RECORDS MANAGEMENT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP84-00780R005500110011-0
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 10
• Document Creation Date: December 19, 2016
• Document Release Date: September 29, 2006
• Sequence Number: 11
• Publication Date: August 1, 1973
• Content Type: MEMO

Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
PPB
973
Subject: Records Management
Per your request, some thoughts stemming from your meeting of 22 Aug 73
concerning the Records Management issue:
1. How does the notion of trying to manage records arise? Apparently,
because records are generated in great volume at great expense, move through
their active lives using up large amounts of manpower and space, and finally
come to rest in inactive storage which again has major costs associated with it.
2. Records are thoughts in physical form. In a way, the difficulty of
controlling the creation and use of records is comparable to trying to control
the thoughts of an organization. I know that Records Management doctrine is
intended to cover the record from creation through its active life to archival
storage or destruction. In practice, however, records management has found
its principal effectiveness limited to inactive or archival storage... in spite of
the determined efforts of earnest people over the years. Somehow, it has proved
very up-hill for Records Management programs to impact significantly on records
generation or the active life of records. There must be a reason for this.
3. Records are the molecules of an organization. Intuitively, it's a very
forbidding prospect to manage a pervasive aspect of an organization by under-
taking to manage the molecules of that organization. It is like trying to manage
a building by giving attention to its bricks. Organizations do not enjoy such
exquisite degree of control... in records or anything else. Perhaps this explains
why Records Management, concentrating as it does on records, has been relatively
ineffective in its efforts to "manage" the creation of records and the active life of
records.
(What we should manage, it would seem, are the plans,
programs, projects, and people in an organization...
but not the records as such which are generated by these
plans, programs, projects, and people. And managing
`y1f a ..V :C F Pages 1 thn_a ti
 Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
the Agency's plans, programs, projects, and people is
the function of the supervisory system of the Agency .
Efforts at strong management of a pervasive aspect of
Agency activity like records from some central body
would be in some way in conflict with the Agency's
supervisory structure and would appear certain to
cause confusion and achieve little good . )
4. Given the thesis that it really isn't feasible to control costs in an
organization by attempting to "manage" records as such throughout the
Agency (except archival records), it would appear that whatever wisdom or
restraint we employ in generating and using records must become an inherent
part of the normal work habits of all employees throughout the Agency. This,
in turn, would suggest that what is needed from some central body (small staff)
in the Agency is a few practical rules which can be followed by all of us in the
paper world in which we work... together with widespread training in those
rules. Such rules (they must be easily do-able) and compliance with them may
achieve the lion's share of all we can hope to achieve in the way of "management"
over active records.' If so, we should avoid ponderous central mechanisms or a
large body of people straining at the full doctrine of records management.
a. We do not need a large central body of people attempting
to "manage" records. (Won't work.)
b. We do not need elaborate and complex mechanisms
imposed on the line command aimed at records management.
(Cost more than it will achieve . )
c. We do need a few simple, practical rules... (and we
need to implant these rules in the daily habits of our people
through training) ... which will guide each and every one of
us towards more desirable practices as regards records
creation and usage. (Should work.)
6. Another line of thought: In terms of Agency-level organs, it seems to
me we have too many. We have the Records Management Board, the Information
Processing Board, the R&D Board, the Contract Review Committee, etc. The
same individual often serves on more than one of these bodies.
 Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
Last year, I proposed that the IP Board be upgraded in membership to
the Assistant Deputy Director level. This was done. Part of my rationale in
so doing was to move us towards a body senior enough to establish policy and
make many decisions in the Information Processing area and, potentially, in
other areas of management concern as well. That is, I felt (and still feel)
that a Board at the ADD level could properly and effectively support the DCI
and the CIA Management Committee in the planning, policy generation, and
decision making required in running the Agency.
Such a Board would need staff support. And it might be that small
staffs specializing in such areas as Records Management, Information Processing,
R&D, and Contract Review would be needed. However, these staffs should be
kept small. Easy coordination among these staffs would also be a requirement--
a factor which might argue for organizational co-location of these staffs.
7. My basic point on Agency-level staffs is that we should avoid establishing
large central staffs with ambitious charters which inevitably run counter to the
lines of authority by which the Agency's day-to-day operations are managed.
8. One last gasp: After the Records Management issue has been thought
through a little more, I think we will want to sit down and consider what mix of
Agency-level staffs and Boards CIA really needs.
 Approved For Release 2006/09/29: CIA-RDP84-00780%,5500110011-0
Approved For Release 2006/09/29: CIA-RDP84-0078OR005500110011-0
 Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
r DD/M&S .L=~?~ ~
OJCS-1090-73
6 September 1973
MEMORANDUM FOR: Associate Deputy Director of Management
and Services
SED/OJCS
SUBJECT Thoughts Concerning Questions Raised at
the 3:30 PM "Records Management"
Meeting of 22 August 1973
REFERENCES Your oral request that we formalize
our views on the subject
A. Question:
Is there an Agency problem concerning the collection,
storage, retrieval, and dissemination of information?
Answer:
My feeling is that there has to be a problem --
rather, problems -- if only because of the combination
of two factors, among others:
1. The four-headed organizational structure of
the Agency -- referred to in the meeting as "the
Confederation of Four". It is difficult to coordi-
nate the information processing affairs of four
nearly autonomous organizations so as to avoid
duplication, and inconsistent systems.
2. Constant advances and change in information
technology. The benefits of this new technology
should be available uniformly to all elements of
the Agency and would help provide:
 -
Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
a. More efficient and timely methods of
information processing.
b. More effective collection, storage and
retrieval of information.
c. Standardization of information processing
methods without unduly sacrificing flexibility.
B. An approach:
Establish an Agency-wide organization (or extend
the charter of existing organizations) whose function
would be to provide technical assistance to Agency
groups who desire to use information processing
technology to assist in solving their problems. But
avoid the "slow-death-for-lack-of-a real-live-problem-
to-work-on" situation by giving this organization an
agency-wide mandate to establish or improve particular
information systems. This makes the organization more
a "line" organization than a "staff" organization and
provides the "organizational focus" (Climenson) to
get the job done.
This organization should emphasize the conduct of
cost/benefit studies in selecting information processing
techniques appropriate to particular problems and
situations. Change must be evolutionary and with the
full support of the group(s) affected by the change.
Pilot efforts are to be encouraged and their findings
disseminated widely.
C. "Non-approaches"
Lengthy studies culminating in equally lengthy
reports are notoriously unsuccessful in producing results
(like Presidential Commissions). Pure staff organizations
are usually weak and thresh around ineffectively.
System En4ineering Division,
Distribution:
Orig &
1
- A/DD/M&S
1
- DD/M&S Registry
1
- OJCS Registry
3
- SED/OJCS/DD/M&S
OJCS
 Approved For R let
f 9 C 4 68{ Q5500110011-0
AGENCY INFORMATION HANDLING PROBLEMS
Hypothesis
1. It is hypothesized that there is an information
processing problem within the Agency, at least as far
as the production analyst shops are concerned. By
"problem" is meant: present day technology is not
being used to a significant extent--even at the present
dollar resource level.
2. It is hypothesized further that there is
a significant difference between the approach to data
processing in support of such activities as payroll,
inventory control, trajectory analysis, mensuration,
etc., and the approach to information processing in
support of the intelligence information needs of
production analysts. The former is characterized
by relatively fixed data, computational requirements,
and reporting formats; the latter is characterized
fflu its afj, i d ?~ fiat s !
v #:azitr ~Jfy
 Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
by the high volume of narrative textual information
any piece of which will have different uses to dif-
ferent analysts, and all of which is subject to severe
semantic-ambiguity problems that must be considered
when processed by computer. (For example, the word
"strike" means one thing when coordinated with a term
that represents the concept "labor;" and it means
another when coordinated with a term that represents
the concept "aircraft" or "target.")
Why Is There A Problem?
3. There are two major reasons. First, there
is no central management of the Agency's information
handling systems. Presently, individual offices are
free to develop information handling projects that
may cost significant resources and offer relatively
narrow applications if successful, and possibly dup-
licating to some extent the work of other offices.
As example, we see CRS's MAD (Machine-Assisted
Dissemination) project developed at a cost of over
 Approved or Release 2006/09/29: CIA-RDP84-0078OR005500110011-0
gg a s a,r~,' t1 o sass g 7
y aailea"~i~ i.ri9 t+t aai? :ai > J wad;'
$200,000 and several man years' effort only to have
the decision made to replace it with another office's
development; no action had been taken to have the
MAD development made part of the big picture. Thus
too, we see the CRS computer center, after many years
in the making, closed and its operations consolidated
with OJCS. This was accomplished without any meaningful
study of the potential benefits of a central computer
facility to support intelligence information processing.
4. Secondly, the overall systems analysis
capability within the Agency is spread very thin over
the many offices. Not only are they thin, but many
are given the title without the benefits of any real
training or background; as though the title somehow
took care of that. Many computer systems analysts
exist in OJCS where naturally enough their emphasis
will be to seek EDP solutions to each problem. Presently,
production analysts do not know where to get advice
on their information handling problems.
 Approved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0
How To Improve The Situation?
5. How to improve? Create a systems analysis staff
under the DCI. This staff would review the information
handling procedures being used within the Agency and would
give special attention to the information handling
development projects that are underway. The staff would
determine the extent to which centralization of the
information handling function makes sense. For example,
perhaps all major development work should take place
within the DDM&S; perhaps each directorate should have
a systems staff that would coordinate its activities
with the DCI's staff.
I'd) It a ji;
ADoroved For Release 2006/09/29: CIA-RDP84-00780R005500110011-0