STANDARDIZATION OF TERMS OF REFERENCE

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP88G01116R001202280002-5
• Release Decision: RIFPUB
• Original Classification: S
• Document Page Count: 3
• Document Creation Date: December 22, 2016
• Document Release Date: March 3, 2011
• Sequence Number: 2
• Publication Date: October 6, 1986
• Content Type: LETTER

Approved For Release 2011/03/03: CIA-RDP88GO1 1 16RO01 202280002-5
6 October 1986
NOTE FOR: All NIOs
NIC/AG
FROM: C/NIC
VC/NIC
SUBJECT: Standardization of Terms of Reference
1. We need to further standardize the drafting of Terms of Reference
(TOR), particularly to reflect recent emphases in thinking that have been
raised by the DCI and others and that we have discussed at staff meetings
over the last several months. Future TOR's should therefore follow the
following outline:
a. Indicate who asked for the Estimate, for what purpose and what the
rough timeframe is for preparation.
b. Critique the last Estimate on the same or related topic. This
should be about 1-2 pages in length, representing a thoughtful and
non-defensive view of the community's last effort on the subject.
How accurate was it? How has it diverged from actual subsequent
events? Any lessons to be learned? Did we overestimate or
underestimate the problem? Did we miss the main issue?. If
footnotes were taken or alternative judgments were made, who seems
right in retrospect?
c. State the problem or issue that will be dealt with in the Estimate
-- along lines that our previous Concept Papers have taken. This
paragraph outlines the issue and its relevancy to policy problems.
d. Key Questions. These should usually be no more than 5 and can be
less. They represent the heart of the Estimate. The Key Judgments
are answers to the Key Questions. Draft Estimates will be checked
against the Key Questions. The questions should be as broadly
conceived as possible.
e. The outline of the paper. The main sections of the outline should
spring from the Key Questions. Sub-topics should be posed as
sub-questions -- which are preferable to simply statements of
sub-topics. The reader should be able to look at the outline and
get a sense of the range of issues that will be touched upon under
each major question.
DCI
EXEC
REG
SECRET /9-.200
Approved For Release 2011/03/03: CIA-RDP88GO1 1 16RO01 202280002-5
 Approved For Release 2011/03/03: CIA-RDP88GO1 1 16RO01 202280002-5
SECRET
f. The TOR should state which other NIOs have been and will be
regularly involved in the conceptualization of the paper from the
very outset.
g. Before the TOR is drafted it should be discussed with VC/NIC and/or
C/NIC, along with other pertinent NIOs, to insure that we are on
the same wavelength and that broad papers are appropriately
relevant and meaningful to the work of other NIOs in related
areas. Some NIOs will fairly regularly be involved as
participants: NI0/ECON, possibly NIO/USSR, NI0/GPF, or others as
appropriate.
h. An in-house attachment to the TOR should mention, at least
generally, what outside consultants may be used in helping
formulate the paper. ese consu ants may be paid or unpaid,
cleared or uncleared, formal or informal. The most effective use
of external consultants comes before the drafting of the TOR and
certainly before coordination. There are many uncleared experts
who are willing to offer their thoughts on how they see a given
country or problem and who need never see any Agency paperwork.
They can help focus the NIOs thinking on how to formulate the
problems and consider various outcomes.
i. The paper should list the key policy level players whom the NIO
expects to consult to insure that the paper touches on issues
important to them. This does not mean the Estimate should only
answer their questions. The NIO himself determines the full range
of the issue -- but this range must include specific policy
questions. It pays to clarify these policy questions with the
policymaker before drafting the paper or even the TOR.
j. All Estimates should generically include the following:
The broader regional/functional context of the variables and
judgments focused upon in the paper.
The major variables that affect the Key Judgments, the effects
of those variables, and what kind of alternative outcomes they
could produce along a spectrum of probability.
Elaboration on the range of alternative scenarios, i.e., most
likely case, and alternative cases, to include plausible best
and worst cases, for example.
A brief statement of indicators which would help the reader to
track any subsequent drift toward better or worst, or more or
less likely scenarios along the spectrum.
Approved For Release 2011/03/03: CIA-RDP88GO1 1 16RO01 202280002-5
 Approved For Release 2011/03/03: CIA-RDP88GO1 1 16RO01 202280002-5
SECRET
2. This format should serve to insure that we are in fact doing those
things for which we have been asked to consider regularly by the DCI, the
DDCI, the SRP, MAP, and SSCI, as well as reflecting our own internal
discussions on ways of improving our own product.
3. Front loading is everything. If we haven't done it right before the
TOR is coordinated, subsequent mouth-to-mouth resuscitation by the NT0 -
doesn't really salvage a dying Estimate.
11
B en Frank . Horton III
Graham E. Fuller
cc: _,BCI
DOC I
SRP
SECRET
Approved For Release 2011/03/03: CIA-RDP88GO1 1 16RO01 202280002-5