THE DI MISSION IN THE 21ST CENTURY

• Collection: FOIA Collection
• Collection: Declassified Articles from Studies in Intelligence: The IC’s Journal for the Intelligence Professional
• Document Number (FOIA) /ESDN (CREST): 0000621389
• Document Release Date: July 30, 2014
• Case Number: F-2013-02322
• Publication Date: December 1, 1990

Approved for Release: 2014/07/29 000621389TITLE:AUTHOR:VOLUME:The DI Missionin the21st CenturyYEAR:1990(b)(3)(c)34 ISSUE:WinterApproved for Release: 2014/07/29 000621389pproved for Release: 2014/07/29 000621389'STUDIES ININTELLIGENCEA collection of articles on the historical, operational, doctrinal, and theoretical aspects of intelligence.eivo,e 16All statements of fact, opinion or analysis expressed in Studies in Intelligence are those ofthe authors. They do not necessarily reflect official positions or views of the CentralIntelligence Agency or any other US Government entity, past or present. Nothing in thecontents should be construed as asserting or implying US Government endorsement of anarticle's factual statements and interpretations.Approved for Release: 2014/07/29 000621389Approved for Release: 2014/07/29 000621389The future is nowThe DI Mission inthe 21st Century(b)(3)(c)Before dawn on 23 October 1983, I wasawakened by a call from the CIA OperationsCenter. The US Marine barracks in Beirutapparently had been attacked with some type ofexplosive. As the political analyst on Lebanon, Icame in to cover the story. During that day, Ifound out that a car bomb had killed over 200 USmarines. By 10:00 a.m., analysts from the Direc-torate of Intelligence's (DI) Near East and SouthAsia (NESA) Division were briefing CIA officials,  other agencies,(b)(1)  (b)(3)(c)Although information on the tragedy was tricklingin, we believed we were better informed thanalmost anyone in Washington about what hadhappened.If a similar tragedy occurred today, the mediacoverage of it would be different. The entirecountry would be tuned to the Cable News Net-work (CNN), which, more quickly than the othernetworks, would devote its air time to showingreports from its correspondents on the scene. Wewould have little reason to doubt the enormity ofthe attack because we would be seeing live cover-age of the devastation. (In 1983, it was not untillater in the morning that we had a firm handle onhow many had died.) The Commonwealth repre-sentatives would not have to come to Langley,knowing that the CNN coverage was timely andgenerally accurate. CNN also would be interview-ing "experts" who would speculate on the perpe-trators of the attack and its ramifications. Thewidespread TV coverage would largely preemptwhat an analyst would say in print the next morn-ing, unless there was overnight intelligence177(175)(1)identifying the actual instigators. In fact, the firstthing the analyst would do when he or she got towork would be to turn on CNN so a timelysituation report could be written.The pace of change in the world during the 1980smeans that "future shock" is now. As the exampleon Lebanon illustrates, the impact on the DI hasalready been substantial. Critics are questioningthe CIA's value, and the alleged failure to predictmajor world events has left the DI uniquely vul-nerable to cheap shots. Whatever the case, the DIfaces revolutionary challenges during the next 10years. These challenges will be particularly formi-dable because of the rapid changes in the area inwhich the DI makes its living?information.Anyone who has worked in the DI during the last10 years has his own idea of the nature of thesechallenges. Based on my 11 years of experience inregional offices, I believe the DI will have to dealwith three major challenges: the information age,the devaluation of intelligence, and a crisis of self-doubt.The Age of Cable NewsCNN recently ran an advertisement claiming thatpeople in Washington who need to be in the knowwatch CNN. The wording of the commercial madeone wonder whether or not CNN in fact was awarethat the major intelligence centers in Washingtondepend upon CNN for timely information. Ana-lysts who have worked on several recent crisesreadily admit that they acquired information fast-est through cable news.Approved for Release: 2014/07/29 000621389,SecnsrApproved for Release: 2014/07/29 000621389(b)(3)(n)It is one thing to say that analysts use CNN. It isanother to claim that the revolution in informa-tion has forced changes in the DI's analyticaleffort. But I would argue that the almost instanta-neous coverage of world events by the media, thevisual impact of wall-to-wall CNN news feed, andeven the more comprehensive articles by the dailynewspapers have made it much more difficult forDI analysts to provide the policymaker withunique and timely current intelligence. We nolonger have a monopoly on information or analy-sis. CNN competes with us on the former, andNational Public Radio, Nightline, and in-houseexperts vie with us on the latter.In some cases, DI analysts have been put on thedefensive because their articles in the NationalIntelligence Daily (NID) did-not jibe with the waythe story played in the press. A case in point is therebel offensive in 1989 in El Salvador. Unlike thepress, the DI believed that the offensive had failedmilitarily from the start. What we failed to antici-pate was that press coverage of rebel efforts wouldkeep the offensive alive for weeks and that, in fact,the rebels would start conducting operations withan eye toward media coverage. This coverage andits different slant forced an adjustment in the toneof current intelligence. We became reluctant towrite unequivocably that the military had thesituation in San Salvador under control, when, forall intents and purposes, it did. It was also difficultto maintain that rebel forces were reeling from thefailure of the offensive when CNN reporters,standing in front of a smoking building, claimedthat this or that attack indicated the offensive wasfar from over. From the military perspective, theywere wrong, but from the journalist's point ofview, they were right.The vast information revolution that CNN onlysymbolizes could spell real trouble for the CIA ingeneral, and for the DI in particular. Unlike thesituation 10 or 15 years ago, we cannot take it forgranted that our coverage of political change inCountry X will be more timely and comprehensivethan that of the media. It becomes more difficultto present analytical judgments that run counter topictures from the scene or the lengthy analysis of aforeign correspondent who can now communicateCenturyinstantaneously with his network or newspaper viafax and satellite. Critics of the press have longcomplained that the media?particularly TVnews?sensationalizes domestic events; now weare watching journalists apply more of their talentsto international stories.Solving the MysteryThere is the argument that the DI product issuperior and more important than the story in theNew York Times because of our access to classifiedinformation. One wishes it were so, but a look at amonth's-worth of NIDs indicates that the over-whelming majority of the articles is not based onclassified information. Evidently, they are classi-fied primarily because of their analytical content.The following numbers reflect NIDS publishedfrom 30 April to 25 May 1990. Items were catego-rized based on their regional subject matter, so thefigures do not necessarily indicate which regionalor functional office drafted the piece. Some of thehighlights:? The NID carried 125 items on the USSR.Only five were classified Top Secret, and only26 were Secret Orcon or Top Secret. Almostall of the stories on Lithuania, arguably themost important international issue at thattime, were Confidential Noforn.? One-hundred sixteen European stories wererun, of which four were Top Secret. Thatnumber expands to 26 when stories at theSecret Orcon level are included.? Stories dealing with Middle Eastern andSouth Asian issues generally had the highestclassification. Out of 108 pieces, 26 were TopSecret and 63 were at least Secret Orcon. Thiswas the only regional area where items classi-fied Secret Orcon and above were in themajority. The India-Pakistan problem and itsnuclear implications contributed greatly tohigher classifications, although the principalfeature article that appeared on the issue wasSecret Noforn.18Approved for Release: 2014/07/29 000621389Approved for Release: 2014/07/29 000621389CenturyThese numbers overemphasize the importance of"secrets" to our analysis. The pieces were catego-rized based on the highest classification of anyparagraph, although many longer items had onlyone or two sentences at the Orcon or Top Secretlevel. Two items in particular stick out because oftheir unusual classifications. One had a lead thatwas Secret Noforn; all the other paragraphs wereConfidential Noforn. The lead of the other wasclassified Orcon, but nothing else in it carried thatclassification.Secrets admittedly were important in the coverageof some issues. In addition to India-Pakistan,these included nuclear proliferation, chemical andbiological warfare matters, and the Soviet military.More often than not, however, the NID earned itsstripes because of-only one or two items thatappeared in the book.It is unclear whether these figures represent achange from five or 10 years ago, but at least somerecipients of the NID have to wonder about all themystery. If the DI product does not provideunique information, if it appears after severalnews programs and papers, and if it can only befour paragraphs or 36 words long, we may not befulfilling the needs of our consumers or ourmission.Blowing Calls, Blowing SmokeDI analysts are the international affairs experts.Almost every policymaker, and particularly mem-bers of Congress, will tell you that CIA briefingsare the best and most objective. So our expertiseshould help us compensate for whatever compara-tive disadvantages we may or may not have. TheDI, however, appears to be undergoing a crisis ofself-doubt about its discipline, an uncertainty thatserves as a drag on our expertise. Moreover, theformats and products that we use in some casesactually work against our greatest strength, dilut-ing our expertise to the point of generality.Two articles in the media illustrate some of theissues that fuel the DI's concern. On 14 May 1990,19(b)(3)(n)NewsweekNewsweek wrote that "one of Bush's chief foreign-policy advisers ranks the agency's recent trackrecord as only fair. . . good at analyzing trends . . .but poor at predicting the timing of events in thecollapse of Eastern Europe." Newsweek quotesformer Undersecretary of Defense Fred Ikle assaying that CIA analysts "collect a lot of facts andorganize them very nicely. But their predictionsare wrong." The next month Roger Morris, in aNew York Times article that belabored the Agency,wrote that "political intelligence could be vested ina revitalized foreign ministry and diplomatic ser-vice, where it belongs."Many DI analysts knowingly react to such criti-cism by joking that if they could predict theprecise timing of events, they would work on WallStreet. But the cynical shrug masks our own disap-pointment over our inability to call every electionor forecast the fall of each dictator, let alonepredict the triumph of democracy. We conductpost mortems on our analyses and invite secondand third opinions, but we remain confused as towhy the DI technique of employing good minds toknow everything about a subject and write for-ward-looking analyses based on intelligence doesnot hit the mark more often.This confusion and uncertainty in some cas(b)(1)affects our analyses. For  example, even thou( b)(3)(n)uban leader Castro ap-pears secure, we are reluctant to make that callunequivocably. Many of our caveats are honest? (b)(1)   (b)(3)(n) Some caveats,however, are the result of confusion about ourdiscipline. We realize that other political an.(b)(1)working on other countries (b)(3)(n)predicted a continuation ofthe status quo, and reversed themselves in the MDone week later. We are spooked. We are dismayedthat on almost any issue we can point to anindividual who was more correct than the organi-zation. What makes it worse is that his or her callwas sometimes based entirely on instinct.Much of this criticism is healthy. We need toanalyze our own effort, but we are sometimesApproved for Release: 2014/07/29 000621389et.reSeertrr? (b)(3)(n)?Approved for Release: 2014/07/29 000621389second-guessing ourselves before the fact. One ofthe most important lessons to be learned is that nomatter how professional we are, we will be wrongon occasion. We predict based on our rigorousapproach or we rely on instinct. My predictionwould be that instinct overall would have a muchworse track record.Communication of our expertise is complicated byour format and processes. Particularly in currentintelligence, the physical limitations of the productlead to analytical limitations. In the survey of NIDpieces, almost every office produced more "inbriefs"?three- to five-line nuggets of informa-tion?than any other category. Anyone who hasever written an "in brief' knows that it is impossi-ble to impart much expertise or any degree ofsubtlety in the item. In addition, the NID only hasroom for a small number of pieces. Every officehas horror stories concerning an item that waitedin the NID queue just long enough to be scoopedby the newspapers. Our advantage in expertise isjust not relevant if it has trouble leaving thebuilding or is oversimplified.Answering the ChallengeThe DI has for years recognized the need to lookhard at the way it does its business. The explosionin the production of less formal typescripts, forexample, has been one successful way of providingour expertise to consumers more promptly, com-prehensively, and informally. Proposals for anelectronic or computer-delivered NID have alsobeen made. The design of the electronic NID,which would allow readers to search computerfiles for more in-depth information on a subject ofparticular interest, anticipated the need to providemore analysis more quickly. Budget constraints,however, have delayed many reform proposals.Given the current fiscal and Congressional cli-mate, the CIA and the DI will probably have tolook for less expensive ways to respond to ourchallenges, at least in the short term. A case can bemade, as Roger Morris did in his article, that theCIA no longer needs to be in the business ofCenturypolitical?and current?intelligence. The DI, how-ever, is too good an organization and its people aretoo talented to abandon this mission. But the timehas come to retool.Decentralize Current IntelligenceThere is no easy way to reform the current intelli-gence process, but perhaps the pain of transitioncould be eased by patterning the process after whatis arguably the DI's most successful product, thetypescript memorandum, which is produced at theoffice level. This would decentralize current intel-ligence. Each distinct regional and functionalarea?in most cases, each office?would produce adaily review. The publication would be standard-ized to the extent that each office product wouldcarry identical cover sheets. It could be called theDI Daily Review, with each office using a subtitleto distinguish its product. Other features couldinclude:? A distribution list, similar to the office'stypescript distribution list, aimed at the As-sistant Secretary level and below.? Washington-only distribution at opening ofbusiness, using, at least initially, secure faxtransmission.? A Monday edition that, recognizing the gen-erally slow information flow on weekends,would feature more op-ed, speculative pieces.Analysts would have the opportunity to usemore of their expertise by writing about ideasrather than facts. Many newspaper articles donot advocate policy but provide a differentway of looking at a problem. We should dothe same.? Use of an early-morning team of analysts andmanagers who would use new information orwhat appeared in the press to update existingarticles. A lot of frustration could be avoidedif analysts could use such phrases as "Despitepress acounts this morning, we still believethat . ." Every office may not always need20Approved for Release: 2014/07/29 000621389Approved for Release: 2014/07/29 000621389Centurysuch a team, but today in crisis situations analystsalready come in early to update the briefers whodeliver copies of the President's Daily Brief (PDB).The creation of daily typescripts need not endan-ger the NID, which has a vast clientele overseasthat could not be serviced by the office dailies.And many of its several hundred Washingtonreaders probably would always want a generalworld overview. Their assistants, however, mightpass on to them articles in the DI Daily Reviewthat supplemented NID pieces. The DI dailies, infact, would be more likely to pick up where theNID left off by providing dailies to the AssistantSecretary level and below, whose current intelli-gence needs are not currently met, at least not bythe DI.The NID production process need not changemuch, and it might improve. Offices would pro-pose items for the NID, and they would also givethe NID Staff their menus for their own dailypublications. The preliminary version of the publi-cations would be sent to the staffs late in the day,offering the NID and PDB a chance to pick upstories that turned out to be more interesting thantheir titles suggested. Similarly, offices wouldshare stories among each other to supplement theirdailies and to ensure that CIA published only oneversion of a story.Decentralized current intelligence also would posesome problems. They might include:? Classification. As the survey of NIDs indicat-ed, the DI uses little Orcon and Top Secretmaterial. Presumably, the Directorate of Op-erations (DO) would allow use of most itsreports in a document that was Washington-only and had a distribution list similar totypescripts. It would be more difficult to useTop Secret material. Nevertheless, it does notmake sense to limit the distribution of ouranalysis and inconvenience our readers by  publishing only in the NID.  (b)(1)(b)(3)(c)(b)(3)(n)21_Secret'(b)(3)(n)? Delivery. This is likely to be a bigger problemand to require funds and bureacratic finesseto solve. Secure fax centers are springing upeverywhere, but more will have to be createdto handle the transmission of as many as adozen different dailies. The real obstacle,however, probably would be delivery within abuilding. One answer might be to create morepositions for DI representatives in Washing-ton.(b)(3)(c)Each regional andfunctional omce neeas to have its own repre-sentative in the appropriate bureau at State.Eventually, the DI might begin to use com-puter linkups for its daily publications. Thematerial could be transmitted from oneword-processing station to another, allowingfor the production of a letter-quality productright in the Department of Defense or State.The delivery problem in an electronic age isan overarching one that the DI will have tosolve, whether or not we change the way wedo current intelligence.? Quality Control. There is a tradeoff betweenquality control and decentralization. An ar-gument could be made, however, that moretimely and in-depth analysis would help re-dress the balance. DI management also wouldstill be able to review the subject, if not thesubstance, of daily office publications.What is GainedSome of the benefits, such as a more timelyproduct and greater coverage of issues, are readilyapparent. But many other benefits are possible.Better Relations with Policymakers. Daily publica-tions that covered smaller parts of the world ornarrower issues should improve DI access to poli-cymakers. The development of a distribution sys-tem that allowed the DI product to be deliveredpersonally in and of itself should facilitate feed-back and subsequent improvement of our effort.The Assistant Secretary of State for Africa, forexample, would receive a document?perhaps fouror five pages in length?that contained more of theApproved for Release: 2014/07/29 000621389Approved for Release: 2014/07/29 000621389(b)(3)(n)information she really needed. If our product wasnot satisfactory, if our readers were not reading it,adjustments could be made almost immediately toreflect real demand. If, on the other hand, ouranalysis proved useful, we probably would in-crease our access to policymakers.Better Relations with Congress. Decentralized cur-rent intelligence probably would allow us to pro-vide more information to the Congress. Congress-men appreciate CIA briefings, but they areunhappy when we are reluctant to provide thingsin writing. Reading the NID is even more difficulton the Hill. Daily reviews, however, presumablycould be shared more easily with some Congress-men and their staffers?at least on an ad hocbasis?because of their lower classification. In anycase, it is in the CIA's interest to be of greaterservice to the individuals who will control increas-ingly tighter purse strings. DIA officials recentlyremarked that in a time of budget constraint theiranalytical arm was emphasizing ad hoc products,including answering Congressional requests, anddata bases. We should look for ways to expandCongressional access to current intelligence.Our inability to get our message across to Congressas a whole may have already hurt us in the budgetcrunch. Many members of Congress seem to haveaccepted media depictions of the changes in East-ern Europe and the Soviet Union as triumphs ofdemocracy. What we may have witnessed instead,however, is the triumph of ethnicity and geogra-phy over hollow ideology. DI analysts in fact havemade that point, which indicates that countryanalysis remains important. Despite what somemay think, change in Eastern Europe required usto pour more resources into studying the area.Some Relieffrom Criticism. The provision of morecurrent analysis in less classified formats?the"demystifying" of intelligence?could help defendagainst unjust criticism. Certainly, the way the DIdoes its business today probably invites some ofthat criticism. We shroud our analysis in deeplayers of classification. The reader, much likeGeraldo Rivera opening Al Capone's safe, expectsto find something wonderful and magical inside?or at least some good secrets. Instead, he finds, forCenturythe most part, facts that he could have read aboutin the New York Times and analysis that onlyleaves you wanting more. Making our intelligencemore accessible and providing more subtlety on adaily basis would probaby boost our ratings orprovide for a better understanding of our limita-tions. Having more space to write about problemswould almost certainly aid in explaining compli-cated issues and avoid overly simplisticjudgments.The DI, in fact, should consider whether or not itsanalysts could generally be more accessible to thepublic. We have already increased briefings of thepress in the last year or two. We could "go public"in a much bigger way: office directors or nationalintelligence officers could appear in public forums,perhaps on TV, to provide commentary on aparticular topic. Such exposure, particularly onTV, would have to be on a selective basis. It coulddo us a lot of good. Our top officials ought to bejust as adept as any other bureaucrat in distin-guishing between classified and unclassified infor-mation in making his comments.The Role of ResearchThe changes suggested in the way we do currentintelligence would require a shift in resources. Amuch more active current intelligence processwould be likely to cut down the amount of timeanalysts could devote to longer papers. One sug-gestion would be for us to write more typescriptsand real research papers and to phase out intelli-gence assessments. Typescripts usually answer areal question; research papers are necessary todevelop analyst skills and a foundation of knowl-edge. Intelligence assessments, however, tend to beneither fish nor fowl. They are schizophrenic crea-tures that seek to satisfy both the policymaker,who does not have time to read anything too long,and the expert, who wants to learn something heor she does not already know.There is some concern in the DI that analysts willfail to become sophisticated thinkers if they arediverted from research and spend too much timewriting current intelligence. But that concern can22Approved for Release: 2014/07/29 000621389Approved for Release: 2014/07/29 000621389Centurybe turned around. Is our current intelligence soshallow that analysts can be proficient withoutreally knowing their subject matter? The answer is"maybe"?many offices let their new analystswrite short items for the NID early on because theproducts are fairly straightforward. Daily reviews,on the other hand, would make more demands onanalytical expertise.Foreign CorrespondentsReforming our Washington work habits would goa long way toward ensuring that the DI remainsthe most important analysis resource for policy-makers. We would, however, still be the onlyinternational information service without its owndedicated foreign correspondents. This problem islikely to become worse for the DI in the 1990s.Secure phone lines between embassies and theState Department have already reduced cable traf-fic in crisis situations. In addition, the revolutionin open-source information and the changingworld environment is leading the DO to reassessits own collection priorities. Internal political de-velopments may become less important becausethe State Department and the New York Timesalready report on them. But where does that leavethe DI? Where do our secrets come from? (b)(3)(n)(b)(1)(b)(3)(c)(b)(3)(n)The DI could look at several options to expand its  international coverage.(b)(1)    (b)(3)(c)  (b)(3)(n)23The Bottom LineSome change is necessary and some is inevitable.And the DI should be in charge of both kinds.Even if Brezhnev rose from the dead tomorrow,the revolution in information technology alonewould force us to alter our work habits. One couldeven envisage a day when the NID evolves into avideo newscast on secure TV nets that are beinginstalled all over Washington.The DI and current intelligence may be creaturesof the Cold War, but both can prosper from itsend. US policymakers need even more informa-tion on world events. More important, they needan organization that will sift through all of thatinformation, analyze its importance, and place itin the appropriate context. The DI's mission todayis to make sure we do that better than anyone else.This article is classified55,CRETApproved for Release: 2014/07/29 000621389(b)(3)(n)--Seterg?