CHAIRMAN'S REPORT DECEMBER 1973

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82B00871R000100300003-6
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 35
• Document Creation Date: December 19, 2016
• Document Release Date: June 23, 2006
• Sequence Number: 3
• Publication Date: January 24, 1974
• Content Type: REPORT

Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Confidential
MHM~
UMEMM
Information Working Group
Chairman's Report
December 1973
MORI/CDF Pages
Confidential
Attachment
IRAC-D-75.1/1
24 January 1974
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Warning Notice
Sensitive Intelligence Sources and Methods Involved
Additional Warning
NATIONAL SECURITY INFORMATION
Unauthorized Disclosure Subject to Criminal Sanctions
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONTENTS
SUMMARY
1. Introduction
2. Discussion
A. CIRIS
B. Related Systems, Characteristics
and Use
C. Needs
D. Evaluation
3. Conclusions and Recommendations
A. CY 74 CIRIS Data Call
B. Interim Evaluation of Response to
Objectives
C. Longer Term Considerations
ATTACHMENTS
Attachment
IRAC-D-75. 1/1
24 January 1974
A. Terms of Reference
B. IWG Membership
C. DCI - Intelligence Community Staff Information
Needs & Options for the Next Data Call
D. An Analytical Approach to Evaluation
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2E9K* I - P82B00871 R0001 00300003-6
SUMMARY
The IRAC Information Working Group (IWG) was tasked to
review CIRIS and related intelligence management information systems
with the objective of developing recommendations for the implementa-
tion next year, and beyond, of CIRIS and/or some other follow-on
system which will:
1. Provide a community-wide basis for characterizing the
use of resources.
2. Provide a data base and means of analysis and evalua-
tion to support senior managers in meeting their current
and future objectives.
3. Provide a management information data base and inven-
tory of assets to support USIB, IRAC, OSD, and NSCIC
activities, and
4. Be responsive, cost effective, and minimize duplicative
effort.
The fundamental requirement is to achieve an information system
approach to help fulfill the President's 5 November 1971 directive to
the DCI.
The IWG reviewed previous evaluations of CIRIS; reviewed the
current CIRIS data base and related information systems; reviewed
user needs and user experience; reviewed related evaluation and
analysis activities; and debated issues and problem areas as they be-
came identified.
The resulting conclusions and recommendations include:
1. Specification of the issues surrounding the CY 74 CIRIS Data
Call.
2. Recommendation of an interim approach in CY 74 to the
DCI need for monitoring community response to Key Intelli-
gence Questions (KIQ's) and Objectives.
3. Recommendation for additional work toward the development
of a future community information system for critical needs,
making maximum use of other systems distributed throughout
the community.
Approved For Release 2QA~jaWa;,JA,-D,P82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 20ffi
IMB2 q4--ELP82B00871 R000100300003-6
1. Introduction
The CIRIS community information base and its predecessor,
a target oriented display,maintained by the IC staff, were developed
jointly by the Department of Defense, CIA, and OMB in response
to an OMB initiative five years ago. It was originally intended to
provide the DCI and the community with resource data. To date,
CIRIS has been useful for selective resource displays, some corre-
lation work, issue studies in support of program reviews, and the
preparation of Congressional briefings and last year's National
Intelligence Program Memorandum. It has also been queried in
the course of a variety of studies, and it supports the need of the
Critical Collection Problems Committee of USIB for an inventory
of intelligence units and activities. Users during the past year were
primarily the IC Staff and DIA.
There are a number of related Management Information Systems
(MIS) in being, or proposed, to serve the needs of individual elements
of the community.
There are potential community management information needs
which are either only partially satisfied by CIRIS together with these
related systems, or are not addressed at all. Some representative
examples of needs derived from Mr. Colby's 6 September 1973 memo-
randum to the President, "Objectives for the Intelligence Community",
the USIB FY 1974 "Key Intelligence Questions" memorandum, and
other sources are:
(1) the IRAC quarterly review of resource utilization by all
intelligence agencies.
(2)
the IRAC R&D Advisory Council review of R&D activities
within the National Intelligence Program to identify
on-going R&D efforts, their costs, their purpose, and
the management responsibility for each effort.
(3) the DCI review of the assignment of intelligence functions
within the community and the worth of intelligence products
to consumers in order to eliminate marginal, inefficient,
unnecessary, or outmoded activities.
1
Approved For Release 2006/06/23: C1A-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
(4) the DCI review of the national/ tactical intelligence inter-
face with respect to assets, capabilities, and mutual
support opportunities in order to reduce unnecessary
overlap or duplication in the national and tactical areas.
(5) the evaluation, beginning in FY 74, of the performance
of the community and the individual agencies in respond-
ing to Key Intelligence Questions.
(6) community efforts to improve insight into substantive intelli-
gence issues and to relate these issues to cost.
These examples imply at least the need for a post audit capability.
On the other hand, such overview needs are not necessarily of equal
importance to Program Managers in the detailed discharge of their
responsibilities. The apparent information needs of the community are
diverse and do not necessarily lead to similar solutions. Questions,
such as data timeliness and credibility, data-call timing as related to
the I ycle, vulnerability resulting from data submitted, standards--
at least in terms of compatible formats and use of common terms, and
the value to the user compared to the cost and effort required to support
an information system, all develop into significant issues in any serious
consideration of the community management information problem.
One ''system'' most likely cannot and should not attempt to
serve the needs of the entire hierarchy of users. The problem then is
one of relating expressed or implied user needs for community manage-
ment information to the options which can be developed for satisfying
these needs and the trade-offs involved in view of these options. This
is the IWG view of the task assigned to it.
The review was conducted through a series of related tasks.
These are described in the Terms of Reference (TAB A). The follow-
ing sections of this report describe the results obtained under each
task, leading to the conclusions and recommendations in the final sec-
tion.
TAB B lists the IWG participating organizations and individuals.
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 25X1 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
C. Needs
(1) Needs of the DCI and the Intelligence Community Staff:
The President's Directive to the DCI of 5 November 197:1
directed the DCI to assume leadership of the community in planning,
reviewing, coordinating, and evaluating all national intelligence pro-
grarn and activities, to review and provide judgments on the efficiency
and effectiveness of all intelligence programs and activities (including
tactical intelligence), and to recommend the appropriate allocation of
resources to be devoted to intelligence. To fulfill these responsibili-
ties, the DCI and his supporting Intelligence Community Staff need
management information for resource distributions, and for perfor-
mance evaluation. To a great extent, these needs are complementary,
but their requirements are very separate and distinct. The DCI's planning,
program review, and reporting functions require such data as: (a) resource
for past fiscal years to trace changes and trends in manpower and funding;
(b) projections for two to five years of resource needs as a result of re-
source decisions already made; (c) distributions of funds by major cost
categories, including R&D and construction as well as O&M, etc. ; (d)
such records as the POM's, PDM's, PBD's, FYDP, and broad issue
studies; (e) descriptions of resource capabilities and subordinations;
(f) during a program year, data on changes from reprogramming, re-
source decisions made within program structures, or diversions of
funds originally programmed for intelligence to uses outside the intelli -
gence community. The basis for these needs from the viewpoint of
the DCI's Intelligence Community Staff is fully described in TAB C.
(2) The Performance Evaluation System: The DCI's re-
sponsibility to evaluate intelligence programs and activities and
render judgments on their efficiency and effectiveness requires a
data system which does not now exist. Data input for such a system
are essentially measurements of performance against tasks, and/or
national intelligence objectives, and there is little precedent or ex-
perience with the generation, analysis, and reporting of such in-
formation. This problem is discussed in Section D.
(3) The Needs of Other Agencies:
As is stated elsewhere in this study (CIRIS, Sec. 2a),
other agencies have no essential or critical need for CIRIS, and make
little direct use of it, except for DIA which currently has no manage-
ment information system of its own. Because data in the agencies'
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 20@3Nj2t ~RE,82B00871 R000100300003-6
data bases must be accurate, reliable, and timely for agency manage-
ment purposes, a future community-wide resource data system to support
the DCI would do well to rely heavily on the separate, distributed agency
data bases. The price of such reliance will be a broader mutual agree-
ment on program structures, cost allocation procedures, definitions
of data, timing of updates, subject and geographic categories and
levels of aggregations required. Working out such an agreement could
be an important task for the IC Staff and appropriate community bodies
in support of the widened community responsibilities of the DCI.
E valuation
Evaluation is expected to assume increasing importance as
means are developed to assess the performance of intelligence entities
against the objectives being established and to serve as a basis for the
formulation of strategies against objectives. The President's instruc-
tions to the DCI stress that he is to evaluate all intelligence programs
and activities and that he is to provide judgments on the effectiveness
of intelligence programs. The ultimate effectiveness is measured in
terms of the value of the intelligence product to its customers.
For the purposes of this consideration, evaluation is under-
stood to be a process of value determinations - performance vs expecta-
tion vs cost - to support judgments regarding improved resource alloca-
tion. Analysis is here considered to include the identification, dissection,
aggregation, and manipulation of data to support the evaluation process.
The objective of both evaluation and its related analysis is, in short, to
obtain as much useful performance as possible from a given level of invest-
ment or to obtain a given level of performance for as little investment as
possible.
Conceptually, the evaluation problem can be approached by
developing information on each of the following:
(1) Intelligence objectives, reflecting current and antici-
pated needs and community tasking against these objectives.
(2) The resources made available to address these
objectives, and their distribution.
(3) The strategies for the employment of these resources.
(4) The implementation of these strategies and the re-
sulting performance against the objectives, including the value of the
resulting products.
11
Approved For Release 200wq.(2,:,1~17RQF$2B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
One could then identify high and low payoff applications of
resources, and evaluate the impact upon future performance of greater
or lesser availabilities of resources, consistent with various manage-
ment strategies.
Today, intelligence objectives, including Key Intelligence
Questions, are being defined. The intelligence community is not in
a position to apply this conceptual approach to them because the re-
quired information base and methodologies are not yet available. A
crucial element in the evaluation process - Step 4 - is a means of
measuring performance in terms which are commonly applied to all
elements within a general intelligence function - collection, processing,
production and support. The development, community-wide, of such
means of measurement, or standards of performance, has understanda-
bly been hampered by political, fiscal, bureaucratic and individual
reluctance to be judged. The financial constraints on the community's
total budget, however, year by year call for a more sensitive means
to differentiate between productive and marginal components and to
distribute moneys to where they yield the greatest result.
Realistically, evaluation today can be considered in terms of (1)
CIRIS and its application to the problem, (2) an interim means respon-
sive to the DCI's evaluation of community response to his objectives
and the Key Intelligence Questions, and (3) considerations for a longer
term solution approaching the goals of the conceptual process described
above.
1. CIRIS and Evaluation
CIRIS data do not contain elements reflecting the effective-
ness of the activity to which the data refer. The addition of effective-
ness measures and automated evaluation data to the current CIRIS
structure is impractical. Moreover, the relationships at present
among CIRIS topical categories, Key Intelligence Questions and DCI
Management Objectives are tenuous and, at the moment, are diffi-
cult to combine in a single structure or system. The choice is either
to add the key questions and objectives to the CIRIS matrix or to set
up a separate reporting system. Because the key questions and objec-
tives are subject to dynamic change over time, inclusion of them in
the CIRIS matrix can adversely perturb that matrix. Furthermore,
the narration required from each reporting entity does not lend itself
to CIRIS formats and suggests an eventual system, separate and dif-
ferent from CIRIS.
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 20061O 3iii tA-'F '.8,2B00871 R000100300003-6
If there is a separate system, it will be important that the
two interrelate as closely as is possible the identified resources in
one system on one hand with the measurements of progress toward
goals, objectives and priorities on the other. Some way must be
found to translate Key Intelligence Questions, even though they will
change, into general subjects which can relate to the resource in-
formation structure. No program has been undertaken to accom-
plish this, but a first step is to seek general recognition of the
problem.
In view of the foregoing, it seems probable that there will
be evolution toward these complementary systems:
a. a budget system to record resources provided
managers,
b. a reporting system relating to objectives and KIQ's
for review of the tasking and performance of managers,
c. an eventual display of a program structure of resource
allocation to show the structural relationships (collection,
processing, production, and support) that are often distorted
in organizational structure but that need to be displayed for
planning purposes, understanding of interdependencies, and
the review of objectives with consideration of those inter-
dependencies.
The role of CIRIS in this evaluation has not been determined,
but it is clear that it should not be significantly modified for CY 74 in,
an attempt to address the evaluation question.
2. Interim. Evaluation of Response to KIQ's & Objectives
The purpose of an interim evaluation capability, operable
in CY 74 is to meet the DCI's need for evaluation of the community re-
sponse to his Objectives and Key Intelligence Questions. GIRIS cannot
do this and is not so structured. The more analytical approaches
which have been considered for evaluation are, at this point, under-
developed, controversial, and not capable of meeting the requirement
during the next calendar year. The following options, however, can
be considered.
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 20(/
(9WD-f82B00871 R000100300003-6
A. Program Manager Option
The simplest option is initially to request the pro-
gram managers to determine the evaluation method best suited
to their operation and report in their own format, their response
to each KIQ against which they were tasked. The Intelligence
Community Staff would then convert these narrative reports into
an Agency/KIQ matrix for overall evaluation.
B. Collection Evaluation Option
A second option directs attention to performance of
collection resources against the target questions embodied as sub-sets
of the KIQ's. In a series of steps this community evaluation process
requires:
(1) Identification of the answers needed to respond
to a KIQ.
(2) Identification of specific information required to
support an answer.
(3) Identification of what data are required to generate
the sufficient information called for in (2).
(4) Cross-relation of these data to the collectors whose
prescribed tasks include collection against such re-
quirements.
(5) Identification of those collectors and the primary
recipients of their ''take''.
(6) Questioning of the recipients as to their judgment
of how each collection system - to the extent they can
identify them - assisted their analysis.
(7) Questioning of the supervisory management level
as to the distribution of reaction to the publications
which embodied the analysis by those in (6)
(8) Questioning of the final recipients or users of
the reports prepared by those contacted in steps (6)
and (7) above.
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
This option illustrates some of the problems inherent in
this conceptually straight-forward evaluation problem. The relation
between collected data, information derived from a number of collec-
tors or a time history of collection, and the relation between this
information and intelligence needs, Key Intelligence Questions or
Objectives can be difficult to identify and quantify. Determining
value of a collection activity or product through questioning analysts
and consumers is vulnerable to variation and subjectivity.
C. Evaluation of Products Responsive to KIQ's
A third and perhaps more direct option considers the
five-part response requested from USIB agencies in response to Key
Intelligence Questions described in Mr. Colby's 30 October 1973
memorandum to the NSCIC (Para. #3). Evaluation of these responses
could be performed as follows:
(1) For each KIQ/Objective tasking, identify products
which will contain the response. Confirm and document
priority tasking of collection and analytical resources.
(2) Document current information baselines, estimates
and uncertainties.
(3) Identify important collection or analytical gaps.
(4) Identify collection and analysis strategies to fill
the gaps, together with the time frame required to
do so.
(5) Monitor through narrative reporting and displays
the results of strategies as applied to the gaps and
the increase in information resulting against products
addressed to the KIQ/Objectives.
Given the results of (5), it would then be reasonable
to consider estimating the cost of providing the incremental intelli-
gence, so that cost-benefit analyses could be performed.
3. Longer Term Considerations
There has been considerable effort applied to developing
evaluation methodologies applicable to community activities, such as
15
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
resource collection and performance evaluation. Some of these are
described in the following few paragraphs.
Dr. IResource Allocation Model
This is a good example of an effort to apply available
analytical techniques to the intelligence resource alloca-
tion problem. It is the basis for a potential future system
candidate described in TAB D. If successful, it could, for
example, show various incremental benefits to be derived
from various funding options. In its present form, its
credibility may be questioned for top management use be-
cause:
(1) The validity of the sample of the community it takes
has not been verified, and some observers feel that
some preliminary allocation conclusions are subject
to question.
(2) The approach uses determination of relative values
and the construction of utility functions by analysts
whose natural biases could flavor the results.
(3) The technique assumes that a valid requirement-
to-collection "observable tree" can be constructed.
(4) The model is so complex that a user cannot easily
identify the effects of individual judgments exercised
during construction of the model.
Mr. IDecision Analysis Approach
This model is another interesting approach to provide
quantitative measures in support of intelligence resource
decision making. It attempts to determine the value of im-
proved intelligence in dollars, so that it can be compared
to the likely cost of obtaining the improved intelligence.
Results obtained from his model are credible, but again
they suffer from a dependence on judgments not easily
visible to the decision maker, and complicate making
intuitive insights into the model.
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
Other approaches include Mr.
appli --
cation of the performance evaluation technique
a DOD-sponsored cost benefit analysis study.
(TAB D) and
One of the problems encountered in this review was an
apparent lack of community agreement on the most essential ques-
tions that justified development of evaluation methodologies, and
a coordinated community effort to address these questions. Before
deciding which of the available techniques should be applied and
evaluating the impact on present or future community information
data bases, it would be useful to seek broader community agree-
ment on the need for such advanced analytic approaches.
3. Conclusions and Recommendations
Conclusions from the tasks described above, can be separated
into three categories: short term considerations for the CY 74 CIRIS
data call, interim evaluation of community response to DCI objectives,
and considerations for the longer term development of an efficient,
responsive community management information system. Criteria
considered important in selecting these recommendations include
responsiveness to the essential needs of the intelligence management
structure, maximum use of existing or otherwise necessary informa-
tion systems and processing facilities, implementation that will en-
courage community support, and equitable distribution of support
effort and expense.
Conclusions
1. One user, the Intelligence Community Staff, considers
CIRIS essential to the performance of its responsibilities
in support of the DCI with respect to program review and
identification of the resources associated with activities
the DCI must evaluate. One user, DIA, considers CIRIS
very useful, and loss of continuity in CIRIS data would
create a hardship in carrying out assigned management
responsibilities. In the view of ASD (I), CIRIS is far too
elaborate and has not been helpful. (Minutes ?:.R..A;_"
meeting - 5 November 1973).
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
2. The usefulness of CIRIS to the IC Staff could be in-
creased by additional detailed information (TAB C). In
the long run, the wider adoption of community data
standards, the potential increased use of data bases
distributed within the community, a desire to minimize
the cost and effort involved, and a need to decrease the
burden on diminishing staff resources should reduce the
detail required in a CIRIS data call. As a practical
matter, little change can be made in the current instruc-
tions if they are to be used for an early CY 74 data cal]..
3. In the view of the Intelligence Community Staff, the
usefulness of CIRIS would be increased if information
could be made available during the July/August review
period, including targeted information for the program
year 1976. This would require acceleration of the sub-
mission schedule and the targeting of information for the
program year, not presently required.
4. This is contrasted to the lack of essential need ex-
pressed by other community members. Within DOD,
DIA is the only significant user and this need will be
substantially reduced when the proposed IMIS system
is implemented.
5. The DCI must be supplied the information he needs
in the exercise of his community responsibilities. One
of these responsibilities with a significant impact on
information needs is the program review function. The
manner in which this is to be done by the DCI and his
staff needs to be established before the essential data
needs can be determined. This can either significantly
simplify or expand data call requirements.
Recommendation
Given a determination of the DCI information needs, IRAC,
through this working group or another ad hoc committee,
should advise him in his consideration of the benefits to
be gained, and of the effort which will be incurred in the
community response to a CIRIS data call, including
changes recommended through consideration stated in
the previous paragraph.
18
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
B. Interim Evaluation of the Community Response to Key
Intelligence Questions, and DCI Objectives
Conclusions
1. This need for evaluation of community response to
the KIQ's or DCI Objectives has been repeatedly empha-
sized (TAB C and minutes - 5 November 1973 IRAC
meeting) as an essential, new requirement for informa-
tion, to be implemented in CY 74.
2. CIRIS cannot meet the requirement and should not
be considered as a candidate means to that end.
3. Initially, a separate, narrative report should be
used pending longer term consideration of means for
future evaluation and analysis of community perfor-
mance.
Recommendation
Unless current IC Staff consideration of this problem
recommends a simpler solution, adopt Option C, Section
D (2). This calls for narrative reports, consistent with
Mr. Colby's 30 October memorandum to the NSCIC con-
cerning Key Intelligence Questions, to do the following:
1. For each KIQ/Objective tasking, identify products
which will contain the response. Confirm and document
priority tasking of collection and analytical resources.
2. Document current information base-lines, estimates
and uncertainties.
3. Identify important collection or analytical gaps.
4. Identify the collection and analysis necessary to fill
the gaps, together with the associated time frame.
5. Monitor through narrative reporting and displays
the results of identification of tasks in (4) to fill gaps
identified in (3) as they affect improvement in the
information in baselines and estimates in (2) to be
reported in products identified in (1) above.
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
6. Given the results of the monitoring in (5), consider
estimating the cost of providing the incremental intelli-
gence, so that cost-benefit analysis can be performed.
C. Longer Term Considerations
Conclusions
1. Despite the fact that CIRIS is little used outside the
IC Staff, future community needs are likely to require
some form of centralized, highly aggregated, manage-
ment information system which will give visibility to
the distribution of community resources and related
resources outside the community and provide support
to the DCI and USIB and IRAC committees.
2. This future system should fully exploit and not
duplicate the separate management information systems
used throughout the community.
3. In order to provide effective crosswalks among the
data contained in these related community systems,
serious consideration should be given to greater uni-
formity of data standards and formats.
4. The needs of the community to perform analysis and
evaluation in support of senior decision makers should
be identified, and development of promising techniques
should be subject to greater coordination and discipline
throughout the community.
Recommendations
1. Establish through an ad hoc working group or the
USIB/IHC an activity addressing community resource
management data standards.
2. Establish through an ad hoc working group or the
USIB/IHC an activity to develop the concept of making
maximum use of data bases distributed throughout
the community and minimizing the need for a redun-
dant central consolidated community data base.
20
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL
3. Establish through an ad hoc working group or the
USIB/IHC a means to continue examination of the
critical evaluation and analysis needs of the community
management structure so as to coordinate and encourage
the development of those advanced techniques which
show promise.
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL TAB A
Attachment
IF AC-D-75. 1 / 1
24 January 1974
OBJECTIVE
Review CIRIS and related intelligence management information
systems with the objective of developing recommendations for the
implementation next year and beyond of CIRIS or some other follow-on
system which will:
1. Provide a community-wide basis for characterizing
the use of resources.
2. Provide a data base and means of analysis and eval-
uation to support senior managers in meeting their current and
future objectives.
3. Provide a management information data base and
inventory of assets to support USIB, IRAC, and NSCIC
activities, and
4. Be responsive, cost effective, and minimize dupli-
cative effort.
TASKS
1. Carry out a Current Systems Review
A. CIRIS
1. Review current CIRIS capabilities and data
structure.
2. Review user experience and use of CIRIS to
date -- beneficial use, problems, and reasons
for non-use.
B. Related Systems
1. Review capabilities, data structure, and attributes
of contributing and/or related intelligence manage-
ment information systems.
oEiD ELI Li C;IA-I~~ TAB "A"
Approved For Release 20&/d123 : h2B0087 1 R000100300003-6
 Approved For Release 20066R ,I 414 P88B00871 R000100300N~6A
2, Review current and/or expected use of these
systems.
2. Identify User Information Needs and Requirements
Identify information system needs of the user
community and current or potential applications
of a community-wide data base.
For each requirement, identify the end use of
results and current means, if any, of satisfying
the need.
Separate user needs by major category, such as:
a.
Planning
b.
Preparation of the Congressional Briefing
c.
d.
e.
Preparation of the National
ligence Program Budget
Program Review
Evaluation
Foreign Intel-
3. Review significant problem areas as may be identified,
such as analysis and evaluation needs and methodologies and treat-
ment of support costs.
4. Summarize and report results with recommendations by
mid-December 1973.
Approved For Release 2006109/~.IUA ff 00871 R000100300003-6
 25X1 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006 Riiq R8L2B00871 R000100300PRJ56C
Attachment
IRAC-D-75.1/?
24 January 1974
DCI - INTELLIGENCE COMMUNITY STAFF NEEDS
User Needs
Within the intelligence community, the management information
systems, with the exception of CIRIS, have been designed to support
requirements other than. those of community-wide management, Late
in 1971, the President directed the DCI to assume leadership of the com-
munity in planning, reviewing, coordinating, and evaluating all intelligence
programs and activities and in the production of national intelligence. He
was also directed to improve the performance of the community, to pro-
vide his judgments on the efficiency and effectiveness of all intelligence
programs and activities (including tactical intelligence), and to recommend
the appropriate allocation of resources to be devoted to intelligence. In
response to this direction, the DCI has assumed a number of responsibilities,
specified in his letter to the President of 6 September 1973, and has also
established a set of management objectives to be implemented within the
community. A set of substantive objectives is now being developed. If
the DCI is to perform his responsibilities and to be confident that the
necessary intelligence support is being provided the national authorities,
more resource and performance information must be made available to
him on a timely basis than .is the present case. 25X1
From the description, it is evident that the data required by the
DCI are diffused widely. The three main sources of useful data are CIRIS,
the FYDP, (in its several states of development) and the CIA Congressional
Budget. None of these supplies him with information on effectiveness, and
at this time he has little basis for addressing the "appropriate allocation of
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/CQ~/?3F,:ISI~NF'A8~B00871 R000100300TAB C
resources". Yet the President looks to him to advise on the potential
impact of alternative levels of resources (usually cuts). The President's
Directive specifies that "All information required from all departments
and agencies of the Executive Branch is to be made available to him (the
DCI) in order that he may provide me with an annual detailed review of
the needs and performance of the intelligence community".
The critical words in this instruction are "information required".
The key variables that affect the requirement are:
The management system.
The timing and content of actions at the DCI level.
Though the requirement might be described in terms of present require-
ments, the information requirements are themselves evolving, hence the
information system to support the DCI should be designed with a view to
some growth potential.
The Management System
The Presidential instruction (November 1971) emphasizes DCI
review, coordination, and recommendation on resource allocation. It
leaves operational responsibility where it is currently found. Accord-
ingly, the DCI envisages a coordinated but decentralized mode of man-
agement in which he provides guidance, defends the community level
of resources that he deems necessary, exerts influence in the alloca-
tion of resources, and asks managers to report on the fulfillment of
DCI-approved objectives.
In a decentralized system with a large number of decision
centers, two common features are that (1) a decision by one unit may
have significant consequences to another unit (need for coordination)
and (2) many units may exercise similar functions (such as training,
logistics, communication, staffs) so that combining such commonalities
might provide significant economies.
The current theory of managing large systems suggests that
where such features are present, the information system should stress
visibility of costs, interrelationships, and consequences, because a
large management system can be self-policing once a problem becomes
visible. It is necessary, however, to recognize that often there are
CONFIDENTIAL
2
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL TAB C
strong elements in current data flows reducing visibility. Among these
are partial costing (such as not budgeting for a tenant relationship on a
base or such as carrying retirement costs outside the immediate budget),
differences in program structure, discrepancies in costing procedures,
arbitrary and inconsistent treatment of indirect costs, failure to relate
funding to function, and absence of output data.
The post-audit function suggests a need for community manage-
ment systems, with resource data as well as output or performance
information. In component level annual reports, where they exist,
many activity measures are cited, but they are rarely reported on any
continuing basis to higher management levels. Lack of output informa-
tion hamstrings any central effort to evaluate effectiveness, except
on the most subjective basis, and also weakens budget defense by ob-
scuring the relationship between workload and resource requirements.
The problems here are (1) the relations between quantity and value
which are not necessarily strong in intelligence and (2) the justification
of much of intelligence not in terms of current activity but in terms of
contingency value. Neither of these problems is insurmountable, how,
ever. Guidance can be provided on capabilities desired; management
and substantive objectives specified; performance against objectives
measured in both qualitative and quantitative terms, as appropriate;
and output measures developed for a wide variety of essential tasks
such as communications, processing, and support services. The DCI
has assigned priority to the task of developing systems for evaluating
collection and production.
The Timing and Content of Actions at the Community Management Level
The data requirements relate to six key functions: planning,
Congressional presentations, the annual report to the President (The
National Foreign Intelligence Program Budget), the program review
process, evaluation of performance, and recording actual resources
available. At the present time, program review, Congressional pre-
sentation, and the annual report dominate the data requirements, but
the evolution of the management system would probably dictate increasing
attention to the other three functions, particularly evaluation.
A. Planning
At the present time, the key elements of community-
level planning are the Perspectives paper and the promulgation of com-
munity-wide management and substantive objectives. Since guidance
3
Approved For Release 2006/0-(Op$~P00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL TAB C
plays an essential role in the coordination of a decentralized manage-
ment system, it is to be expected that the planning input behind the
setting of objectives will increase. There will also be greater atten-
tion to the out-year consequences of program and major project
decisions.
Data requirements and timing: DCI-level planning guidance
for the forthcoming program year is normally presented within the
November-December time frame*. Useful inputs to it are (1) data
for the preceding fiscal years that will aid understanding of structural
changes within the community by depicting trends in manpower and funds
allocated to programs/program elements, organizations, and target
categories. (Agency budget and FYDP are now the major sources. )
(2) Data on the effectiveness of collection systems, processing, and
production (not in a management system now). (3) Projections of
future consumer priorities for intelligence (these are currently outside
a management information system). (4) Projections (3-5 years) of
resource requirements related in major program decisions already
made (these are in the FYDP only). (5) Personnel data depicting past
trends and future projections, with particular emphasis on retirement
and quit profiles, career patterns, age/grade structure, education/
skill inventory, and anticipated changes in the education/skill mix
requirement. (Such personnel data are not currently used to any
significant extent at the community management level. ) (6) Data on
performance against objectives. A point to emphasize is that inter-
agency comparability of program structure becomes more important
as the emphasis on planning increases, because a well-designed pro-
gram structure gives a picture of the structural relationships in the
entire process of intelligence, while relating in a meaningful way to
planning parameters.
B. Congressional Presentation
The system is geared well to provide Congress budget
and program data justifications, so far as the justifications go. The
weakest link is the absence of any comprehensive assessment of the
effectiveness of intelligence programs, As a result, it is virtually
impossible to give any convincing portrayal of the impact on effective-
ness of the resource level proposed or of a change in resource level.
' SecDef Guidance is issued in February, DCI guidance can be made
concurrent.
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL TAB C
Data requirements and timing: DCI Budget presentations have
to be ready by April or May preceding the budget year, although they
may not be given until well into the budget year. Resource data are
presented according to functional (program), organizational, and budget
cost classifications. The presentations often give some highly aggre-
gated data on resources applied against targets and give, usually in
anecdotal fashion, some sense of the effectiveness of activities. The
data base should also provide some capability to provide information
on activities that might come up for Congressional queries; there is
perennial interest in R&D, construction, and external contracts, for
example
C. National Foerign Intelligence Program Budget
This report serves as the "annual detailed review of the
needs and performance of the intelligence community". As such, it is
usually prepared by October or November, when the data are available
for the performance during the preceding fiscal year and when PDM and
CIA program data are available for the forthcoming program year
(currently, FY 1975). Last year, this report provided information on
the national intelligence program with respect to: programmatic structure
of the program year compared with preceding years; the time series for
the NIP in current and in constant dollars; cross-program structure for
collection (and major systems), processing, production, and covert
activities; cost data by appropriation categories, and some highly
aggregated data on targets and priorities. This year the report will
include a section on the development of management and substantive
objectives. Next year the report will likely review progress toward
the accomplishment of the objectives that have been set.
D. Program Review
"Recommendation of the appropriate allocation of
resources to be devoted to intelligence" hinges upon DCI participation
in program review. The key inputs are those used by program review-
ers in the Agencies and the Departments, especially the POM and the
CIA program submissions. The IC Staff participates with ASD(I) in
the review of DoD intelligence programs. This participation impacts
somewhat upon the PDM and does affect the DCI's recommendations
with respect to the National Intelligence Program.
Data requirements and timing: The main crunch period is June
through August, preceding the PDM's, with a secondary crunch during
September through November, preceding the PBD's. The central data
inputs are the program and budget submissions, which for the DoD take
5
Approved For Release 2006/00/3&1qtj f 00871 R000100300003-6
 Approved For Release 2006@1~(I?lhlTrpf'fO00871 R000100300 C
the form of the PO.M, the PDM, and .issue studies. Reviewing officers,
who must consider capabilities and effectiveness, would benefit from an
attributes file descriptive of intelligence units and capabilities and from
a target/regional breakout of the program data. The inventory file,
which exists in rudimentary form today, should be expanded to serve
multiple purposes by describing mission, funding, manpower, locations
of activities, capabilities (including targeting), cumulative capital
investment, and subordination. It probably should be expanded to
include separate entries for major intelligence projects (operational
and R&D) and to list cryptonyms and nicknames wherever appropriate.
CIRIS currently does not present target/geographic breakouts
for program year data. As a result, review officers make little use of
CIRIS except for background, when they use earlier year data. A break-
out of target/geographic allocations for the program year would be help-
ful.
As the community moves toward management by objectives, it
will be necessary to keep a systematic file (probably in narrative form)
of objectives, progress to date, and resources attributed to the effort
against each. The file should be limited to objectives of DCI interest
and those Agency and Departmental objectives that immediately support
community level objectives.
Both program review and budget review at the community level
would be assisted if a computer-assisted file could be established within
the IC Staff which could run updates on the data starting with the POM
submittals, picking up the PDM, the PBD, apportionment updates, and
significant reprogrammings. This could be handled internally with some
formatting instructions to the contributing entities.
E. Evaluation
Evaluation is expected to assume increasing importance
as means are developed to assess the performance of intelligence entities
against the objectives being established and to serve as a basis for the
Formulation of strategies against objectives. The President's instructions
to the DCI stress that he is to evaluate all intelligence programs and
activities and that he is to provide judgments on the effectiveness of
intelligence programs. The ultimate effectiveness is measured in terms
of the value of the intelligence product to its customers; the product
review process is being designed to tackle this aspect.
6
Approved For Release 2006/06/23: ClA 1NR5 M fR000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
CONFIDENTIAL I' AB C
Evaluation is a function that should be performed at each level
in the organizational hierarchy with respect to subordinate activities.
At the DGI level, it should be concerned with relatively aggregated
data but should have access, upon request, to more detailed informa-
tion where required.
Some of the key data requirements likely in any pre-evaluation
(strategy) or evaluation system are: a matrix of programs/sensors/
platforms arrayed against a matrix of objectives /subjects /geography
displaying effectiveness values and some breakout of aggregated costs
for O&M, procurement, R&D. The system for deriving the effective-
ness values is likely to be judgmental. The cost data will be profoundly
influenced by the way in which joint costs are allocated. The Key Intel-
ligence Questions are likely to generate this kind of display.
F. Recording Actual Resources Available
Monitoring the actual allocation of resources is
usually an integral step in the management process and is a necessary
step if the effectiveness of the decentralized system is to be assessed.
As a practical matter, however, the intelligence community is not a
closed system and other authorities (Services, Commands, Agencies,
and Departments) can make independent decisions that affect the avail-
ability and use of intelligence resources. Even more, resource report-
ing is incomplete because important data on reprogramming and on the
availability of manpower (on-duty strength) are not even in the informa-
tion system available to the DCI. Finally, the absence of effectiveness
measures is a gap that thwarts effective coordinated management.
Reporting of progress against DCI and program manager objectives
(managerial and substantive) would tend to reveal problems of coordination
faster than they would otherwise be revealed and hence may strengthen
cooperative management in spite of the basic weaknesses within the data
system. Systems are being considered that might provide such report-
ing.
G. Special Studies
Special studies are a means to get information that is
impossible or impractical to collect routinely; hence they are an impor-
tant adjunct to the management information system. Of course, matters
that are frequently the subject of special studies are candidates for
inclusion in routine reporting.
7
CONFIDENTIAL
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006406j :I0IARiDfNO00871 R000100300 3j C
The options for the next GI.RIS Data Call are limited for prac-
tical reasons to Incremental changes in the data submission instructions
which would "tighten up" the accuracy and/or relevance of the data base.
Additional types of management information may be found necessary for
effective intelligence community management. Requirements cannot be
specified in time for the CIRIS 74 Data Call.
The options described are intended to reduce some of the ob-
stacles to data utility which have been Identified by the group. Some
may be in conflict with others; for example, requirements for additional
descriptive data on Reporting Entities will increase the data preparation
time and tend to work against an earlier completion for data base up date.
Therefore, options should be chosen carefully to insure that system reach
does not exceed its grasp.
Previously, the annual update has not been completed
until after program reviews. The deadline for Reporting Entity submis-
sion could be changed from late May to early April, resulting In CIRIS
74 data being available in mid-June. This compression could produce
an increased data accuracy problem. Attention to accuracy of submitted
data would have to be increased since a great amount of preparation
effort must occur after data reach the Data Support Group of the IC Staff.
B. PROGRAM YEAR DATA
Given the above, there is still no provision currently
for targeting program year resources. The extremely detailed resource
targeting procedures required by CIRIS are not compatible with the level
of uncertainty in Reporting Entities concerning Program Year resources.
Experience shows that even Budget Year figures tend to be straight line
projections of the Current Year targeting. One other option is to target
Budget Year and Program Year resources in percentage level of effort
terms. This option should be considered in combination with the one
above since the purpose of each is increased utility for CIRIS in program
review.
Approved For Release 200616,1 ' 8-,2,600871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001003,r%?3c6
CONFIDENTIAL
C. SUBJECT /GEOGRAPHIC CATEGORIES
The current lists of subjects and geographic areas
for targeting is a reasonable compromise between (a) the require-
ment to report targets fairly specifically if the data eventually are
to have any analytical value and (b) the reality of the vacuum cleaner
characteristic of many collection and associated processing systems.
The option exists to adjust this compromise in either direction, but
the experience to date with "basket" categories indicates that more
specificity in the data is an illusory goal at present. Simplifying the
subject/ geographic lists would probably not degrade the system for
most managerial purposes. It would reduce the utility of the system
for special studies and detailed analysis, such as assessing program-
med effort against specific substantive objectives.
Aside from the benefit of simplifying (and presumably speeding)
data preparation, a more concise subject/geographic matrix would offer
the DCI an opportunity to establish that as the official Community matrix,
considering compatibility with DCID 1/2. All other information systems
should be capable of aggregating data to the mandated categories.
D. COST DE FINITIONS
Although CIRIS definitions are derived from official
DoD sources and are technically sound, it is evident from looking at
time series data on certain support cost categories that submitters
have a great deal of latitude as to how costs are classified.
One option would be to combine the current Positive Mission
Support and General Support categories. This would reduce some of
the inconsistency in CIRIS, albeit at the expense of some detail. It
would also lead to a lower mission/support cost ratio.
Another option would be to expand upon the explanations of
cost classification in the CIRIS instructions. In effect, this would mean
developing accounting rules for the Intelligence Community. Since this
would be an iterative process, the option actually is to make a start
rather than to complete the project this year.
E. MANPOWER DATA
CIRIS now concerns itself with end-of-year authorized
strengths. If the advantages of using FYDP data as control figures are
9
Approved For Release 200l?/PYA1PI&2B00871 R000100300003-6
 Approved For Release 2006/06/23CORENQ$L7,1 R00010030-j
to be realized, this characteristic should be retained. An option does
exist to report actual strengths, also. This would necessarily be his-
torical data rather than projected, but it would provide useful post
audit information, with a limited predictive capability.
F. INVENTORY UPGRADE
The amount of descriptive information each Reporting
Entity provides could be increased by either redesigning the forms or
by prescribing certain entries in the narrative portion of the current
form. Data required for the facilities inventory of the CCPC, for
example, could be submitted with CIRIS data. This would entail specify-
ing how conspicuous equipment and capital investment is to be reported.
Also, personnel assigned to the Reporting Entity but stationed
in a different country from the Reporting Entity headquarters could be
identified,
As a forerunner to a more definitive capabilities inventory, the
descriptive narrative could note contingency capabilities of the Reporting
Entity which are not evident from the submitted data.
G. RESOURCE CATEGORIES
Currently, resources in the CIRIS data base are cate-
gorized into approximately 35 cost and 10 manpower types. Each cate-
gory must be targeted, and this accounts for much, if not most, of the
complexity and volume of data submissions. The option exists to reduce
the resource categories and simplify data requirements, although there
would be a cost in utility.
CON FhO EN TIC L
Approved For Release 2006/06/23: CIA-Rb'l B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6
 Approved For Release 2006/06/23: CIA-RDP82B00871 R0001 00300003-6
Confidential
Confidential
1 L4 W 18 Z 9z
Approved For Release 2006/06/23: CIA-RDP82B00871 R000100300003-6