyne $550,000 under
tion,” but instead used “operational ca- that contract and Paradyne had delivered one
pability demonstrations” as the precon- product—a workplan for conducting the mod-
tract award testing mechanism. These ification and documentation. (Paradyne sub-
were supposed to demonstrate processing mitted a final bill for an additional $252,000
in July 1984.28)
and printing speeds and general opera-
tional capabilities; and By 1982 SSA had purchased the 841 leased
SSA did not, however, enforce the opera- Paradyne terminals already installed in SSA
tional capability demonstrations provi- offices, and had a lease on the other 1,000 ter-
sion—i.e., did not ask vendors to demon- minals. As of mid-June 1984. SSA was still con-
strate actual equipment or document the sidering whether to buy, or continue leasing,
testing, or provide programs or workload the remaining 1,000 terminals, although it was
file mixes, but instead allowed each ven- unclear how they could be used, since the Para-
dor to structure its own demonstration dyne equipment was no longer part of the fu-
ture district offices under the SMP.
‘“1’hey included four hardware changes, four versions of the
operating system soft ware, five vesions of the hardware, in The Paradyne terminals ultimately did work
21 different combinations, As late as August 1982, 17 differentt as planned although they had severe startup
versions of the cent roller were being used by SSA. GAO, op.
cit., IMTEC-84- 15, p. 16 ——.—
-“Contrary to GSA guidelines, which strongly recommend -;(; AO, op. cit., IMTF:C-Wi5, p. ]J1.
that agencies use bench mar-k tests, in which agency computer ~ “1’his was ~ittacked as an “$800,000 w [)rkplau’ note, how-
programs and workloads are run on vendor equipment to vali- ~’~rt’r, that it maj’ hate co~’er-c’d work betwwln Sept R, 1982 and
date system performance. Ipr. 29, ] 9k3, 9 m o n t h s rr}t re])r(’sentt’(1 b~r the ~r(wkplan.
problems and excessive down time. They will
be replaced, beginning this year, with desk-top
termininals l as described in chapter 2.

SSA Failures in Managing


the Paradyne Project

1. faulty system development practices,


2. faulty procurement practices, and
3. underlying structural weaknesses in the
procurement oversight procedures.
other Factors in the Paradvnc
Contract Problems

GAO said that expressing requirements in


terms of general equipment performance speci-
fications for individual terminal components,
as SSA did, may have biased the solicitation
toward particular vendors. Moreover t this
The 1982 restructuring gave primary respon-
method does not allow vendors to address over-
sibility for planning and managing ADP/data
all systems processing requirements but in-
stead forces them to address specific subre-
quirements.
Having failed to analyze its requirements
sufficiently, or to fully conceptualize how an
130

communications procurements to the Office of ment made by a competitor and altered to ap-
the Associate Commissioner for Systems In- pear as Paradyne’s; that it altered other
tegration, merging the functions of specifica- equipment so that it falsely appeared to meet
tions development into one office—the Office the processing rates required; that it falsely
of Systems Engineering. This also lowered the represented that its microcomputer would
level within SSA at which judgments are made meet SSA needs; and, in short, that the tests
about the adequacy of proposed developments. were rigged and that Paradyne sold SSA a pro
It was under these conditions that implemen- totype rather than the off-the-shelf terminal
tation of the Paradyne contract proceeded and SSA thought it was buying.
the sole source software contract for $2.5 mil- In February 1984, the former Director of
lion was given to Paradyne. SSA’S Office of Data Communications (which
The Department of Health and Human Serv- played a key role in the contract award) was
ices (DHHS) is responsible for monitoring SSA charged in criminal court with attempting to
computer acquisitions through its Assistant extort more than $400,000 from a California
Secretary for Management and Budget. DHHS software company in return for assurances that
did review the SSA procurement request and the firm would be selected as a subcontractor
conducted a postaward review of the terminal on a $4 million data communications software
replacement contract with Paradyne but did contract to be awarded to Paradyne.
not become involved in key phases of the In March 1984, Sigma Data filed a civil com-
procurement such as definition of require- plaint asking $70 million in compensatory and
ments, development of the solicitation, prea- punitive damages from Paradyne, claiming
ward testing, acceptance testing, or measur- that it (Sigma Data) would have received the
ing of performance. In effect, according to SSA contract had Paradyne’s misrepresenta-
GA0, 33 DHHS “in accordance with its nor- tions been identified earlier.
mal practices, re-delegated management and
oversight authority for these activities to SSA. In September 1985, SEC and Paradyne
. . . As a result, SSA received little, if any guid- agreed to an out-of-court settlement on charges
ance from HHS. . . .“ of commercial fraud. Criminal investigations
of SSA and Paradyne personnel, and several
An Unfinished Story civil suits, are continuing. The settlement re-
quired no admission of wrongdoing by Para-
In early 1983, SSA began developing a new dyne but simply the promise to comply with
technical approach described in detail in chap- Federal securities laws in the future.35 But on
ter 2, for providing field office claim represent- December 12, 1985, Paradyne, eight current
atives with terminals for direct interaction and former executives, and one former SSA
with the public, but not for distributed data official were criminally indicted for bribery,
processing. conspiracy, and lying to government investi-
gators concerning the 1981 contract with
SSA’S dealings with Paradyne became the
Paradyne. The former SSA director of telecom-
subject of litigation in both civil and criminal
munications allegedly accepted a $500,000 con-
courts. The Securities Exchange Commission
tract for software developed from Paradyne.
(SEC) filed a civil suit against Paradyne in
March 1983, charging the firm with violations
of the Securities Acts.34 SEC alleged that Aftereffects and Implications for SMP
Paradyne, in the preaward operational capa- The Paradyne case was a severe blow to
bility demonstration tests, used dummy equip- SSA’S reputation just at a time when outside
3 3 ~P. cit., IMTEC-84-15, PO 7.
G A 0
support was needed to assure funding of the
]~%cu~itie~ Act of 1933 a9 amended, 15 U.S.C. Par. 770(A)~
1976; Securities Exchange Act of 1934, as amended, 15 U.S.C. 35
’’ SEC Settles With Paradyne, ” Computerworlcf, Elept,. 16,
pars, 78J(B) and 78 M(A). 1985.
731

Systems Modernization Plan. Hindsight sug- The perception of these deficiencies account
gests that there were three basic flaws in SSA for much of the skepticism with which SSA’S
procedures: critics view SMP. They question whether SSA
now has any more rigorously examined objec-
1. SSA had not thought through how it
tives than it did in 1979 to 1981; whether its
wanted to do business, and had not sys-
systems personnel are more capable now than
tematically defined its information re-
they were then; and whether the reviews and
quirements;
checks on the system are now more likely to
2. SSA probably did not have the onsite per-
catch mistakes or detect fraud. Moreover, the
sonnel capable of making a thorough
criminal indictments of SSA personnel have
study of its requirements and translating
not been reassuring. SSA, in addition, has
that into a full modernization plan;
probably been made even more conservative
3. the merging of the specifications develop-
and cautious, more likely to stick to short-term
ment and review functions was a mistake,
solutions and nonrisky options—for instance,
compounded by failure to bring in exter-
its insistence on “proven technology’ for SMP,
nal consultants capable of criticizing the
which may make its decisions worse rather
procurement; this widened the possibility
than better.
that SSA personnel could be fooled and
defrauded bv . vendors.
Chapter 8

The Oversight of SMP, 1982-86


Contents
Page
Oversight Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
GAO Audits of the SMP Effort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Congressional Hearings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
1981-82 Hearings: How Accurately Was Congress Informed?.. . . . . . . . 136
Hearings Since 1982: How Well DidSSA Report to Congress? . . . . . . . 137
Executive Branch Reviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Inspector General Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
OMB Directives.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Chapter 8

The Oversight of SMP, 1982-86

OVERSIGHT INSTITUTIONS
In the executive branch, the Social Security a large number of vendors to submit competi-
Administration (SSA) is within the Depart- tive bids satisfying those requirements.
ment of Health and Human Services (DHHS),
and like all agencies is subject to directives The Competition in Contracting Act of 1984
from the Office of Management and Budget seeks to strengthen the Brooks Act; among
(OMB). Information systems procurement and other provisions, it permits unsuccessful bid-
management receives additional oversight. ders to go to a Board of Contract Appeals,
GSA has final authority to purchase auto- within GSA, which can suspend all procure-
mated data-processing (ADP) equipment but ment during the appeal. (SSA’S current tele-
can delegate purchasing authority to agencies. communications procurement is tied up by pro-
OMB is responsible for overall policy. The Na- tests from potential vendors who thought that
tional Bureau of Standards provides techni- SSA’S specifications were unduly restrictive.)
cal resource support. In practice, the effect of the procurement
The principal congressional oversight bod- process requirements has usually been to em-
ies concerned with SSA are four House Com- phasize least initial costs rather than broader
mittees and four Senate Committees. In the lifecycle concepts, which also include the costs
House these are the Committees on Appropri- of software, maintenance, and manpower. The
ations, Ways and Means, Government Opera- initial hardware cost usually drives the pro-
tions (sometimes called the Brooks Commit- curement decisions.
tee), and the Select Committee on Aging. The Even before passage of the Competition in
active Senate Committees in recent years have Contracting Act, the process of systems pro-
been the Committees on Finance, Appropria- curement was a lengthy one, as is almost any
tions, and Governmental Affairs, ] and the process involving formal procedures necessary
Special Committee on Aging. to assure accountability and fairness. Accord-
The 1965 act governing procurement of Fed- ing to many Federal Information Resource
eral ADP equipment (the Brooks Act) seeks Managers z this often results in a major sys-
to assure competitive and fair procurement, tem being far behind state of the art by the
and sets forth central management responsi- time it is installed. The Competition in Con-
bilities for ADP. The Brooks Act restricts the tracting Activities law has added a protest pro-
capability of an agency to carry out a sole cedure, which some Federal procurement of-
source procurement for large systems (that is, ficers say can be abused to the detriment of
to order a system from one vendor without orderly procurement procedures. SSA officials,
competitive bidding). The agency must estab- for example, privately say that:
lish functional and technical requirements for ● vendors have protested procurements
the system or equipment it needs, and invite solely to damage the financial standing of
the winner by delay3;
1 In 1985 the Senate Committee on Governmental Affairs and
the House Cornmi ttee on the Judicary, Subcommittee on Courts,
Civil Liberties, and the Administration of Justice requested an 2
Proceedings of an OTA Workshop on Federal Information
OTA assessment of Federal Government Information Technol- Resources Management, September 1984.
ogy, which contained a series of three reports released in 1985 sThe winner of the contract award, may for example ‘ave ‘m-
and 1986. The Senate Committee on Governmental Affairs sub- ediately ordered equipment and material or engaged work-
sequently requested this SSA case study, as an additional probe ers; even if the contract is likely to be upheld, that is, does not
of the kinds of generic problems that had been identified in the have to be recompleted, the contractor has suffered a cash drain
earlier and broader assessment. that could threaten its financial stability.

135
136

● have withheld information about techno- aware of basic, deep-seated problems with
logical capabilities when potential respond- effects that are persistent, cumulative, and
ers were given opportunity to comment, relentlessly destructive. The tightly focused,
prior to a formal request for proposals, only highly detailed nature of GAO reports, which
to protest subsequently that the specifica- allows them to answer congressional questions
tions in the request for bids do not allow
them to offer this improved capability;
with pRecision, may at times prevent them
● or have protested on the final day of the 45- from revealing larger patterns of management
day protest period in order to delay the weakness. GAO reports are also focused pri-
process long enough to complete the devel- marily on the question of whether existing leg-
opment of their proposed system. islation and policy guidelines have been fol-
lowed, rather than raising questions about
These tactics can delay a procurement for whether they are appropriate for achieving
8 months or more. Some States, to avoid simi- desired objectives.
lar problems, are requiring protesters to post
bonds. There are many critics of Federal pro- GAO is, however, currently carrying out a
curement procedures who maintain that they major management review of SSA, one of a
result in control of equipment purchases be- series of GAO reviews of management of Fed-
ing separated from consideration or knowledge eral agencies undertaken to support implemen-
of the activity to which it will be applied, and tation of the recommendations of the Grace
sometimes add years to a major procurement. Commission. These management reviews are
But the Brooks Committee has clearly been broader than traditional GAO audits and rep-
responsible for bringing rationality, profession- resent a new initiative, begun in 1982, to re-
alism, and accountability to Federal informa- view the overall management of Federal de-
tion systems procurement. partments or agencies in terms of effectiveness
in achieving their missions. Recognizing that
The effectiveness of all congressional over- good management is essential to achieving pol-
sight is, however, only as good as the informa- icy objectives, the GAO management reviews
tion that Congress gets about Federal agency are intended to demonstrate that:
actions, and there are serious structural prob-
lems in assuring that quality. Past insufficient attention to management
has led to chronic, unresolved problems in pro-
All of the congressional committees are as- gram delivery and administrative manage-
sisted in their oversight role by the General ment, including financial and information
Accounting Office (GAO), which continually resources management; (and) inadequate man-
studies and audits SSA, having a continuing agement structures or systems have often led
onsite presence at SSA for this purpose. Spe- to crisis management or darnage control rather
cial studies are conducted from time to time than real progress.4
by the other congressional support agencies, All congressional oversight is ultimately de-
the Congressional Budget Office, the Congres- pendent on information made available to it
sional Research Service, and the Office of Tech- by Federal officials. As noted repeatedly at
nology Assessment. But GAO’s detailed au- points in this report, agency officials are often
dits, with the benefit of immediate access to unwilling or unable to call attention to emerg-
SSA operations, are particularly essential, ing problems, or are required to shape their
since none of the other congressional support estimates of resource needs to fit the direction
agencies can mount the resources to study SSA of Administration policy and priorities.
at the same level of detail; nor do they have
the inside access that GAO has, so that they ‘From a description supplied to OTA by the General
are largely dependent on SSA spokesmen for Accounting (Mice [GAO), June 1986. GAO management reports
have been issued on the Departments of Housing and Urban
some kinds of information. Development, Labor, and Justice, and on the Defense Logis-
tics Agency; others are being completed on the Department of
However, even the GAO audits have some- Transportation and SSA, and are underway at a number of other
times not been sufficient to make Congress agencies.
137

GAO AUDITS OF THE SMP EFFORT

management.

GAO also identified as indicator-s of serious


problems:
● a 15 percent attrition rate in systems per-
sonnel in 1980 to 1981,”
‘ I J.S. (’ongmss, ( ~ener-a] :lc(ol]ntin~ offic’(~, .$~I]J jng Soc’]ul
Securit.}r Computer Problen].q: Comprehensi\t> Correc(i\’e.4 c
tion Planned and Better Alana~remenr ,Veeded. r[~port of the [ J.S.
(’ontroller-(;eneral, 111?1)-$+2-19. I)ec, 10, 19H1
(’l)mmisslon[~r NlcStecn told the I{ouw Appropriate ions
( ‘onlmitt[~e in ] 9H5 that t h(, norn~al attrition rat t is 4 to 6 per-
138

● the low level of skills of systems person- underestimating the magnitude of corrective
nel, and actions necessary in software improvement
● 45 recorded acts of sabotage or vandal- and data validation.
ism between February 1977 and Febru-
ary 1981. SSA estimated that 65 to 70 percent of the
12 million lines of code then in use could con-
From 1982 to the present, GAO reports have tinue to be utilized, but had done no studies
emphasized errors caused by ADP systems, to validate this estimate. The System Modern-
poor field office management, and poor con- ization Plan calls for data verification and file
tracting procedures in purchase of telecommu- cleanup to be done within 3 years; GAO doubted
nications and ADP equipment; and were in- whether this could be accomplished.
creasingly critical of SSA planning for systems
modernization. 8 GAO found that SSA was In spite of these problems, GAO concluded
...—— in early 1982 assessments that SMP is a defi-
{continued from previous page) nite turnaround step in the right direction, and
cent. However, the average annual turnover rate for full-time gave it a strong green light. More recent critical
permanent General Schedule employees in nondefense Federal reviews of the SMP by GAO indicate that SSA
agencies in 1984 was 11.8 percent (U.S. Congress, Congressional has been able to solve many of its hardware
Budget Office, Employee Turnover in the Federal Government,
special study, February 1986, table 1). problems, but that in the areas of software and
%ee GAO reports: Social Security Fieki Office Management databases serious deficiencies remain.g
Can Be improved and M*”ons Can Be Saved AnnU&y Through
Increased Productivity, HRD-82-47, Mar. 19, 1982; Complete
and Accurate Information Needed in Social Security’s Auto- Adm”m”stration Progress in Modernizing Its Computer Oper-
mated Name and Number Fdes, HRD-82-18, Apr. 28, 1982; Ex- ations, IMTEC-85-15, Aug. 30, 1985; SSA Computer Systems
amination of the Scw.ial -“ty Admi.m”stration Systems Mod- Modernization Effort May Not Achieve Planned Objectives,
erm”zation Plan, HRD-82-83, May 28, 1982; Social Security IMTEC 85-16, Sept. 30, 1985; Income Security: Selected Dis-
Administration’s Data Communication Contracts With Para- abih”ty Payments, HRD-86-47FS, December 1985; Issues Re-
dyne Corporation Demonstrate the Need for Improved Man- lating to Agency Field Offi”ces, HRD-86-71BR, March 1986; Cur-
a~ment Controls, IMTEC-84-15, July 9, 1984; Ad&”tionai In- rent Stat us of the Federal-State Arrangement ts for
formation on the Social Secudy Ati”stration Management Adrm”ru”stering Social Security’s Diability Programs, HRD 85-
of Data Communication Contracts With Para&”ne Corporation, 71, Sept. 30, 1986.
IMTEC-84-83, Aug. 27, 1984; Review of Two Proposed Auto- ‘U. S. Congress, General Accounting Office, SociaJ Security
matic Data Processing Procurements by the Socitd Security Adrni”ru”stration Progress in Modernizing Its Computer Oper-
Adm”m”stration, IMTEC-85-7, Apr. 10, 1985; Social Security ations, IMTEC-85-15, Aug. 30, 1985.

CONGRESSIONAL HEARINGS
1981-82 Hearings: How Accurately been “alerted to the magnitude of the systems
Was Congress Informed? problem by the earlier testimony of three
former SSA commissioners, ” summed up in
In May 1981, the House Ways and Means’ a report prepared for subcommittee use by the
Subcommi ttee on Social Security and Subcom- staff.11
mittee on Oversight jointly held hearingsl” to --
“begin identifying some of the problems that In September 1981 the House Committee on
are facing the SSA in the management of its Government Operations also heard from Com-
ADP Systems. ” They heard newly appointed missioner Svahn, and others.12 In these and
Commissioner John A. Svahn talk about what
1 lu s Con=e SS, The soci~ Security Ad~”nistrations Data
Chairman Rangel called SSA’S “state of cri-
Processing System Crisis, a report prepared by the staff of the
sis. ” The Subcommittee on Social Security had Subcommittee on Social Security of the House Ways and Means
Committee, 97th Cong., 1st sess., May 22, 198i.
1°U.S. Congress, Automated Data Processing Systems, 1 2 s. C
U
. , vja~~.ty of tjhe Social Security Ad~”m”stra-
onWegg

Hearing Before the Subcommittee on Social Security and the tion Computer Systems, Hearing Before a Subcommittee of
Subcommittee on Oversight, House Committee on Ways and the House Committee on Government Operations, Sept. 23,
Means, 97th Cong., 1st sess., May 22, 1981. 1981, 97th Cong., 1st sess.; this was the Brooks Committee.
139

other hearings Svahn presented a dark picture progress, or of risk, are more diverse, less ob-
of an agency in real danger of collapse. Long vious, and less accessible when they are hid-
lists of delays, backlogs, and critical problem den in mammoth databases. The temptation
areas were presented, and appeared again in to selectively pick and present such measures
SSA’S 1982 SMP. is stronger as the resources needed for (or al-
ready sunk in) systems become greater. The
There was no way for the congressional com-
flow of work and the definition of discrete
mittees to challenge these statements and
tasks or operations changes as the technology
figures presented in support of SSA’S plan to
changes, so that it is difficult to compare per-
salvage its operations with a 5-year systems
formance at different periods. Thus, even an
development effort. Indeed there was little rea-
onsite auditing capability, such as GAO has,
son for them to do so, since both critics and
may be frustrated by the difficulty of defin-
supporters of SSA agreed that the situation
ing and tracking real progress.
was bleak.
Yet there was tension and resentment within Hearings Since 1982: How Well Did
SSA, between Commissioner Svahn and his SSA Report to Congress?
aides and consultants, who put together the
testimony and the 1982 SMP, and the long- There have been six major sets of congres-
time SSA managers who had been struggling sional hearings relevant to SSA information
to cope with the problems and to keep checks systems since 1982.13 The 1983 and 1985
coming out on time. The latter resented hav- House Appropriation Hearings, and Hearings
ing their performance pictured so unfavorably. Before the Senate Special Committee on Aging
Five years later, with Mr. Svahn gone, many in 1983, were especially important.
of these managers heatedly dispute the figures
In the 1983 Appropriations Hearings, the
used in 1981 to 1982 to measure error rates,
focus was on future solvency, the impact of
lost time, backlogs, and vulnerability to secu-
budget cuts on SSA activities, and SSA Com-
rity violations and disruption of procedures.
missioner Svahn’s presentation of the SMP.
If these performance or quality measures are The Committee members were, in general, in
in dispute, however, then SSA’S own measures favor of SMP and ready to provide funds to
of improvement and progress since 1981 also carry it out.
can be disputed. It is reasonably clear that
The hearings before the Senate Special Com-
some of the ways of measuring or counting er-
mittee on the Aging in 198314 built on a so-
rors and time expended have changed. Possi-
phisticated, critical staff background report
bly these changes are necessary because of the
combining analysis of external events affect-
changed systems, but SSA is not careful to
ing SSA and internal management actions.
point this out to its oversight committees.
These contradictions are not important now
except to illustrate the general possibility that
congressional oversight can be misled by in-
formation presented by organizations in sup- lsHemings Before the House Committee On Appropriations,
port of or in defense of their actions or of ex- Subcommittee on Labor, HHS, Education, and Related Agen-
cies, 97th Cong., Mar. 9, 1982; Hearings Before the Senate Com-
ecutive branch policies and directives. This mittee on Appropriations, Subcommittee on Labor, HHS, Edu-
problem has always existed. It is made worse cation, and Related Agencies, 97th Cong., Mar. 10, 1982;
by advanced information technologies that Hearings Before the Special Committee on Aging, U.S. Senate,
98th Cong., Nov. 29, 1983; Hearings Before the House Com-
make performance data more difficult for the mittee on Appropriations, 99th Cong., 1985; Hearings Before
layman to grasp or to question. Evaluation of the House Committee on Ways and Means, Subcommittee on
agency decisions related to design, procure- Social Security, 99th Cong., Apr. 3 and 11, 1985.
“U.S. Congress, Socitd Security: How Well Is It Serving the
ment, and management of systems requires Public? Hearing Before the Special Committee on Aging, Sen-
more highly technical knowledge. Measures of ate, 98th Cong., 1st sess., Nov. 29, 1983.
140

Several areas were cited in which management SSA then requested an additional $125 mil-
problems had exacerbated systems problems: lion as a reserve fund in 1986 because of the
staff cutbacks of 5,000 positions between unanticipated costs related to automation and
1977 and 1984, and internal promotion the implementation of the Disability Benefits
Reform Act of 1984. However, Commissioner
and retraining practices that lowered the
McSteen pointed out that these costs did not
overall quality of personnel;
reflect badly on SMP progress; SSA was re-
measures of work performance that re-
questing 2,308 fewer work-years for 1986, as
warded initial claims processing and data
a result of “automation improvements and pr~
collection but not quality of service to ben-
cedural changes. ”
eficiaries;
complexity of instructions and forms that Many Congressmen appeared less interested
the staff had to use in these measures of progress than in the star-
newly aggressive enforcement of debt col- tling discussion in the Washington Post (Feb.
lection and disability redetermination, 19, 1985) of alleged plans to close 200 SSA dis-
which added to the workload just when trict offices and reduce the work force by nearly
the staff was being reduced; and a quarter (17,000 positions). It was feared that
three internal reorganizations since 1975, this would, for constituents, decrease both ac-
with no visible benefit. cess to SSA service representatives and the
quality of the services provided. The Commis-
The message of the Committee to SSA was
sioner responded that this reduction would be
that Congress would measure SMP’S success
made possible largely by systems moderniza-
not in terms of its technical sophistication but
in terms of its improvement of services to ben- tion, i.e., automation. She argued that it could
be done without degrading service delivery and
eficiaries. The committee staff report also ques-
largely without firing workers, since the nor-
tioned the “marginal strategy” of seeking to
mal attrition rate of 5 to 6 percent would ac-
preserve most of the existing software instead
count for about 4,000 workers each year, and
of developing new software.
5,000 part-time workers would be dismissed.
The 1985 House Appropriations Hearings, Other displaced workers were to be retrained
after Commissioner Svahn’s departure, gave and relocated. She added however that an “im-
Acting Commissioner Martha McSteen the op- balance of staff” would be SSA’S greatest prob-
portunity to announce her management goals lem, i.e., matching people to the right job.
and also to announce the first improvements Normal attrition is of course unlikely to oc-
in service delivery as a result of SMP. She cur selectively in just the jobs that are being
pointed to a number of improvements in proc- eliminated by automation, but Commissioner
essing time and reductions of backlogs be- McSteen did not offer any estimates of the
tween 1982 and 1985. amount of relocation and/or retraining that
would be necessary if SSA relied on attrition.
The SMP had been projected, in 1982, to cost
$449 million over 5 years. In 1985, only $101 Questioning of Commissioner McSteen re-
million had been spent, although the 1982 SMP vealed that OMB had originally demanded a
had projected that $293 million (61 percent of reduction of 19,000; SSA had negotiated this
the total) would be expended by that time. The down to a goal of 17,000 SSA workers by 1990.
total projected cost, however, had risen to $863 Some Congressmen were incredulous; some
million, so that less than 18 percent had been protested the absence of any SSA studies of
expended. the potential effects on clients of the proposed
closings and reductions. Congressman Natcher
asked:
‘“An SSA office in San Francisco was said to have received
an average of 28 pages of instructions per day in the fall of 1980, Don’t you know as well as 1 do, that this is
which tests showed would require 17 years of education to under- not going to work?. . . last year. , . we added
stand, as compared to 11 -years for the W’al) Street Journal. the $60 million to maintain a staffing level at
the 1984 level We were very specific in the In the 1985 Appropriation Hearings, the
report. Tell us again, if you will, Mrs. McSteen, committee members generally had praise for
why your current plans me to support 2,180 the social security program, and for Commis-
fewer employees than the Congress directed sioner McSteen as a manager. The~’ hoped that
for the current fiscal year. 16 the SMP would improve SSA operations. On
Congress had authorized 80,253 full-time the other hand the Brooks Committee was now
equivalent (FTE) positions for the agency in highly critical of SMP because it emphasizes
both fiscal years 1984 and 1985. At the end hardware problems and appears less satis-
of fiscal year 1984, SSA actually had only factory in addressing software problems. All
79,951 FTEs, and at the end of fiscal year 1985, of the oversight committees have raised seri-
it had 78,038, about 3 percent under the au- ous questions as to whether the efficiency and
thorization. The President’s fiscal year 1987 rationalization promised by SMP will also
request cal!.s for 73,270 FTEs, or a reduction bring about a reduction in service, especially
of 6,681 (8 percent) over 3 years. A con- in rural areas, or a reduction in face-to-face in-
tinued reduction of 3 percent per fiscal year teractions between SSA employees and clien-
would mean about 13,000 fewer jobs in fiscal tele. Many are highly critical of OMB policies.
year 1990 than in fiscal year 1984. The goal In particular, OMB policy makes it difficult
of 17,000 fewer jobs could be reached in fiscal to spend money on training and retraining,
year 1992 at the present rate of shrinkage. which is much needed at SSA.

Despite the Washing-ton Post story that dis- From 1983 to 1986, the acting commissioner
turbed Congressmen, it is not clear that OMB was a long-time career employee and former
directly ordered SSA to close 200 field offices. regional commissioner who had a high level
The original proposal was reportedly to close of approval within SSA and in Congress. In
arly offices with fewer than 25 employees (a March 1986 a new commissioner was named,
large proportion of the field offices), and SSA who was until then a Deputy Secretary of HHS
gave Regional Administrators the power to but is a newcomer to SSA. Based on OTA inter-
veiws, there were indications of foreboding and
close offices within those criteria, without fur-
ther authorization. GAO reported in March dismay in SSA, its union, and its oversight
1986 that 228 reviews of field offices had been groups at the prospect of further policy shifts
conducted by .SSA in the past year, but no or internal reorganizations.19
offices had been closed as a result, and it was For about two decades, and especially since
‘‘unlikely that many offices will be closed when the SS1 crisis of 1973, many people in Congress
the reviews of all offices are completed by De- have been disturbed by the apparent misesti-
cember 1987. GAO noted that the effect of mates of the adequacy of SSA resources to
the Emergency Deficit Reduction and Bal- carry out congressional mandates for changes
anced Budget Act of 1985 could change that in social security benefits, procedures, or pro-
forecast. It is possible that SSA could be un- grams. There is continuing and recently re-
able to keep some offices staffed. Because of newed uncertainty as to whether these mis-
the Gramm-Rudman-Hollings Act, SSA is now estimates result from failure by SSA officials
under tight restrictions on both hiring and in- to estimate realistically, or the failure to com-
ternal personnel transfers, ’x so that offices municate these needs to Congress in a way that
that lose staff through attrition may not be is clear and credible, or from conflicting pres-
restaffed, sures and directives imposed on SSA by its
“ 1985 Appropriation\ Hearings, p. 915 (9ee no(d 13J. ——.——
‘- Numtwrs supplied bj congressional committee staff. “’I+’or example, Rep. Rol’bal. Chairman of the House Select
1“Staff memo of Feb. 13. 1986, from Dr. Otis Howen, Secre- (’ommitt,ee on .4ging, issued a press release warning that “n~an-
tary of HHS, to Heacis of operating Divisions. etc.; and staff agement of such an import ant agency should be more stable,
memo of Feb. 16, 19/36, fr{)m .S,SA Acting L)eputv Commissioner and that ‘‘from a public policy’ standpoint it would be pr~fera-
for Management and Assessment to other SS,4 I)eputy, Asso- ble that, the agenc~”s chief administrator have at least a fev
ciate. and Regional (’commissioners, ?Tt,ars of hands-on experience.
142

multiple congressional oversight committees, policies such as budget reduction. In regard


or from constraints placed on SSA in regard to SMP each of these factors appears to have
to its communications with Congress by DHHS operated at different times; the latter may be
and OMB, in the interest of Administration of increasing importance at present.

EXECUTIVE BRANCH REVIEWS


SSA’S representations to the Administration Audit, Felix J. Majka, conducted a review of
as well as its communications with Congress the Claims Modernization Project from late
must go through its parent agency, the Depart- 1983 through May 1984, and found numerous
ment of Health and Human Services, which deficiencies. The DHHS Inspector General,
has many components and programs to defend. Richard Kusserow, issued a report on Janu-
The budget examiners within OMB—who act ary 30, 1985, calling attention to problems with
as the President controller, closely involved the Claims Modernization Project of SMP:
in developing the budget, controlling the ● a formalized planning process was not
money flow, and monitoring expenditures—
completed, the scope of the project was
thus play a powerful role in relation to SSA.
not clearly defined, and interfaces with
DHHS itself of course must review and ap-
other systems had not been defined (as of
prove many SSA actions, such as major pro-
May 1984).
curement plans and personnel actions; al- ● adequate minimal standards were not in
though the force with which this supervision
place to guide the systems development
is exercised varies over time. Anew oversight
process; the most critically needed stand-
mechanism, the Inspectors General, also pro-
ards were data definition, documentation,
vides monitoring and oversight for Federal
and planning; and
agencies, including SSA. ● although much has been done to identify
potential control weaknesses in claims
Inspector General Reports processing, SSA did not yet have a for-
Congress created, in 1978, a new position or mal methodology for identifying new sys-
institution, “Inspectors General, ” to aid in the tem vulnerabilities and implementing
oversight process. zo Inspectors General, in controls. 21
every major agency, are especially concerned In the same report-memorandum, however,
with seeing that funds appropriated by Con- the Inspector General noted that:
gress are properly used; they report both to
Congress and to the agency. Our recommendations were generally con-
curred with by SSA and have either been im-
In the Department of Health and Human plemented or are in the process of being im-
Services, the Assistant Inspector General for plemented.
20
The first Inspector General (IG) post was created by stat- Another report horn the Inspector General
ute in HEW in 1976. IG posts were created in 1978 first for the following month criticized SSA’S admin-
the Department of Energy and then for 12 other major depart-
ments and agencies. The Department of Defense was added to
istration of a contract with a software vendor
the list in 1982. Nonstatutory IGs had existed earlier in the for obtaining “modern automated software
Department of Housing and Urban Development, in NASA, tools, ” and said that the software tools in-
and in DOD. IGs can initiate audits and investigations of sus-
pected fraud, abuse, or management deficiencies. Reporting both
stalled (for $24 1,916) did not filly meet the re-
to Department heads and to Congress, they can also bypass
department or agency counsels and take problems directly to 21U.S. Department of Health and Human Services, Office of
the Criminal Division of the Department of Justice. John D. Inspector General, Memorandum to Martha A. McSteen, Act-
Young, “Reflections on the Root Causes of Fraud, Abuse, and ing Commissioner for Social Security, “Audit Report-SSA’s
Waste in Federal Social Programs, Public Adm.im”stration Re- Redesign of the Claims Processing System Under the Systems
VieW, July/August 1983, p,366. Modernization Plan (SMP), ” ACN 15-52654, Jan. 30, 1985.
143

quirements defined by SSA, did not improve The Paperwork Reduction Act of 1980 (Pub-
operational programs, and were no longer be- lic Law 96-511) promulgated the concept of
ing used.22 The effort of the vendor to convert information resources management, or inte-
and improve 150,000 lines of COBOL code (at grated management of all basic information-
the cost of $150,000) was also unsatisfactory. handling activities and functions within an
agency. It charged OMB, assisted by GSA,
Again in June 1985, the Inspector General
with periodically reviewing information re-
criticized SSA for wasting over $1 million in
sources management by each agency (in prac-
the procurement of useless software.23 Kus-
tice, OMB delegates this task to GSA). OMB
serow criticized contractors for delivering proci-
is to provide guidance on all matters of bud-
ucts late and untested, the GSA for faulty over-
get allocation and procurement for informa-
sight, and SSA for hasty preparation and poor
tion technology, through its Office of Infor-
quality of the specifications and for poor
mation and Regulatory Affairs (OIRA). This
project management. He pointed to the Claims
office has not, however, played a strong role
Automated Processing System upgrade, say-
in review or guidance. It was not reauthorized
ing that software purchased from a vendor was
in 1983, but has continued to exist within
unusable. A similar result occurred with an up-
OMB. Representative Jack Brooks, now Chair-
grade of the Manual Adjustment Credit and
man of the House Government Operations
Award Process (MADCAP), and the conver-
Committee, in March 1986 asked the House
sion of earnings program software. Assistant
Appropriations Committee to refuse funding
Inspector General Majka told OTA in mid-
for OIRA because it has “concentrated its ef-
1986 that because software development “is
forts on the President regulatory reform pro-
the most difficult systems area with the most
gram rather than the functions assigned to it
failures, and because it receives relatively less
under the (Paperwork Reduction) Act. ‘ZS
focused attention from congressional oversight
OMB’S Office of Federal Procurement Policy
authorities than does hardware development,
also has played only a minor role.
his office “will continue to concentrate our
SMP reviews on software. ”24 Major OMB budgetary initiatives with re-
gard to SSA, some of which have been noted
OMB Directives throughout this report, are summarized here:
OMB’S role with regard to SSA has been ex-
● efforts to reduce disability roles by severe
ercised chiefly through its budgetary func- enforcement of the Disability Amend-
tions, i.e., efforts to constrain and reduce the ments of 1980;
agency’s work force, rather than through di-
● insistence on reducing the debt carried by
rect monitoring of systems modernization or SSA due to overpayments or erroneous
information technology management. payments; and
——--——
● staff reduction demands, originally a re-
~2U. S. Department of Health and Human Services, Office of duction of 19,000 in 3 years,- negotiated
Inspector General, Memorandum to Martha A. McSteen, Act- downward to 17,000 in 6 years, and pres-
ing Commissioner of Social Security, “Audit Report—SSA’s
Use of a Contractor To Improve Software, ” ACN 15-52649, Feb.
sure for closing some district and branch
6, 1985. offices.
‘W.S. Department of Health and Human Services, Office of
Inspector General, Memorandum to Martha A. McSteen, Act-
ing Commissioner of Social Security, “Audit Report—SSA
Needs To Redirect Its Software Improvement Efforts, ” ACN
15-51662, June 13, 1985. ~5According to news reports; see “Brooks Slams OIRA for
“Letter from Assistant Inspector General Majka to OTA Not Doing Its Job, ” Government Computer News, Mar. 28,
project director, June 24, 1986. 1986, p. 5.
*
.—

Glossary of Technical, Institutional, and


Legislative Terms Used in This Reportl
A-76: An OMB Circular that directs Federal agen- data that they have collected against data col-
cies to privatize, or contract out, government lected by other Federal or State agencies, to iden-
operations under some circumstances (see ch. 3). tify overlaps. This allows SSA to determine, for
ADP: Automated data processing; see information example, whether beneficiaries are receiving pay-
technology. ments under more than one assistance program.
AFDC: Aid to Families with Dependent Children; CMP: Claims Modernization Project, part of the
a program first established by the Social Secu- Systems Modernization Plan, designed to auto-
rity Act of 1935, that provides matching grants mate the filing of social security benefits claims
to States for financial assistance to dependent in SSA’S field offices (see ch. 2).
children in families in need because of the inca- Data dictionary: A comprehensive set of definitions
pacity, death, continued absence, or unemploy- of the data elements that are in a database, con-
ment of a parent. Administered by SSA. trolling the form they are given and the terms
AFGE: American Federation of Government Em- used to call them out of the computer’s memory
ployees (AFL-CIO), the union which represents (see ch. 2).
many SSA employees (see chs. 3 and 5). Database architecture: The plan or framework defin-
Artificial intelligence, or AI: A field of research con- ing the structure of an information-handling sys-
cerned with giving computers some human men- tem, the software and hardware used, and the
tal capabilities, such as “understanding” speech relationships between them (see ch, 2).
and visual images, choosing among options, etc. Database integration: The systematic combination
Expert systems are an early commercial appli- of all sets of data or information used by an orga-
cation of a rudimentary version of such capabil- nization, so that they can be accessed, through
ities already in use (see ch. 4). the use of common terms, by many users and
Black Lung Program: A Federal program adminis- systems (see ch. 2).
tered by SSA that provides monthly cash bene- DDS: Disability Determination Services; States
fits to miners (and their dependents or survivors) carry out the determination of disability of ap-
disabled by pneumoconiosis caused by occupa- plicants for Disability Insurance. OMB has pro-
tional exposure. posed that these services be privatized (see ch. 3).
Brooks Act: Public Law 89-306, passed in 1965, Debt Collection Act, 1982: Financial management
which regulates Federal information technology legislation that led to withholding all social secu-
procurements to assure that they are competi- rity payments from beneficiaries who had re-
tive (see chs. 5 and 8), ceived overpayments (see ch. 3).
Brooks Committee: An informal name often used DHHS: The U.S. Department of Health and Hu-
for the House Committee on Government Oper- man Services, of which SSA is a part; formerly
ations, which oversees implementation of the the U.S. Department of Health, Education, and
Brooks Act (see above); chaired by Representa- Welfare (HEW).
tive Jack Brooks of Texas (see chs. 6 and 8). EDP: Electronic data processing; see ADP and in-
COBOL: Common Business-Oriented Language; a formation technology,
computer language used by SSA (see ch. 2). Fourth-generation languages: Advanced computer
COLA: (Automatic) cost-of-living adjustments in languages that use “everyday” (English) vocabu-
social security benefits, to compensate for infla- lary and syntax, and are useful particularly for
tion; first legislated in 1972. administrative and management information
Competition in Contracting Act, 1984: Strengthens systems not used by computer specialists (see
the “Brooks Act” (see above) governing Federal ch. 4).
procurements of information technology; pro- GAO: General Accounting Office; a congressional
vides an appeals process for losing bidders to as- agency that monitors and audits government
sure that competition has been fair. programs and operations and makes recommend-
Computer-matching: A process by which Federal ations for improving their effectiveness and
agencies (including SSA) electronically check efficiency (see ch. 8).
Grace Commission: The President’s Private Sector
‘References to chapters identify the primary or most full discussion Survey on Cost Control in the Executive Branch
of thesubject matter, not necessarily the first use of the term or phrase. of the Federal Government, established in 1982
147
148

(Executive Order 12369), consisting of 161 high- amendments, to certain severely disabled per-
level industry executives and chaired by J. Peter sons under 65). Administered by SSA.
Grace; some Administration policies such as OASDI: Old-Age, Survivors, and Disability Insur-
those aimed at Federal work force reduction, are ance, popularly referred to as social security; pro-
derived in part from recommendations of this vides monthly cash benefits to replace income
commission or task force (see ch. 3). lost by retirement, disablement, or death of a
GSA: General Services Administration; the execu- worker. Covered employees (nearly 95 percent
tive branch agency that monitors and manages of American workers) pay social security taxes
government procurements, including procure- on their earnings; these are supplemented by em-
ment of computer systems (see chs. 5 and 8). ployer taxes to finance benefits. Established in
Independent agency: An agency that is not part of 1935.
the Executive Office or a Cabinet-level depart- OCR: Optical character recognition, or optical scan-
ment; such agencies, usually regulatory in na- ning; a technology that allows paper-based data
ture, report to both Congress and the President to be read and stored by a computer without be-
and their heads do not serve at the will of the ing rekey boarded (see ch. 4),
President but have fixed terms of office; the Fed- OMB: Office of Management and Budget; part of
eral Trade Commission and the Federal Commu- the Executive Office of the President (see chs.
nications Commission are examples. It has been 3 and 8).
proposed that SSA become an independent Optical disks: A new technology for storing data,
agency (House of Representatives Bill 5050) (see using lasers to write on disks, and offering orders
ch. 3). of magnitude more density, or storage capacity,
Information technology: Computers, telecommuni- than magnetic disks used in most computer sys-
cations, and electronic databases; other techno- tems today (see ch. 4).
logical devices or systems used for automated Oversight: The exercise of congressional power to
data handling. monitor and investigate the performance of ex-
Inspector(s) General: A post created by Congress ecutive branch agencies in carrying out laws and
in 1978 for all major Federal departments; In- expending public monies. Oversight is performed
spectors General carry out audits, investigate by designated congressional committees, through
fraud, and generally aid the oversight process; hearings and through studies and audits by con-
they report both to the department head and to gressional support agencies (see ch. 8).
Congress, and can carry a charge of wrongdoing Paradyne: The 1979 to 1981 Paradyne procurement
directly to the Department of Justice. of terminals (supplied by the Paradyne Corp.)
IRS: U.S. Internal Revenue Service, which issues was to replace the aging SSADARS equipment
the benefits checks authorized by SSA. (see below) that provided data communication
MADAM: Master Data Access Method, a software between SSA field offices and headquarters. The
program developed, used, and maintained by procurement was highly controversial, and fi-
SSA for extracting data from its many data- nally resulted in indictment of some SSA offi-
bases (see ch. 2). cials (see ch. ‘7).
Management information systems: Software/hard- PCIE: The President’s Commission on Integrity
ware systems and databases used for administr- and Efficiency, established in 1980, to advise
ative and management purposes rather than pri- President Reagan on improving government fi-
mary daily service operations or research (see nancial management (see ch. 3),
chs. 2 and 4). Privatization: The policy of contracting-out tradi-
Medicaid: A program established by the Social tional government services and operations to be
Security Act, Title XIX, that provides match- performed by private sector organizations (see
ing funds to participating States (now all but Ar- ch. 3).
izona) which provide for the cost of medical care Relational databases: Ways of organizing large
and services to low-income persons through di- amounts of data that allow great flexibility in the
rect payments to care providers. See Medicare. ways of asking for information; they use fourth-
Medicare: Established by the Social Security generation computer languages (see above), and
Amendments of 1965; the contributory Medicare make it easy for those not highly trained in in-
program includes compulsory hospitalization in- formation sciences to use a database (see ch. 4).
surance and voluntary supplementary medical SMP: The Systems Modernization Plan, first an-
insurance to persons 65 or over (and since 1972 nounced in 1982, for thoroughly improving or
149

replacing SSA’S information technology systems SS1: Supplemental Security Income; one of SSA’S
(see chs. 2 and 7). major programs; it provides monthly cash ben-
Software engineering: A set of techniques, tools, and efits to aged, blind, or disabled persons whose
standards for use in software development and other income is less than a specified amount
testing; a software engineering program is a ma- ($4,032 in 1986). SSI was established in 1972 to
jor component of the SMP (see ch. 2). replace categorical State assistance programs
Social Security Act, 1935: established the Social (see ch. 6).
Security Administration, then an independent Supercomputers: A term often used for the most
agency (see ch. 5). powerful computers available at any one time,
Social Security Disability Amendments Act, 1980: generally used first for scientific research. While
Public Law 96-265, a law for purposes of Fed- today’s computers use sequential processing, the
eral debt collection and financial management, next generation of supercomputers will probably
requiring a review of the status of nonperma- use parallel processing (see ch. 4).
nently disabled recipients of benefits, which re- Title II benefits: Retirement and disability monthly
sulted in the dropping of many recipients from cash benefits.
the SS1 rolls (see ch. 3). Unemployment insurance: State programs, under
SSA: Social Security Administration; now a com- Federal standards, to provide benefits to those
ponent of the U.S. Department of Health and involuntarily unemployed but able and willing
Human Services (see ch. 5). to work. The Social Security Act provides tax
SSADARS: SSA Data Acquisition and Response offsets and grants to induce States to maintain
System, a telecommunication system instituted these programs. Administered by the States.
in 1972 to provide interactive communication be-
tween headquarters computer operations and
data technicians in field offices (see chs. 2 and 6).