GUIDELINES FOR CIA FOREIGN INTELLIGENCE ACTIVITIES: AN OVERVIEW OF THE FOREST

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81-00261R000500030056-7
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 24
• Document Creation Date: December 9, 2016
• Document Release Date: March 17, 2000
• Sequence Number: 56
• Publication Date: May 15, 1978
• Content Type: MF

Approved For Release 2000/08/30 : CIA? kl5ftf 161 R000500
ME1ORANDUM FOR: Director of Central Intelligence
FROM . Graham Allison
SUBJECT . Guidelines for CIA Foreign Intelligence Activities:
An Overview of the Forest
Better late than never--I hope. In any case, I aut sorry that this is
so late.
This memo summarizes my response to your assignment:. to explore
whether it is possible to devise useful guidelines for the intelligence
community's activities abroad. As your last memo emphasized, this
question applies both to clandestine collection and to covert action.
The delay in my response to your assignment resulted not only from,
the time-consuming demands of learning my new job as Dean_ Even more
important has been the extraordinary difficulty of this assignment. At
first, I thought the difficulty stemmed largely from my lack of familiarity
with the community. But as I've talked to hundreds of people within and
without who don't suffer this handicap, I have concluded that the problerrr
is much deeper. It is relatively easy to take any single tree or other,
and develop the case for pruning it, or fertilizing it, or even cutting it
down. The hardest problem is to step back from the trees and get a broad
view of the forest.
My "report" attempts to present an overview of the forest and problems
of forest management. This memo extracts seven major ideas or points that
are presented in a lengthier, and more orderly form in the report.
1. Your first question is: is it possible to devise useful
guidelines for the intelligence community's ar_i ties abroad?
Many people at CIA 'and elsewhere argue: no_ For
example, Ambassador Harlan Cleveland stated this
view clearly at the Agency last year: "A written-
code of ethics can never be comprehensive enough
personal behavior as a public servant- General
prescriptions, whether in the form of dos or don'ts,
are bound to be so general as to be. useless or so
specific as to be unworkable."
My unambiguous answer to the question about guidelines
is: yes.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 ' s - tai l i ty
Apprc~~F~,,~els~t3~q(~~(~b~4-~T0~1000500030056-7
of devising useful guidelines or
simple and clear: such guidelines now exist.
Though nowhere stated as such, FrlassenbT d. inc-
precisely that form, the body of informal mores
orders, internal regulations,.
do constitute a network
governing CIA and restrictions
that provides useful direction to,
on Agency activity.
2. Why are we interested iin acode of ethics?
-- The point is not primarily or exclusively to constrain bad behavior. Contrary to the assumption
on which much current public discussion is based,
the objective of the exercise is not to tie down,
or drug a rogue elephant.
The point is to help motivate, shape, and constrain
- the behavior of professionals in the intelligence-
community in ways that will ~re toreconfidence
external 1y n lawful i ntel g e
pride_internall in a most difficult and -important
profession.
Especially in your position, you should emphasize the
positive as well as the negative side of the i4 to
the purpose
guidelines. On the positive side,
inspire intelligence professionals toecourage,
one
inventiveness, and effectiveness in ! reti
critical fundamen-ta~h.ile
of the most difficult and
assuring relevant publics
lawfulness of intelligence activity.
the purpose is to restrain behavior from violating
basic rights and other values.
3. A System of Guidelines
While any effort to write guidelines will contain a 'Fist-
of dos and don'ts, that codewill not syIn fa ct., it is
stem of guidelines
but one element in a "system" of guy
includes not only a clear statement of rules, butnalso apa rocess for,
a 1 in the rules to hard cases, a Pro ?s pian and an independent process for overseeing the community practices.
this presentation of an overview is organized around
concept of a system of guidelines:
eci.-fi:c. guidelines and
Rules (from general principles to sp
perhaps even a formal code of conduct);
A Process for Applying the Rules;
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approve F6QFW e1 {61 ' 0 : I'd'1 ` 8`1 `b 261 R0005d0030056-7
-- A Process for Overseeing Practice in Rule-'rlriting,.
Application, and Enforcement-
4. The Lar er System of Which Guidelines Are a Part
The system of guidelines on which my assignment focuses
is but one part of an even larger conglomerate of the functions
that shape, motivate, and restrain behavior of individuals in an
organization. The larger conglomerate is called by my colleagues
at the Kennedy School a "personnel management system," though it
doesn't have much to do with what ordinary personnel officers
think about. This larger system encompasses the array of activities
from recruitment, selection, socialization, and training of entrants
in the organization through assignment, career development, continuing
education, rewards and punishments, to exit--all of which shapes,
motivates, and restrains the behavior of individual members of the
organization. As part of the total "personnel management system,"
a system of guidelines is important. Standing alone, its effect
on the behavior of individuals will be quite limited. In comparison
with other elements of the total "personnel management system,"
the system of guidelines may be less important than various other
components.
-- This larger "t)ersonnel management system" is
beyond m assi nment here. As.you know from the
tsaflier memo Andy Marsha an sent I have
variou G %hat larger
system, especially as it relates to the production
of first-rate analysis.
5. General Principles
For the system of guidelines on which this assignment
focuses, the most important element is the big picture: general
principles within which more specific guidelines and processes
are established.
-- In my view, the fundamental principle is that the
President (and government) should not undertake
actions in secret that cote? nil. i- -R nr'
e defended
test of political viabilit_Y?_
-- Because clandestine activity cannot be subjected
to the normal test of open public debate, the
institutional challenge is to devise an
., ropriate s2rro4ate pro( that engages surrogates
for the interests that would participate in full
3
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
public discussion, involves them inl serioungcaing
t in Oars
disciplined private review of proposed
covert actions, and requires thoir suppcrat"ic
substantially equivalent to the normal & mo
test of public support.
The paradigm of a surrogate process in our society
- nrl` rirl+r-~ic,
is the mechanism for b vs.,the need t9
onstitutional rights to rivac
ar~quire information that may prevent crimes.
__;-
1?{ ow i s this- s"~ne. Wiretaps on US citizens can
- has issued a
be authorized only after a court cause to believe
warrant on the, basis of probably
an individual is engaged in crirninal activity_
6. Specific Guidelines Are a Can of 'lo m`s
-- Lengthy, and perhaps endless ar(JUMent can be made
on both sides of almost every specific prohibition..
Within limits, what is suitable or even permissible
will vary substantially with circumstances. For example,
"measures that should not'be and?rtakenpinmpeacetiema,
or against a democratic state, may
during actual or threatened hos'; ? ties ust?therefore
a totalitarian regime. El-e'
be preserved to adjust to ctr_'c,u', csinaand httoQe'
modify rules and procedures as o
- Because of the extraordinary ar)raY of possiblities,
it would seem best to state general pdesumptioons
that capture core va1_ues and to pr vid principles
for making ece`options to these
where exceptions are justified.
Thus, across the* array of specif
P ; and thereby affirming ic issues, rececommend
establishing presum - the value-~- but creating
a process that allows appropriate individu-alantotbalanceeeacwceptions
these values against other important valuers
in extraordinary situations.
This leads me to believe that brath as a matter of
R strategy and tactics, you should tional circumstances.
nr1IRI~"1S and permits exce Lions in exc-ep
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 -Approved For Release 2000/08/3~0 ? CI~~~,@ldRQ 0500030056-7
xc ___~~ _- ~.-.Q
7. A Final Footnote That-_.E
Current Assignment
1aosely organized
As you know very well, the present, rapid and substantial
"system of guidelines" is now undergoing
change. This change proceeds apace, piecemeal- It is
initiated sometimes by Congress, sometimes by NSC or Justice
(as in the long list of topics being addressed by the second
half of PRM-11 and its follow-on) and sometimes by the Agency
itself. But as far as I have been able to determine,,this:
erview
process of change is not informed by any comprehensive live that
of the problem, or any comprehensive strategy change
explicitly recognizes:
the u" ^f trror~hyl actic act
the tradeoffs between one change and another;
the cumulative impact of piecemeal change on the.
actual working of the entire system-
So, for example:
-- The Hughes Ryan Ame mant has produced an effect
that was not intended or anticipated by most of
its congressional supporters. Their votes were
won by the argument that the President should
.personally have to authorize and certify to Congress
tions were necessary for the
that specific covert ac
nation's security (and thus himself be clearly and
personally accountable for such actions, unlike-
the earlier Castro episode, for instance)..
But what has.been the result? This requirement
now means that covert actions are reported riot just
to an oversight committee but to eight separate
committees; these committees' lack of sanctions
for members' unauthorized disclosure givesindy?j~.1
con ressmen a virtual u over the
majority o in Congress about these matters;
this whole prosbus everel restricts the-
r aCt;~,?~
number and sco . This chain of
effects was presented in the discussion
preceded the vote on
The current-draft bill of the SenateeSelecteCommtttee
on Intelligence addresses aI ^f SnPC.ifici.ty
altogether any attempt to address
penalties for unauthorized disclosure of -intelligence
reflecting coner-er_ L V U
ate for 7 0 isl-ation and what for executive
appropriate --_ ~_ nA nmittincr
Approved ForRelease 2000/08/30 :,CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30: CIA-RDP81-00261 8000500030059
roin follow-on is
PRM-11 working g
trying to try up oe problem after another,
n
one at a time, with little overview of possible
tradeoffs and interactions.
/7/
fnr h; t,kina. My "report" is presented, therefore, not only as my
answer to your narrower question about guidelines, but also as a
first cut at an overview of the forest that might--after appropriate
more comprehensive deeper view of the problem, that CQa
take the offensive and provide at every point.a more thoughtful,
What is to be done? Somewhere, someone should have a
strategic overview of the problem and should be pulling together
the relevant parties and pushing them towards a purposive conclusion.
Candidates for "strategic director" include NSC (possibly Aaron),
the Senate Committee (maybe Miller), the Agency (perhaps STATINTL
I suspect that the first two candidates will not play this role.
If you, in close collaboration with someone in the Agency, wanted to
revision and expansion--provide some of the basis -For your carrv1ng
t~ wit h p... 1
To develop an overview that is simultaneously comprehensive
and deep; to develop' a strategic plan; to-do the strategic management--
each is a big job that will require substantial investments. The stakes
are high here--but they are high in all the issues on your desk. Not
sitting in your seat, or even being aware of many of the other issues
on your desk, I won't presume to recommend that you invest your time
and energy here rather than elsewhere.
If after you've read the memo that follows and thought about
the larger strategic problem, you want to hear more on this last
subject, I have a few thoughts about a pla coon.
After too long a delay, and more hemming and hawing than
I like to acknowledge, the ball is now back in your court.
6
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
GUIDELINES FOR CIA
FOREIGN INTELLIGENCE ACTIVITIES
The report that follows Isdoirgcauniszesedd it
will . points. After discussion and
wiserve to highlight the major p and quickly into a briefing
revision, I can translate this directly a q
paper or a long memorandum or whatever. The form will depend on how
you might want to use it.
The major ideas in this report have been summarized
zedein thercover
memo. Here they are presented in a more logical
This report is organized as follows:
I. THE SETTING
-- why should the United States engage in any clan-
destine activity that clearly violates other
nation's laws?
-- can any action the U.S. takes against non-U.S.
citizens abroad harm important U.S. values and
objectives?
II. A CODE OF ETHICS
-- the importance of guidelines
of useful guidelines
li
ty
-- the possibi
--- the role of guidelines and a code of defensive
-- should the intelligence community
about guidelines and a code of ethics? d 4
III. ELEMENTS OF A SYSTEM OF GUIDELINES
IV. RULES
-- general principles
specific guidelines
V. A PROCESS FOR APPLYING THE RULES
VI. A PROCESS FOR ENFORCING COMPLIANCE WITH THE RULES
VII. PROCESS FOR OVERSEEING PRACTICE IN RULE-WRITING
CEMENT
rPLICATION AND ENFOR
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
THE SETTING Congress,
or any broader presentation of theee nissuea to n4
F al 1 F,
th urpose
the public, or even within the Agency, A concise, i
preamble that puts the problem in context.
hard to do well. People come to this ssue with
preamble is very prejudices:
such varied backgrounds andsrejuld engage in any
some wonder why the U.S. should
clandestine action at all;
others are skeptical about any restrictions
ever on U.S. clandestine actions abroa ;
what
still others are confused: offended by Soviet
citizens' phone
electronic eavesdropping on U.S. but.
to U.S. Congressmen, i cal 1 s or Korean payments
uncomfortable about the basis facticesrin~~theg
if CIA engages in equivalent
Soviet Union and Korea.
'For frame mast iofnterereferent arties,.
sted
reamblE. should establish common
A good p specific i ss
problem and some
ddress r who the
within which to a of
are thinking a
it can serveted coclus s the people coe of w a thse a pro general int boutroductithese
on
i o he made
ssueyor the fi it should s Dints t
issues f for the fir time, problem. Among the key p
to major dimensions of the are the following- daily in clandestine
A. The U.S. government engages that violates the
activity abroad that is not acknowledged, and that would, if
laws of nations against whom it is taken,
stitute a violation of
taken against U.S. citizens, con law. Some clandestine activity
American's rights and U.S.
satellites taking pictures of missile sil~rom wheat
involves s lucking
some involves electronic devices some is just good
fields; documents
air phone calls or other communicat1olv~s the U.S.
espionage where a spy gives
old-fashioned esp government's secrets or a foreign government
containing a foreign in
asses false documents offirgovernment about U-S.
in the hop nmen
p e of deceiving
Lions or actions. questions:
This activity poses two fundamental l) Why should the U.S. engage in any clandestine
( activity?
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Ap roved For Release 2000/08/30: CIA-RDP81-002611`000500030056-7
ci tize ns harm on important U.S. takes against non--
(2) Can any ortant U.S. values and
objectives?
cl anciesti ne
ngei n `~
States e
Why shoul d the_Un United Because
aws.
B. or and
majcause
activity that clear viol overnments intsecret Can do
actions taken by forei national sec
sometimes err ara h rm c l o s e d .S societies like the Soviet Union,
ll
ot
y
some nations, especia
ts to hide from outsdc
vle"ncedb
advay
f
or
e ef
a
massivs
ecause U.S. interests
engage
have-so
U.. . interests.
S
information that othWesha?e not
a
collectedsit Because
acquiring use U..S-
other
,na
long as they do not know events in
n 1 ?pnc =
wry and without
interests can s
countries by instruments b other vt e lotions has
our role. The recent orgy of rev
But it has provided
focused acknowledging rather than successes.
sed on failures, that make the point:
a number of examples histirated
h
t
without the development of
agiilitiesr the ' and reliable clandsSinand aSoviet strategic arms
Treaties limiting negotiated. These treaties
tain
could never have been commit the Soviet UnionotototbuildeanaABe M certain
actions, for example, and the
the closed nature of Soviet society
ite
Giv's un willingness to allow on-s ets
S,ovi
Sovienet
S. could not have confidence that
t fore what
?excep
th U S
t
-
eemen
were adhering to the agr means 0
the treaty refers to as anatioae collectioneri-
fication." Effective c1 rerecuzsite
ential capabilities are an evensioreofscurrent SALT ;
of any successful
negotiations.
ry
hijacked an Air
when Black September teas
France 707 with 100 Israelis aboard, took the plane
to Uganda's Entebbe Airport, and threatened to kill
s an equivalent number of Arab
ard, toexisted?
the hostages unless released, what
prisoners in Israeli Jails were,
frhe
"legal" surveillance of the situation..
observed local Ugandian law were not itselfya Capability
the hostages. Should the U.S . e
for covert action like the Israeli raid on Entebbe
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30: CJA-RpQP. '~-SQ264R 06 030056-7
if terrorists were diverting ri 3t of a large
any ts
uranium from a nuclear reactor in
number of countries, U.S. security and interestS
could require a covert capability to identify fact and, if necessary, to take action. the
C. Can any action the U.S. takes auainst non-U__5.. citizens
harm m~ ctant S val aes and aa~ect?ves? This Is the
th abroad - o
o ntelli-
other extreme. I have been surprised t;o discover how many
people are candidly skeptical of any restriction on U.S- i
gence agency's covert actions abroad--beyond thetradit1pna~
question has identified at least
calculation of w"?~rQr the V
an My refection on this
st
three ways in which U.S. intelligence action again non-U.S.
izens abroad can harm important American values and objectives:
tit under-
estimated The power of the U.S. exarp e is greatly by
most Americans- George Washington s
contention that teabroadravexample erstat~essther
most powerful instrument
point. But the opposite view is even less tenable.
As the most open--loci ety__i n _the wart d, the U.S. i s
most vulnerable to int~erro~~S~ltechr~alagiPSplike
terrorist activities,
plastique, and even assassination. We~kcasaitiis,
one of our strongest defenses against
is to be found in international legal andmoral
than
taboos. Our nations role in weakee ingaraaser than-
strengthening such restraint on po
tion will, I believe, stand as one of the blackest
marks on our record of the late 1950s and early 1960s-
Foreign intelligence activities can contravene and
even u der
foreign policy objectives are a multi-faceted and
often not entirely compatible amalgam- Because the
instruments of American foreign policy include many
lanizations, it is not possible to achieve?
a f fi nely lyg-tuned consistency. But where an agency'
a ,
engaging in covert actions is giventowide tee diss'cretjoretionn
the likelihood of actions contrary The
thrust of American foreign policy
most celebrated recent example--Chile--_1nAfact
ions
illustrates just the opposite. There=
were consistent with direct Presidential order.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30: PC.Iy~tDP8e -6J.Q05Q0030056-7
- While the U.S. Bill of i s do
the U.S. government has always
non-U..S, citizens, i President
asserted the concept of iman r-iah:5 in our dent gn
Carter has made this a higher priority
policy and penalized governments that regularly
and systematically violate their own citizens human
ctivities, were to
a
of those same -tnalvl ua ,
only be inconsistent, our actions would belie the
very values that the President's policy proclaims-
Moreover, revelations that the U.S. has engaged in
of itical assassi--
rights. If U.S. inteli~gence
stematic violation of the ri hts
engage in regular, s - co ur os t 611WUUF not
certain categories of act'on, e-g-,
nations, raise deep questions in many Americans
mind about neir owngov rnment. - s t,. C
A CODE OF ETHICS
Productive discussion of a "codoff etafcunklrequi ?~pre-
similar preamble that clears away a number
and (2) the likelihood t a
A interests. i _,i
the widespread view at the Agency and elsewhere
that it is not possible to devise useful guide--
lines for intelligence activities abroad;
the view--prevalent in Congress and the intelli-
gence community--that the overriding purpose of
guidelines and a code of ethics isato tie down
a rogue elephant so as to prevent
the suspicion--in the Agency and elsewhere--that
anyone who advocates a code of c ethics mustnbeaaout
naive moralist with unreal s
the effect of such codes on the actual behavior of
members of an organization.
Among the points to be made here are:
A. The Im ortance of Guideline ai .ndoa `once
of guidelines for i ntel l i gence
home emerges from the juxtaposition torgdarante facts
1
(1) the necessity for cla oante objectives;
national security and advance important foreign p
h t clandestine activity will violate
conceptions:
Approved or R ease 2000/08/30 : CIA-RDP81-00261 R000500030056-7
an
important U.S. values
 Approved Fo1rhRej? , O%W3A: I pF_.S1-QQ 6 OO08OOO30056-7
that can be served by clandestine activity and the. important
values that can be violated by clandestine activity must be
addressed by a system of guidelines that determines how competing
values are to be weighed in specific cases.
The purpose of a code of ethics (or system of guide-
lines) is to motivate, shape, and constrain the behavior of
professionals in the intelligence community in ways that will
restore public confidence in lawful intelligence and give intel-
ligence officers ride in their fir , ,f,.Ssion? More specifically,.
the purpose of a code of ethics is twofold:
positively, to inspire intelligence professionals
to courage, inventiveness, and effectiveness in
performing one of the most important and critical
functions of American government;
negatively, to restrain behavior from infringing
basic rights and other important values.
The positive and the negative are two sides of the same coin.
B. The Possibility of Useful Guidelines. In talking
to people at CIA and elsewhere, I ve been surprised to dis-
cover how many people believe that it is not possible to
devise useful guidelines for CIA. Some base this conclusion
` on little more than the conviction t But a
"~'
nclusio
hi
n
s co
number of more thoughtful individuals come to t
after hard thought about the extraordinary diversity of
circumstance in which clandestine action may be taken, and
the inherent ethical ambiguity of activity of this sort..
This view was stated pointedly by Ambassador Harlan Cleveland
in a speech at the Agency last year when he said: "A written
code of ethics can never be comprehensive enough or subtle
enough to be a satisfactory guide to personal behavior as a
public servant. General prescriptions, whether in the -Form
of dos or don'ts, are bound to be so general as to be useless
or so specific as to be unworkable."
Having walked around this problem more times than
I like to admit, I can give you an ambiguous answer to at
least one question. Ambassador Cleveland and other who argue
that it is not possible to devise useful guidelines for the
intelligence community's activities abroad are incorrect.
Those who answer "no" to the central question have been misled
by too narrow a conception of the problem.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 My grounds for confidence about the possibility of
App oiwdfor lllea ec2DOO1881 :0(CIAaRDp8 ?026$R90050? "056-7
xis.. Though nowhere stated as such, or
assembled in precisely that form, the body of law, executive
orders, internal regulations, and informal mores governing
CIA do constitute a network of guidelines. These guidelines
[Tlay floc hP s7~.lr;ar at4n.L noun to he. Nonetheless, in the
day-to-day operation of the Agency, these guidelines provide
important, useful direction to, and restrictions on Agency
activity. For example, why does the Agency not target U.S.
citizens for clandestine collection within the U.S.. (or when
it does so, stand in clear violation)? Because the legisla-
tive charter embodied in the National Security Act of 1947,
as amended, states "the Agency shall have no police, supoena,
law enforcement powers, or internal security functions." Why
do station chiefs not use missionaries in foreign countries
for cover? Because a DDO log notice declares missionaries
off limits. Why do case officers not enter into contractual
relations with working members of the media? Because you and
your predecessor issued internal regulations prohibiting this.
Why are case officers so protective of their agents? Because
of their conception of their professional obligation and
,personal relationship of trust with another individual whose
life they have compromised.
C. The Role of Guidelines and a Code of Ethics. Any
effort to write guidelines will contain a list of dos and
don'ts. But that list of dos and don'ts, whether embodied
in internal regulations like DDO log notices, or perhaps even a
formal code of conduct for intelligence professionals, will not
stand alone. In fact, it will be but one element in a " ystem of
guidelines." This system of guidelines includes not only a clear
statement of rules, but also a process for appl~inig the rules to
hard cases, a process for enforcing compliance, and an independent
process for oversees the community's practices.
This concept of a system of guidelines serves as
the organizing principle for the discussion:
~- Rules (from general principles to specific
guidelines, and perhaps even a formal code
of conduct);
-- A Process for Applying the Rules;
-- A Process for Enforcing Compliance with the
Rules;
-- A Process for Overseeing Practice in Rule-
writing, Application, and Enforcement.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 nt to assemble
wa
may
is ackage, one enr to try
1 ~{ } f 14-1~81sqt500030056-7
~
~~~~
h ~i
roue e of conduct- uc
a worki nor ? orma l CO d y - ~rofes-
gF
nCe i
to formulate a formal Central values for i nt~ c
hard cases
articulate a half dozen
al s and then provide some i 1-at i ve ex~r~i1pl ld es be ofweighed and
shou
tem
sion competing values
with suggestions about how
traded off in these specific instances- an
trad executive orders;
ethical will include ray no assassination;
notices; etc.--this system of rules
ethical nregUl -laws that
log
internal regulations;
cannot be specified to the level of detail ofthe intelligence
n ui rules to
Because situatisexercisecdriscretiondin app lying
between
professional mustst frog to
hes. The virtue of the rules is toid.ti ance in attemp on hard
ard cas hard cases and to provide some g
easy and considerations that bear
weigh the multiple and competing
cases.
ntelligence Community Be Defensive Abou
Should the I ~ The ethos of the community that
D.
Guidelines an`d a Code of Ethics_ I believe, however,
pro f i l e and be defensive. that
to take a low
a strong case can be made for talon the guilty
As the accused--and-an agency evidently g
ious abuses--the Agency is the target
of some ser
especially in Congress.
of many reformers,
Refensiveness u has
guidelines the Agency
In fact, the system tiaver the last
been developing, partcularly
ears--when tidied up and though
several tfully
Y
presented--can be shown to be `~ofoguidance
in of guidance
thoughtful, and effective a system
as there is for any major agency If the Agency
including especially the Congress.
cressive in ex lorin
took the offensive and was a the case or t e exsen..
the t. 5' in making lawful
tial role of intelligence, specifically
intelligence in a free society, and in arguing the
stem of guidelines will
case that the emerging sy it might both restore
guarantee lawful intelligence, i
eternallY and the
the standing of the community
morale of its members internally- hors under the
the intelligence community la Indeed,
Today
ow of revelations rterabacksyawe cann now mseer in
as Monday-morning qua mmunit s u
clearly that the intelli through the 19 s'...
an o en democratic societ was
anomalous. You have moved vigorousiy to
f essent,a
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved Fors
6-7
pos (as
e? ial guidelines forte
an example for all major agencies of government
should at least be explored.
III. ELEMENTS OF A SYSTEM OF GUIDELINES
The organizing principle for this discussion is a "system of
guidelines." The key elements of this system are:
A. Rules
-- General Principles
-- Specific Guidelines
B. A Process for Applying the Rules
C. A Process for Enf orcing Compliance with the Rules
D. A Process for Overseeing Practice in Rule-Writing,
Application, and -Enforcement
RULES
A. General Principles.
Rules for the intelligence community consist of both
general-principles and specific guidelines. Most important
are the general principles that establish the context fororespecific argument. While it is possible to multiply principles,
I have tried to formulate the minimum number that could serve
to establish a working context.
*Recall this system of guidelines is one part of a much larger
that includes recruitment, selection
socialization, and training through assignment, career development,
continuing education, rewards and punishments, to exit. This entire
system shapes, motivates, and restrains the behavior of members of
the organization.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 1. Es iona e is an extraordinary~remedy: because
Approved ctra ' %-4 7 '-p 'R0QMQO&A056-7
values and interests, clandestine action should be under-
taken only as an extraordinary remedy, and where the benefits
of the clandestine action, and the advantage of pursuing the
objective through clandestine means rather than overt means,
have been clearly identified.
advocating clandestine action.
ti -. It puts the burden of goof on those
c c V ' !This principle establishes a
(Contrary to the impression created by recent
revelations, clandestine activity has mainly been considered
and authorized as an extraordinary act. The most dramatic
indicator is budget. Even Marchetti and Marks, whose thesis,
is that American intelligence is dominated by covert action,
estimate the budget for covert 'action in '75 to be only $750
million. In discussing this subject, I've found that many
seeming lIinformed New York Times readers, and even some
writers, believe that the covert action budget must be many
billions.)
2. The most fundamental principle is that the President
(and government) should not undertake actions in secret that
could not in principle be defended to the American -1iubl i c
and meet the test of political viability. There is a view
that the American people are not sophisticated enough to
appreciate the need for covert action, don't understand the
exigencies of state, and are naive about the dangers inherent
in the world. While there is considerable evidence to
support this view, the U.S. government is constructed on a
quite different presumption.. That presumption is that, on
balance, and- v r time, a process that forces the President
and the government to seek and win the support of the
American people to sustain a chosen course of action is
preferable to any other process yet invented for selecting
and sustaining public policy. The record is mixed.. But on
balance, and overall, this democratic presumption has a better
batting average than any competitor. (Recall Churchill's
remark: democracy is the worst form of government- except
for all the others.)*
*NOTE: Political viability is not measured by a direct poll of
public opinion. In our system of government, political
viability is determined by an amalgam of views of the
President, Congress, courts, and citizens.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 What does this principle imply for clandestine
activity, that is covert a `io
Approved FoflReLeasav2*/p,$/;9re #A6## 0q6 0056-7
einq
known that we are the influencing agent) and clande
stine
collection (the effort to collect information that others
do not want us to have and without anyone knowing that
we have collected it)? In the abstract, the implication
is clear: the U.S. government should not undertake actions
t hat ul d not in on nci le win the s~rn.~ tk_~p,a
l
peop
e ( i f 1 . V, r no~eibl p n haves f,rl 1frank nu
.,.,. _ z__-v jzi ----,-_ bb l i s
di C('IICCinn Th
e
activity must by definition be secret Janducannot utherefor?e
be subjected to the full, open processes of public debate to
determine whether it meets the test of political viability.
This dilemma poses the institutional challenge: to devise
principle 3.
3. Because clandestine activity cannot be subjected
to the normal test of open public debate, it mus be
authorized and overseen by an app-p-riate surrogate rocess.
_
a process- that engages surrogates for the interests that
would participate in full public discussion; involves
them in a serious, disciplined private review.of proposed
and ongoing covert actions; and requires their support in
ways substantially equivalent. to the normal democratic test
of political viability. No easy task.
The present process for authorizing clandestine
activity, for overseeing the process, and for checking
abuses represents at least one attempt to meet this prin-
ciple. The question is how well current procedures meet
this test and how they can be improved.
Two slightly more specific implications of this
principle are:
V7
This is clearly controversial; and contrary to the
practice of the past. According to this principle,
the U.S. is going to provide clandestine support,
(1) that_clandestine activity must be consistent
with openly announced substantive olic-les vn-"o U etc_
1.1:?s tnalhave Been established by the norma en
process of government; and
(2) that the President and Director of CIA
should be prepared to defend in public the broad
categories of clandestine activity in whictT the U.S.
en a es, though not the specific actions themselves..
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved F&rRdt#agea2bDCV/0 /39s: C1A~ r8~'P- drl 'tion parties are being funde y aer na general class of action should be defended in public
debate. This does not mean, however, that the U.S.
government's action in funding any particular party
should at the time have to meet the test of public
debate (since it will not be possible to make public
all the information on the basis of which a full and
open debate could take place).
It has sometimes been proposed that we add a
fourth principle of publicity,sometimes stated as follows:
if a particular clandestine activity were made public, would
you be proud of your action? I find this proposed principle
h
t.
unsatisfactory. It is both too loose and too tig
STATINTL On the one hand, the principle is too loose: individuals'
notions of pride differ. would be proud to
have made public many actions that most Americans find
offensive. On the other hand, the principle is too tight
and restrictive in that clandestine action-is by its nature
secret and unacknowledged. The grounds for the decision to
engage in clandestine activity in a particular case, and
the circumstances that surround that decision, encompass
many facts and factors that cannot be made public. Conse-
quently, it is not possible to present the full case in -
public for any particular action. Absent that full case,
the public does not have full grounds for judgment. Con-
sequently, the public test of support and "pride" is not
appropriate. That's the point addressed by the surrogate
process above.
B. specific Guidelines
Each class of clandestine activity presents a juicy
target for endless juridicc.l ai-gumeat--both pro and con. You
will recall the lengthy debate about the Harvard CIA guideline
requiring professors to inform their dean about any paid work
they do for CIA.
As an academic, I find it difficult to resist joining
such arguments, for instance, about whether the intelligence
community should be permitted to try to overthrow democratically-
elected governments, and if so, under what circumstances.
I have lengthy notes on both sides of proposed guidelines about
a half dozen major classes of clandestine activity.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 For the purpose of a broad view of the forest however
App rTFOIr Rele"eo4V(DGi08t3Ol:iCiA-FQnI -W2tI &&Q04056-7
major i sues to be weighed in choosing any specific guidelines?
and to offer some general criteria that should be applied in
choosing specific guidelines. The major issues to be weighed
devising specific guidelines are four:
(1) the pros and cons of s atute y. executive
order v. internal regulation;
(2) the relative merits of fl_ at prohib__;~s
v. prohibitions subject to wa? ver_
(3) the advantages and disadvantages of engaging
m.Qr_external surronates (courts and Congress) at
successive levels of specificity; and
(4) the imp of each guideline in con-
straining abuse, encouraging undue timidity, and
motivating professionals in the intelligence community
to the desired mix of initiative, inventiveness,
and restraint.
Each of these issues invites lengthy argument. None
is easy to weigh in specific cases. But for the sake of brevity,
I will propose three general criteria for choosing specific
guidelines--criteria that should, I believe, be widely acceptable..
1. The necessity for flexibility. The appro-
priate course of action will vary substantially with cir-
cumstances. For example, measures that should not be
undertaken in peacetime or against a democratic state,
should be permitted during actual or threatened hostilities
or against a totalitarian regime. Circumstances change.
Guidelines must therefore preserve flexibility to adjust
to circumstances and to modify rules and procedures as
conditions change. (This criterion has strong implications
for the balance between legislative statute and executive
order subject to Congressional veto.)
2. Guidelines should express
f~_ xs. in wags
that establish strong presumptions al ainst vi ofatkon of
those values. Clandestine activity runs a constant risk
of violating important American values. Guidelines should
therefore state these values as clearly as feasible in
order to affirm the values and to place the burden of
proof on whoever proposes to risk harming the value.
Approved For. Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 3. Except u-~~ ~~~~u , . the yuVr_X it--
that values com
circumstances. Recognizingcess that allows appropriate
Appr pete,
?avtl F BI@A ' 16 RUP -0026+1AO1~~~ 3~~c2
individuals to balance e
thus make exceptions to the guidelines in extraordinary
situations.
The specific guidelines proposed in the Senate Select
Committee's charter legislation include flat legislative prohi-
bitions on:
assassination of foreign officials;
special activity that has as its objective or is
likely to result in:
the support of international terrorist
activities
the mass destruction of property
the creation of food or water shortaces
or floods
-- the creation of epidemics or d: series
1 or other
the use of chemi cam Dl~ -
weapons in vi o of treaties
verth
the violent
country
the torte of individuals
the support of any action, which violates
conducted by the police,
foreign intelligence, or internal security
forces of any foreign country;
the use for certain intelligence activities of
onsor
persons who follow a full-time epansorship and
who travel to a foreign country under
part of a U.S-
su rornote education or
government program designed to p
ural.._aLIai rs ;
the arts
organi zati on-
anynU.Sacmedi vities
the use for intedttoelli
jt~rnal ls~s, cce
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-T
 In my view, each of these prohibitions does re i ect. sou'
-however ill expressed. These prohibi-
tions also appro fa ~-'Y'/M @ tCIArRQPR11 2611 OO~06 3oO56-7
writing specific implementing regulations.
The chief probl m with these guidelines is their fora.
If stated as presum_ iois rather than flat prohibitions,
accompanied by an apprupr1Q4; V,
a reformulated (and, ohopes, more felicitious formulation
of the values) should be acceptable.
Though the draft charter does not state prohibitions
or presumptions relating to col1ecti you m hgat want :a_yQse
l e -
m
-_ l e..s i ntrusi ve means of col l ecti o ar~nto'=-P``?
reed to more intrusiVP means (thus gathering ir-forma-
tion rom unwitting individuals is preferred to
unobtrusive electronic surveillance whi h is, preferred If -
to breaki g and e teri g
e1evant for internal Agency
r
A final point may e m re
purposes. If one takes each class of clandestine activity, for
example, "propoganda" or "economic warfare," discovers whclabarate and specific guidelines now exist, one quickly
network of guidelines from e
to DDO Log Notices. For
could array these giwethncarefulyattendtianetaE~thexma~artissues
reviewed and refined
identified above.
V. A PROCESS FOR APPLYING THE RULES
The process for authorizing special activities,, special
collections and counterintelligence is spelled out in Execut:'r.ve,
prouided_oE the sur-
stagef cothat uld beprocess
Order 12036. Abrief
eachsketch
rogates involved
VI. A PROCESS FOR ENFORCING COMPLIANCE WITH THE Rill-ES
The Inspector General, the General Counsel, and the, 1QEL
VII. A PROCESS FOR OVERSEEING PRACTICE IN RULE-WRITING APPLICATION
AND ENFORCEMENT
The Senate and House Intelligence Oversight Committees,
especially if they would get their acts together.
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
p
several presumptions. f'UJ exa
 Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
Sections V, VI, and VII can be filled out whenever You like..
16
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
ADDITIOtJS To ALLISON CONCEPT
Section on public as an oversight process
policy of openness
limitations on sharing with public
Acknowledgement of
Limits on FOIA
Sanctions against disclosure
III
Section on rewards
ecial role and sacrifice of
Acknowledgement of special
intelligence professionals
System of incentives
Medal s
Retraining program for retirement
Retirement program
Section on domestic activities
FBI-CIA jurisdiction
Special concerns for rights of Americans
Relations with American institutions
Press
Peace Corps
Cover
Electronic surveillance - domestic
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7
 Approved For Release 2000/08/30 : CII-RDP81-00261 R000500030056-7
IV Section on liaision
Limited to intelligence, not security
Need to respect privacy
22 May 1978
STATINTL cc:
Approved For Release 2000/08/30 : CIA-RDP81-00261 R000500030056-7