PROCEDURE FOR MANAGING INTERDISCIPLINARY INTELLIGENCE PRODUCTION

• Collection: FOIA Collection
• Collection: Declassified Articles from Studies in Intelligence: The IC’s Journal for the Intelligence Professional
• Document Number (FOIA) /ESDN (CREST): 0000617220
• Document Page Count: 10
• Document Release Date: July 30, 2014
• Case Number: F-2013-02322
• Publication Date: September 1, 1981

Approved for Release: 2014/07/29 000617220TITLE: A Procedure for Managing InterdisciplinaryIntelligence ProductionAUTHOR:(b)(3)(c)VOLUME: 25 ISSUE: Fall YEAR: 1981Approved for Release: 2014/07/29 000617220pproved for Release: 2014/07/29 000617220INTELLIGENCEA collection of articles on the historical, operational, doctrinal, and theoretical aspects of intelligence.All stlatemepts 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 000617220Approved for Release: 2014/07/29 000617220A PROCEDURE FOR MANAGING INTERDISCIPLINARYINTELLIGENCE PRODUCTION(b)(3)(c)The most popular buzz-word in intelligence production circles these days is-interdisciplinary.- One hears a great deal about the needs of intelligence consumersfor integrated products?reports that pull together the economic, political, military,technological and societal aspects of an issue so that a policy maker gains a broad un-derstanding of it. This is necessary, it is argued, because policymakers themselves haveno time to perform the integration. Hence, intelligence reporting that focuses on onlyone dimension of an issue at best can provide some useful, but incomplete,information; at worst it may be dismissed by the consumer as irrelevant.That one of the functions of intelligence is to provide policymakers with completeand relevant information on an important issue is obvious; the problem has been howto organize and manage the production of such information. This article describes onemethod that has been used with some success in the National Foreign AssessmentCenter (NFAC) and has now been institutionalized in the research planning process.My purpose is to record the lessons that we learned in producing a major interdisci-plinary study on the development of the prospects for Soviet military power so thatthose charged with managing issue-oriented research can profit from our experience.(Those more interested in our substantive findings can refer to the study, TheDevelopment of Soviet Military Power: Trends since 1965 and Prospects for the1980s.)Origins of the StudyFor several years it has been evident to the Intelligence Community that theenvironment for Soviet decisionmaking on national security issues is changingmarkedly.? Growth of the Soviet economy is slowing and is likely to decline further in the1980s as energy and manpower shortages compound the longer-standingproblem of poor productivity.? Many of the top military and political leaders will be replaced soon: most arein their 60s and 70s and a number are in poor health.? Growing Western concern about Soviet military power, China's opening tothe West, and instability in Eastern Europe and the Third World are alteringthe international setting.? Advances in technology are providing new opportunities to upgrade militarycapabilities.C011AL 37Approved for Release: 2014/07/29 000617220Approved for Release: 2014/07/29 000617220CON F NTIALThese complicating factors make it particularly difficult for intelligence analyststo forecast Soviet behavior. But the tools available to analysts have become morepowerful:? The Office of Strategic Research (OSR), drawing on analysis performedthroughout the Intelligence Community, has compiled an all-source data baseon trends since the 1950s in Soviet military production, order-of-battle,manpower, and expenditures.? The Office of Scientific and Weapons Research (OSWR) has gained valuableinsights into the Soviet system for managing missile and space programs andthe development capacity of the Soviet aerospace industry.? The Office of Economic Research (OER) has developed new models to assessand predict Soviet economic performance.? The Office of PolitiP1 A jyis OPA has instituted research projects aimed(b)(1)atviews on policy issues.oviet policymakers and their possible? Analysts throughout the community?in OSR, OSWR, OPA, and in DIA andthe military services?have greatly improved their knowledge of the Sovietpolicy process.It was this coincidence of an important policy issue and new analytical tools thatled OSR to initiate the process that led to production of a major interdisciplinaryassessment of the prospects for Soviet military power.The ProcessThe process began in early 1979 with planning for a two-day seminar on "SovietDefense Decisionmaking.- At that seminar analysts from the NFAC production officesbriefed each other and their managers on their current understanding of Sovietdecisionmaking processes and on the factors likely to influence Soviet choices onmilitary forces and policies for the 1980s. Out of the discussions a consensus emergedin favor of a broadly-based, interoffice research effort aimed at producing anintegrated report on the prospects for the development of Soviet military power in thecoming decade. The group also identified the major research themes that theindividual offices should investigate.Following the seminar, an interoffice working group was formed to turn thegeneral research themes into a concrete plan  of action. (I  chaired the group; othermembers were (b)(3)(c)  of OSWR.  of OPA; of OER; andf OSR.) The working group was tasked with evaluating the current NFAC re-search program, recommending redefinition and rescheduling of planned research,and proposing new projects to meet the goals of the plan.Initially, some members of the working group expressed reservations about theplan, and some managers resisted the notion that an interoffice group should play arole in determining their production priorities. We also touched some sensitive nervesby recommending a few projects that the offices, for one reason or another, were lessanxious to undertake. Eventually, however, we hammered out a consensus at theworking group level.The working group codified its research proposals in a formal document, signedby all of the NFAC office directors, and set up a management information system to38Approved for Release: 2014/07/29 000617220Approved for Release: 2014/07/29 000617220CC.:T?1<NTIALkeep the directors aware of the program's progress. The program plan consisted of aninterim report on recent trends in Soviet military forces, som(b)(1  ),uilding blockpapers on individual factors affecting Soviet policy, and an overview report forpublication in the spring of 1981.By the summer of 1980, enough work on the building blocks project had beencompleted to warrant a working-level seminar to outline the major theses of theoverview paper, and identify areas for further research. During October andNovember, a first draft of the overview paper was put together. Coordination of thepaper was completed in March and in May the DCI transmitted the report to thePresident, Vice President, Secretary of State, Secretary of Defense and Assistant to thePresident for National Security Affairs. Several hunderd other consumers received thepaper through normal dissemination channels.Reaction to the paper has been highly favorable. Several high-level recipientshave praised the integrative character of the analysis. The report has contributed todebates on US military programs and has been made required reading for seminars atone of the war colleges.Lessons LearnedIn the process of producing our paper we learned a great deal about planning andmanaging interdisciplinary research. First, we learned that it is possible, within ahierarchical organization, to carry out an integrated research program. But we alsolearned that managing such a program is difficult and demanding: both carefulplanning and vigorous follow-up are vital.More specifically, we found that there are a number of essential steps in aninterdisciplinary project:? Choosing the issue properly.? Sharing information and views among analysts with different backgrounds.? Defining the issue in terms of its major research themes.? Appointing a project leader and working group to develop a program plan.? Assessing the adequacy of currently planned research in answering keyquestions about the research themes.? Preparing a program plan with milestones for completion.? Formal commitment of resources to the effort.? Monitoring the progress of the research program and taking action to keep iton track.? Integrating the findings into an overview paper.As a result of our experiences, I believe that all of these steps are needed,regardless of the organizational structure established to carry out interdisciplinaryanalysis.Issue IdentificationThe first order of business is to frame the issue so that it can be an effective guideto research. If this isn't done properly, the project team may be unable to assess theeffectiveness of its plan. In the words of psychologist David Campbell, "If you don'tknow where you're going, you're likely to end up somewhere else."A particular challenge in this first step is to be sure that you are dealing with anissue. It's easy to confuse the process of defining research to address an issue or answera question with the (equally important) problem of ensuring that there is adequateanalytical coverage of a topic or geographic area. An issue like -the ability andCOrF.113?E<AL 39Approved for Release: 2014/07/29 000617220Approved for Release: 2014/07/29 000617220CON FIRlTIALwillingness of the NATO allies to shoulder a larger defense burden" can be addressedby the method described in this article. But "Africa" and "Nuclear Proliferation" areareas and topics, not issues, and different techniques probably are required to manageresearch dealing with them. A good test is to try to frame the issue in terms of a ques-tion that can be addressed in a single paper. If you can't, it probably isn't an issue.Information SharingThis is an indispensable step in an interdisciplinary research effort. It is necessaryto combat a common fallacy?that the information held by an individual or group isgenerally known by others. (This is rarely the case, even when individuals ororganizations have close working contacts.) It is also important to have a common basisfor planning the research program.Successful information sharing, like the other steps in an interdisciplinaryresearch program, requires careful planning.? The atmosphere must, be right. We found that holding a two-day initialplanning seminar in a remote location permitted an intensive exchange -gfideas away from day-to-day pressures.? The briefing topics must be well-chosen. They should provide i'wide range ofperspectives on the issue, but they must be relevant. There may be a tendencyto brief on a topic not related to the issue simply because the information isavailable. This should be resisted.? The group dynamics must work. The conference planners must get the rightmix of personalities and expertise, even if this results in an imbalance inorganizational representation. They must also, however, ensure that the keyanalysts who will be working on the program and managers from theparticipating organizations are present. This is needed to ensure from thebeginning a commitment to the goals of the program.Theme DefinitionThe culmination of the initial sharing of information should be the identificationof themes for organizing the research effort. If this is not done, there is a danger thatparticipants in the information sharing phase will leave, having been entertained withbriefings but without a clear idea of where the effort will lead.This is a tricky phase of the process, and the planners must be careful to explainits purpose to the participants. There will be a tendency for many to want to arguesubstance or draw conclusions, but this is not the time. Instead, the participants shouldstrive to understand all of the facets of the issue and to record their understanding as aset of themes for further investigation.A technique that we used successfully was a variation of brainstorming called"funneling applied creativity." After immersing seminar participants in substantivebriefings for a day and a half, we solicited from them their ideas on research themesthat should be followed up. The ground rules permitted anyone to make a proposaland prohibited any criticism or discussion until all of the suggestions had been put for-ward and recorded on flip charts. We then opened the floor for discussion and deletedsome themes and refined others. The themes that we settled on were:? Soviet defense decisionmaking and management.? Soviet military goals, threat perceptions and requirements..? Defense and the Soviet political succession.? The Soviet defense-industrial establishment.? Defense and the Soviet economy.40 CC"FLI<NTIALApproved for Release: 2014/07/29 000617220Approved for Release: 2014/07/29 000617220COrtikk.NTIA? Trends in Soviet military forces.? Soviet military technology.Appointing a Management TeamThe final act of the initial planning seminar should be to appoint a managementteam. The team will have two tasks:? To formulate a program plan.? To monitor and report on execution of the plan.The team should include representatives of all organizations that have substantiveresponsibilities related to the research issue. But more importantly, it must have theright blend of personalities and expertise. A critical task for managers, then, is to selecta knowledgeable group of senior analysts or first line managers who know both thesubstantive issues and the research and production capabilities of their organizations.Assessment of Current ResearchOnce there is agreement on the definition of the research issue and its themes, ther'first task for the management team is to assess the adequacy of current plaps. The dif-ficult parts of this step are defining the criteria for measuring adequacy and findingout exactly what the current plans were.For our measuring stick we used a set of key questions structured into the sevenresearch themes (see box below for examples). The concept was that research would beadequate if it would permit confident answers to all the key questions. To uncover in-formation about current plans, we found personal contacts to be far more useful thanpublished research programs. (The published programs were often cryptic, out of date,and unrealistic.) For this reason, it is important that the members of the working groupbe in a position to find out what really is going on in their offices and to provide realis-tic assessments of the intended scope and completion dates of currently scheduledprojects. It is particularly important to go beyond the titles of planned researchprojects and find out more about their purposes and substantive theses. Without thisinformation, it is impossible to judge the appropriateness of the planned research.Research Theme 3: Defense and the Soviet Political SuccessionKey Questions: What are the possible scenarios for the Soviet politicalsuccession? Who are the likely candidates for top leadership posts? What aretheir views and preferences in the area of defense resource allocation andmilitary policy? How much flexibility will a new regime have to alter resourcepriorities and policies?Research Theme 6: Trends in Soviet Military ForcesKey Questions: What have been the key trends in Soviet military forces?Are they consistent across service and mission lines? What programs are nowunder way or known to be Planned? When will they impact on force structureand capabilities and to what extent?What deficiencies do the Soviets perceive in their forces? What can weinfer from this abbut new programs initiatives for which we currently have noevidence?CON F1Q ENTIALApproved for Release: 2014/07/29 00061722041Approved for Release: 2014/07/29 000617220514:ZENTIALPreparation of a PlanOnce the criteria for success are established and currently programmed researchis identified, the project leader and management team can begin to prepare a programplan. The first step is fairly mechanical?array the scheduled projects by theme,compare their estimated completion dates, and identify necessary adjustments. Someprojects might logically be combined or made into joint projects.For those key questions that will not be adequately answered by currentlyplanned projects, the project manager and his group should prepare project proposals.These should be specific enough to be a basis for resource decisions by officemanagers. They should include, for example, a statement of the scope and proposedschedule, level of analytical effort and any special requirements such as ADP orcontractor support. It's useful to identify the analysts (or at least the branches) bestsuited to carry out each proposed project.At this stage, the management team should also identify requirements foroverview papers to integrate the findings of the building-block projects. If possible,group should should identify potential drafters so that they can be involved in, the entire re-search program.After all of this homework is done, the project manager should prepare aconsolidated program that identifies all of the building-block and overview papers,sets forth schedules and assigns responsibilities, and recommends new projects to fillgaps in current plans.Formal Commitment of ResourcesAfter the plan has been prepared in draft, it should be submitted for approval bysenior management. The approval should be formal and specific. To make a programwork, it is necessary that managers at all levels assign sufficient analytical resourcesand protect them. When approving the plan, managers should be made aware thatthey have made commitment to deliver their projects and that their own supervisorswill be apprised of the progress of research.Once the plan is completed it should be circulated to all analysts and managersinvolved. (We gave our plan a little extra dignity by having it bound in hard covers.) Itis essential that the participants understand their role in the plan. One of theimportant functions of a program plan is to alter the directions of individual projectsso that they are responsive to the key questions. This can't be done unless people knowwhat the questions are and how their own work relates to them. The only way to en-sure this is good communications. It's helpful to record in the Advance Work Plans ofboth analysts and managers that their projects are related to an interdisciplinaryresearch issue.Monitoring and TroubleshootingEven the most carefully prepared plan will fail without aggressive follow-up.Periodic reports to senior managers are essential to keeping the program on track. (Wesent reports quarterly.) The project manager and team must be frank in identifyingproblems and pressing for their solution. We found that most problems solvedthemselves once they were made known. But occasionally we had to keep remindingmanagers of their commitments and on one occasion had to convene a formal meetingof office directors.42CC-5-Rlsti*TIALApproved for Release: 2014/07/29 000617220Approved for Release: 2014/07/29 000617220CN'T4IZZ1TIALAnother essential ingredient is an effort to promote better communicationsamong the analysts working on the building-block projects. Merely publishing aprogram plan will not make analysts aware of their responsibilities and theirimportance in the overall process. It will be necessary to organize briefings, seminars,and retreats to facilitate communications, and the members of the working group mustbe prepared to spend a great deal of time and effort trying to break downbureaucratic and disciplinary barriers.Production and EvaluationThe final step in the process is to integrate the results of the building-blockprojects into a coherent overview paper. The danger in this stage is that it maydegenerate into "analysis by staple." Interdisciplinary analysis must be more thanediting inputs from various specialists; it should entail a real effort to combine themethods of several disciplines to create new insights.It's particularly important to select the right drafting team. It's not enough toform a group of, say, an engineer, an economist, and a political scientist. At timesfrom different specialties have such different mindsets and vocabularies that theyliterally cannot communicate. Ideally, each drafter should have experience in at leasttwo disciplines. And whatever their background the drafters must work effectivelytogether?have the right body chemistry.We found that the close personal relationships that were forged during theworking group's seminars were an important ingredient in promoting interdisciplinaryanalysis. The participants knew each other, appreciated the work of other organiza-tions, and learned the concepts and vocabulary of other analytical disciplines wellenough to communicate effectively. Moreover, most coordination problems werecleared up during preparation of the building-block papers, leaving only a fewcontentious issues for the overview report.This is not to say that we had no problems. We faced two serious substantive dis-agreements. Opinion was sharply divided between OSR and OER on prospects forSoviet defense spending and between OSR and OPA on the extent to which Sovietforce procurements were driven by military or political considerations. These issueswere to have been addressed in building-block papers, but the schedules for the papershad slipped so that coordinated conclusions were not available for the overview report.Eventually we solved the problems?the first by more analysis and the presentation ofalternative projections; the second by textual changes and a realization that thedifference was largely semantic. But it would have been more efficient to identify thedifferences of opinion earlier, in the building-block projects, so that they could beworked out in a more rigorous fashion.The process of managing interdisciplinary research should not stop with publica-tion of an overview paper. It's necessary, then, to have a formal evaluation of the pro-gram's accomplishments and of the remaining gaps in knowledge. The mechanism weused was yet another two-day seminar at which we reported to NFAC managers onwhat they had received for their resource commitment and solicited their help inidentifying and sponsoring follow-on research.Final ThoughtsThe process described in this article worked. There may be other ways ofproducing a major interdisciplinary report, but based on my experience I believe thereare five indispensable ingredients.CONF NTIAL 43Approved for Release: 2014/07/29 000617220Approved for Release: 2014/07/29 000617220COl?-ssirltors<IALCommitment by senior managers to the importance of the issue and awillingness to devote resources to it and protect those resources.Careful planning that involves all interested parties from the beginning.? Good communications both up the line to management and across organiza-tional lines among analysts.? The right body chemistry among planners and analysts.? Follow-up to ensure that all participants retain their commitment and carryout their responsibilities.There are commodities to be rationed carefully among issues that are genuinelyhigh in priority. My own opinion is that an organization the size of NFAC probablyshould not undertake more than about half a dozen major interdisciplinary projects ata time. A larger number would dilute the commitment of senior managers and reducetheir ability to protect the resources devoted to the programs. But if aggressivelymanaged and strongly supported, the process I've described has a high probability ofpayoff. 7- -"It is at all times necessary . . . that we frequentlyrefresh our patriotism by reference to first principles. It isby tracing things to their origin that we learn to under-stand them, and it is by keeping that line and the originalways in view that we never forget them."Tom Paine44 cIgIALApproved for Release: 2014/07/29 000617220