Before the Subcoemtl ttee on International Security
on Fonlgn Affair*
ar grateful for tha opportunity to appear beforeCommittee
and to aislst In It* Inquiryubject of treat nationalprovision of Intelligence requiredhr policy processgreat difficulty and its complexities are not well understood. the Conmlttee's findings trill help us to do It betterhe future.
Chairman of the national Intelligencea responsible
the production of Rational Intelligence Estimates and other fores of
intelligenceupport of the policy process at the Rational Security Council Level. Hy remarks today kill be primarily focused on the Estimates
theaiporunt fonaal product of the Intelligence Coemwnlty. X
erill address first the nature of Estimates; second, the process by which
are produced; and third, how me are seeking to Improve thorn.
j jEstimates are produced by the DCI. They may be prepared
;on requestenior policy officer who Is concerned about the trend of eventsarticular situation; an example Mouldaper em
Soviet objectivesis Afghanistan and Iran. Some are
on an annual basis and are tied directlyeriodic policy
process such as that of the budget; an oaaaple here mould be our annual
on Soviet strategic programs. Lastly, others are done because
Maur staff caneed for bettor understandingroblem
that trill probably call far policy constat rationanlor levelha next ft. years; some of our work en energy problems falli In this araa.
One essential featureotional Estimatehat It'draws on all of tht Information and wit don available to tht US eoverramwit, wthtthtr fror official sources or private ones, on tht subject at Issue. acond ftatura Is that it goes bayond tht descriptionituation to project tne course of events Into tht future. It dots not predict, butange of posslblt futures and attempts to assess their relative probability. Obviously, the shorter the tin* span wa project the grtater confidence we ctn glvt to our assessacnts. hirdery laportant ones that our goal In producing Estlaatasot to achieve unanlalaity. On many questions of treat national Impor-tance thereide range of views. The national Estimate oust not suppress dissenting views or water down the differences between then. Rathersust identify those differences that are substantive andand present them so that the policy officer can understand not only what these views are but why thay arainal feature Is that an Estimate, however great the human effort and however extensive tha amss of data that have gone Into it, oust be clearly presented,written, and above all short enough so that tha requester bat time to read 1t.
Turning to the production process, tho key figure hare Is the motional Intelligence Offlcar. The motional Intelligence Council consists of eight regular NIOsall with specific faooraphlc responsibilities and two responsible respectively for strategic farces and general purpose
forces. In addition there Mill In tine be RIOs-at-Lerge whosewill discuss
Each NIOhe DCl's senior aUff officer for his area. As such, be is responsible for assisting then Ms participationhe Rational Security Council, the Policy Review Committee, and the Special Coordinating Committee. Helso responsible for maintaining close liaison with senior staff of the RSC and the Departments of State and Defense In order to Identify the questions arising In the policy process on which Intelligence can assist. Equally Important, hemportant for seeing that Intelligence production Is In fact responsive. In particular, he oversees the production of Rational Intelligence Estlootes within his field.
Once the need for an Estimateecognized and the DCI has approved Its production, the NIO convenes representatives of the Intelligence Ccemiunlty agencies to prepare terms of reference. He also Invites policy officers and, wherever possible, the requester. The purpose of thiso ensure that the terms of reference are In fact responsive to the requirement. hould also not* that while we very such want the policy officer to help us define the problem, we believe that that should be Ms last contact with the process until the paper Is completed. To preserve the Independence and objectivity of our response,must maintain this separation.
Tho RIO then arranges for the Estimate to be drafted. He has the entire Intelligence Coaaamlty to draw on, and Ittandard practice for him to go to the best man for the job that he can find, regardless
of agency. In the case of elaborate papers, especially those on military subjects, the NIO may form several Interagency working groups to prepare parts of the paper. In any case, the drafters work under the MO's supervision rather than the supervision of their parent agency. The agencies do not have to take responsibility for the draft; theyhance to make their views knownater stage In the process.
It is normal practice for the NIO toanel of outside consultants to review the paper at various stages, of which the most Important Is at the point when an Initial draft has been prepared.
Once the NIO and his superiors are satisfied with the draft and have taken Into account the contnents of consultants, the draft 1sto the various agendas for Interagency coordination.oted earlier, the purpose of the coordination process Is not toland homogeneous paperhich all differences have been carefully smoothed over or buried. There will be differences that will be resolved In the course of argument. There will be others In which two sides are operating from different Information. There will be still others of the half-full or half-empty variety. Is It significant whether something is sharply reduced or severely reduced? Disagreements like these can be resolved. But there will also be real differences of interpretation, of substance, of Judoment. The NIO must lead the Coawunlty through the processay that eliminates the trivial differences and Illuminates the important ones. To do this he must ensure that these alternate Interpretations are presented In parallel language, with the supporting arguments arrayed so that the reader will understand what the fight 1s
about and whyttars to hla.hould aaphaslze that wt ara less concerned with who holds uhet viewh what tho view 1s; thli mans that differences aaoog tha analystsing It agency arc just as important at dlffarances between agencies.
Tha final staga In the production of an tstlaetetsto the National Foreign Intelligence Board. Th*ssuesMlth the advice of the Board. Each principal of the Board hai the tight to dissent from the findings of the pape-. In practice, this usually aeans that the Boardaper that Its representatives have already worked out. Including any expressions of difference that Itcontain. Nevertheless, the principals retain the rightand often exercise Itto have their own dissenting views expressed in their own My,
I have described the process asormally proceeds, but we retain the flexibility to prepare Estimates very rapidly by short-cutting these steps. While the najor Soviet strategic Estlaate usually takes six months or sore to prepare, the shorter special Estimates directly linked to current policy problems can be turned outew days. This probleai of speed Is oneill return to later.
ant to discuss sooet of the changes that we have recently wade In the National Estimates awchlnery. To doit of history Is necessary. Since thes responsibility for the production of Estimates lay with the Board of National Estimates aad withsupporting Eitleates Staff. In slaplest term, the staff was responsible for drafting Estimates and the Board, actingollegia!
body, rt.iced then and advised then their Issuance. r. Colby, fooling that the Board had bacon* too detached from the policy process and that the staff had become too detached from the analysts of the Coamunlty. eliminated both and established In their stead tho NIOs.ave described them earlier. Because each NIO mas also the Director's personal representativeheocess, he could see tohat therelose linkage between that process and Intelligence production. Because he had to depend for his drafting assistance on the Coanunlty agencies, they were brought back Into the Estimates process.
Speaking very generally, the NIO system was moreuccess In provldirg Intelligence support to the policymaker thanas InNational Estimates as such. NIOs. especially those on areas like the Kiddle East, are extraordinarily busy and events In their areas arc fast-moving. This has meant that the Estimates machinery has often not been able to move fast enough to produce the papers that are required before they are overtaken by events. NIOs have Instead found less formal ways of making Intelligence Inputs. Moreover, there hasrend sinces for Intelligence to be woven Into policy papers rather than presented separate. Thus the number of National Estimates produced, outside of the military field, has fallen off,ould say that the Input to the policymaker of estimative material (with aV) has Improved sharply under the NIOs.
Another problem with tha NIO system has been that of quality control over Estimates. Whereas the old Board of Rational Estimates was
collegium In which papers were subjected to critical reviewroup Of experienced generalises, the HIOs have operated as Individuals an- werable only to the DCI. Not only arc they busy, but the OCI Is of course busier still. There has never been time enough to give papers the kind of review for relevance, organization, coverage, and just plain wisdom that they deserve. Neither the process of Interagencynor the use of consultantsny substitute for this kind of review.
Uhen we formed the National Intelligence Council on Januaryur Intent was to preserve the best features of the MO system while seeking to recapture those of Its predecessor Board and Staff. hould caution that we are onlyhe early stages of this venture and Itar too early to judge 1t. The reoular NIOs, their number reduced to the present eight as described earlier, have become the Council,hair. We are in the process of adding to their number four generallst NIOs-at-Urge, who will not have the quasi-operational staffof their colleagues, and whose primary function will be to provide that element of peer review tbat has been missing. Me are developing ways In which the Council can workore colleglel manner without detracting from the Individual responsibilities that some of Itswill have for serving the DCI. In other words, the close linkage to policy will continue.
Another change will be the creationmall staff oflerienced analysts available to the NIOs for the quick preparation of wide-ranging analytic papersupport of policy, and especially for
th* drafting of National Estimates. Thii will differ from tha earlier Cstlmates Staffng firmly rooted In thef th*ith It* personnel drawn from those agencies on rotation. On* of tho disruptive features of the NIO system was the need for th* NIOrafter whenever an Estimate was required. This slowed down th* process, complicated planning for th* contributing agencies, and lead to product not of uniformly high quality. Byedicated1 staff, we hope to reduce these problems without departing from thethat the best analyst, wherever he or she Is located In thekill be sought out for the Job.
Finally, there will be some simplification of th* Estimates process Itself to make It less auscle-bound.
CIA was directed by tbe National Security Act7 toand evaluate Intelligence relating to the nationalational Estimates aatody that function, whicht least as Important today as It appeared It 1s. however, ouch more difficult In the more complex and fast-moving world Theesponse to this world. Theynable us to deal more effectively with policy demandshe future.Original document.