APPROVED FOR RELEASE 1994
CIA HISTORICAL REVIEW PROGRAM
2 JULY 96
CONFESSIONS OF A FORMER USIB COMMITTEE CHAIRMAN
David S. Brandwein
To many people the word "committee" triggers a reaction which ranges between revulsion and displeasure. Within the intelligence community, the likeliest targets for committee-haters are the USIB committees. Typically, criticisms laid against them are that their judgments tend to be waffled, they don't respond quickly to urgent tasks, and they don't come up with imaginative solutions to difficult problems. Hardly a year goes by without a fresh study by a high-level official or group of the "problem" of some or all of the USIB committees. Usually the objective of these studies is to improve the committees' effectiveness through reorganization and reallocation of functions. Actually, changes have been few and far between, and for the most part the committees have proved to be very durable. One cannot escape the conclusion that they must do some good, and that we have as yet not figured out how to come up with a better scheme for ventilating interagency problems and for achieving coordination on them.
Surprisingly, the reproaches visited on the USIB committees are seldom reflected back to their parent body, the USIB. One might say this stems from a natural reticence in openly criticizing the boss. But there is more to it than that. One needs to look at the people who make up the USIB, and then look at the people who staff the USIB committees. The USIB principals on the one hand are the top officials of large organizations. It can be assumed that they have reached these prestigious positions through a process of natural selection which rewards those individuals who combine a high order of intelligence, stamina, and dynamism. On the other hand, it is sad but true that the USIB principals have sometimes chosen mediocrities to represent them on the committees. The message here is that one cannot examine the USIB committees as abstract organizations. The caliber of the people involved is at least as important.
The foregoing is to introduce my subject, the committees. I propose to describe the inner workings of two USIB committees from the vantage point of the chairman — how they are staffed, what things they do, how they do them, and how they might be improved. Because this presentation is based mostly on my own experiences as chairman of two different committees — the Guided Missile and Astronautics Intelligence Committee (GMAIC) and the SIGINT Committee — it is probably more subjective than it is objective.
I Become a USIB Committee Chairman
My first deep involvement in the world of USIB committees was in November 1988, when I became Chairman of GMAIC. I had lobbied for the job, partly because of the attraction of the prestige attached to it. I served as Chairman of GMAIC until the summer of 1972, when I was asked to chair the SIGINT Committee, a job which I held until April 1973. The bloom was well off the rose by 1972, and my acceptance of the SIGINT job was not characterized by the same enthusiasm which I felt in 1988. Nevertheless, the experience was broadening, because the two committees concentrate on different aspects of intelligence. GMAIC is concerned primarily with the production of intelligence on guided missiles, while the SIGINT Committee is oriented towards collection tasking for certain forms of raw intelligence data — SIGINT, ELINT, and telemetry. Because my exposure to GMAIC was much more prolonged than that on the SIGINT Committee, the larger part of the discussion which follows relates to GMAIC.
One of the first things I did after I became Chairman of GMAIC was to study the committee's charter. There I saw that "GMAIC shall coordinate guided missile and astronautics intelligence production activities of the government." Here was not just prestige, but power! But it didn't take too long to realize that in spite of all those fine words in the charter, my real responsibility was quite limited. I could bark a lot, but biting was not allowed.
Actually, this lack of authority should not have been surprising. Each member of the committee was in the pay of a different agency or department of the government. He looked to his own organization for direction and career development. His work on the committee was usually a collateral assignment, not to be confused with his real job. Even the CIA member could take issue with the chairman, and indeed he sometimes did so. In the face of such nonexistent command authority, it was really remarkable that a respectable amount of positive work was in fact done by the USIB committees.
[Top of page]
Theoretically, each USIB committee has members representing the same organizations as are in the USIB. Actually, there are differences. In the case of GMAIC, the members for each of the service intelligence agencies are voting members, whereas in USIB, the service representatives are observers. (As chairman, I soon discovered that on issues brought to a vote, the service members tended to follow DIA's lead, but that if they were asked to vote before the DIA man, they tended to vote more independently. The seating around the table, going clockwise, was always Army, Navy, Air Force, DIA, CIA, State and NSA. Therefore I always took votes clockwise.) GMAIC also had a man from the National Aeronautics and Space Administration as a non-voting participant. The SIGINT Committee also had voting service members, and a representative of the National Reconnaissance Office as a non-voting participant.
Since the AEC and the FBI are members of the USIB, they exercised their prerogatives and named individuals to the committees. But neither the AEC nor the FBI man ever attended a meeting. And it was the sensible thing for them to do, because we almost never had items on our agenda which involved them. (Toward the end of my tenure as GMAIC Chairman, I did get to meet the AEC member — the committee sponsored a junket to Charleston, South Carolina, to see Polaris and Poseidon missiles and the submarines which carry them. The AEC man "activated" his membership temporarily, and joined us on the trip. He turned out to be a likable person, and took the ribbing he received from the other junketeers with good grace. But we never saw him again.)
To paraphrase Orwell, all members of the committee were equal, but some were more equal than others. The kind of representation differed dramatically from one organization to another. Some were senior officials with easy access to their USIB principal and commensurate authority, but others were very junior officers who were primarily note-takers and message-passers. Some organizations were represented by men who had been on the committee for more than a decade, but other organizations chose to rotate the membership at frequent intervals. It is also unfortunately true that sometimes totally unqualified individuals with no prior experience in the intelligence business were assigned to the committee. All of this made for a mixed bag. At no time was there a true team effort encompassing the whole group. On the other hand, it would be fair to say the quotient of intelligent, informed, and active members was always high enough to permit the committee to discharge its responsibilities reasonably well.
The caliber of the secretariat for a committee probably is as important to the success or failure of a committee's efforts as the capability of the chairman to lead or the quality of the membership. This was particularly true in my case. Running the USIB committees was a part-time job for me, and having a capable person who could spend as much time as was needed to do all the staff work connected with committee affairs was indispensable.
[Top of page]
The main work of GMAIC is to produce intelligence, and within this category the key job is to make inputs to the national estimates devoted to strategic, military, and technical matters. For many years now, these estimates have gone along two tracks. First, each USIB member agency is free to send its own estimate input to the drafting team assigned to an estimate. Second, GMAIC can provide its own estimate input to the drafting team. There is an apparent redundancy here, but it seems worse than it is. This is because GMAIC does not write a complete estimate; rather, it tries to identify those areas within the estimate where there are likely to be different judgments. It examines the evidence and tries to resolve the differences, or failing that, it tries to describe their nature as lucidly as possible.
Within GMAIC, the actual work on estimates is done in standing subcommittees devoted to particular substantive areas — Soviet offensive weapons, space, ABMs, etc. In these subcommittees, called working groups, the issues are discussed and documented by working level analysts. This interchange at the working level is, I think, the best way to achieve interagency coordination on intelligence production. Granted, there are many issues which fail to be resolved. But even in those cases the fact that there is a difference is disclosed early in the estimate process. The specialists who are most familiar with the intelligence data have a chance to determine if all parties worked with the same data and how the various analytical groups came to differing conclusions.
If the above process works properly, then by the time the representatives of the USIB principals meet to consider a draft estimate, confusion is minimized, and the task of preparing the estimate for submission to the USIB principals is made easier. Of course, sometimes the system doesn't work properly. Data may arrive too late for incorporation at the early draft phase, or a senior official may choose to change an estimate judgment at the 11th hour, for reasons which may have very little to do with intelligence. An interagency committee like GMAIC can do little in such situations — perhaps the new system of National Intelligence Officers will be effective in handling such last-minute panics more smoothly.
Collection requirements also absorb a fair part of GMAIC's time, in two ways. First, GMAIC is called on to identify intelligence gaps and to suggest collection efforts which might fill them; and, second, it is asked to evaluate the contribution of various ongoing collection programs to the solution of its problems.
The tasks undertaken run the whole gamut from trivial ones to those having a major impact on the country's collection efforts. For example, described below is the committee's response to the Soviet ABM problem. Because of a poor collection posture, there was in the mid-60's great uncertainty as to whether a particular major new Soviet development program was for an ABM or for an air defense missile. Resolution of this uncertainty was desperately needed by U.S. defense planners, and later on by our SALT negotiators. USIB asked GMAIC to study the problem and to recommend new collection programs to fill the gap. The committee did so, using not only its own resources but consultants drawn from outside. It delivered an exhaustive report containing a number of recommendations for collection, some of them with very major cost impact.
There followed in the late 60's a burgeoning of expensive collection programs whose primary rationale was the Soviet ABM problem. All this didn't happen just because of the GMAIC report — the whole community was by then sensitized to the ABM problem and there was a general consensus that something had to be done. But the GMAIC report was at least the key document used by managers of the new collection programs to justify their systems to the budget people.
Multi-million-dollar technical collection programs take many years to go from the design concept stage to an operational system, and by the time they were in use, the ABM/Air Defense Missile question they were intended to shed light on had pretty well been answered by less exotic methods. And, as it turned out, the community wound up in the 70's with a capability to collect data on second-generation ABMs and on new air defense systems which was infinitely superior to the collection posture a decade before. But one can question whether the money was well spent. A case can be made that GMAIC delivered up collection recommendations with not enough regard to their cost impact, and that in their panic to do something about the ABM problem the collection system managers brought out some systems of questionable cost-effectiveness.
In addition to its work on estimates and collection problems, there are many lesser matters which require the committee's attention. For example, GMAIC provides the mechanism by which the U.S. exchanges missile and space intelligence XXXX XXXXXX XXXXXX. This is done through periodic workinglevel meetings on a variety of subjects. The scheduling of the meetings is done by a three-member steering group consisting of the GMAIC Chairman and his counterparts from XXXX XXXXX XXX.
These international meetings are a tender topic. There is no denying the fact that because of the huge size of the U.S. intelligence community compared to that of XXXX XXXXX XXX it is unlikely that in any exchange we will get as much as we give. For this reason it has been easy for people to snipe at the meetings by suggesting that their primary benefit is the chance for the participants to get trips XXXXXX XXXXXX XXX at the government's expense.
But I believe that an objective study would show that the U.S. is ahead of the game. Keeping XXXX XXXX XXXX analysts informed of activities here allows them to channel their work into more productive areas. And, in spite of XXX XXX XXX XXX intelligence groups, they keep coming up with nuggets in the form of new unique analysis which have been very helpful to us.
On the ludicrous side, I came to dread those sessions devoted to nomenclature. Most people are unaware that GMAIC is the authority responsible for naming such beauties as the SS-X-18 ICBM. In the example given, it stands for the 18th surface-to-surface ballistic missile system brought out by the Soviets, with the "X" signifying that the program is still in the experimental test phase. Sounds simple, doesn't it? But I remember very well the interminable wrangling we got into the day we decided to attach names such as CSS-1, CSS-2, etc., to Chinese missiles. One member concluded we were obliged to rename all the Soviet missiles SSS-1, SSS-2, etc. Another member wanted CCSS-1 and CCSS-2, so that it would be clear that they were Communist Chinese missiles and not those owned by the Nationalist Chinese. The problem here was compounded by the fact that even the committee members who were ordinarily passive at our meetings became instant and vociferous experts when it came to choosing names for missiles.
So much for GMAIC's activities — not a complete catalog, but rather a representative sampling. The time involved in these affairs was usually a halfday a week for the meeting itself, and perhaps another eight hours a week devoted to committee-connected matters. It was time worth spending, not only to discharge the statutory responsibilities of the committee, but for some intangible benefits as well. The intangible benefits derived from the fact that the committee provided a fairly informal but systematic mechanism for passing information on and raising questions about missile and space intelligence across agency boundaries, at a number of levels in the hierarchies of these agencies. This observation is equally true for the SIGINT Committee and I suppose the other USIB committees as well. Given the present structure of the intelligence community, composed as it is of a group of autonomous agencies, it is hard to visualize how else interagency coordination and information exchange can take place except through committees such as USIB's.
[Top of page]
The SIGINT Committee
The work of the SIGINT Committee is devoted primarily to providing guidance to SIGINT collectors. Some of the other things the committee concerns itself with are the evaluation of collection programs, sanitization and decontrol problems, and recommending policies for cooperative SIGINT collection programs by other countries. The greatest share of the time is devoted to communications intelligence (COMINT). In fact, in its early years, virtually all SIGINT collection was COMINT. This situation has changed with time, so that nowadays collection of emissions from foreign radars (ELINT) and foreign instrumentation signals (primarily telemetry) have become equally important.
Looking at the work content of the SIGINT Committee is not the whole story, because one needs to understand the environment in which these affairs are conducted. The National Security Agency (NSA) is a huge organization, and it is singularly preoccupied with the collection and processing of SIGINT. The NSA member represents a production organization, while the other members represent user organizations. NSA, for the most part, tries to avoid explicit direction of its work. It prefers to get tasking in the most general terms possible. The users — State, CIA, DIA, and the Services — would like to be as restrictive as possible, and tend to write tasking documents in great detail. There results a tug-of-war between NSA and the other members over many of the issues discussed in the committee. Usually the committee resolves the issue by doing that which committees do best-finding a compromise solution acceptable to both sides.
There is a similar relationship between the representative of NRO and the committee members. NRO is a huge organization, too — not in numbers of people, but in the fact that it exercises great control over a large share of the intelligence budget. Overhead sensors used to collect SIGINT are procured with money doled out by NRO to the agencies managing the particular sensor programs. There results a different sort of tug-of-war, in which NRO tries to get the committee to specify requirements for SIGINT data in general terms. But the user organizations would like to know how NRO proposes to satisfy their requirements, and in so doing inevitably get involved in the details of the design and capabilities of the satellite collectors. Some of the committee members represent both user groups and satellite project management teams, and this causes additional trauma.
One of the most difficult documents for the committee to get out is its annual statement to the USIB of the 5-year guidance for the SIGINT portion of the national reconnaissance program. There is no requirement that the 5-year guidance take into account the availability of funds to satisfy the requirement, and if the committee chose to be completely irresponsible, it would merely put out a "wish list" of all the things it would be nice to get. Fortunately, the committee has tried to inject some realism into the process, by doing such things as putting requirements in priority order, identifying those requirements whose satisfaction would result in changes in the national reconnaissance program, and estimating their cost impact. Even so, the process is not without defects. It has not been possible so far to interleave COMINT, ELINT and telemetry needs into a single priority list. This makes for great difficulty for those who have to make choices between costly overhead collection sensors which specialize in one or another form of SIGINT.
An even more fundamental problem is the one alluded to earlier, the lack of a system for matching the value of intelligence to the cost of its collection. The SIGINT Committee, in doing the 5-year guidance, has problems in making authoritative choices of the most cost-effective programs. It must screen proposed requirements collected from analytic organizations, and it has no guarantee that they have been submitted with a consciousness of their cost impact. The degree to which the final document is useful to the policy level therefore is very dependent on the maturity and good judgment of the committee members.
[Top of page]
Why Not Do Away with the USIB Committees?
Alvin Toffler in his book Future Shock concludes that bureaucracy is on the way out, and is being replaced by management by "adhocracy." Similarly, it has been said that the USIB committees should be abolished and the work be done instead by ad hoc committees convened to handle specific problems. There is a certain attraction to the idea. In recruiting people to serve on an ad hoc committee, care could be taken to see to it that each individual had the necessary expertise to contribute to the issue in question. Also, agencies having no interest in the issue would not be burdened to provide representation.
It is probably true that we are likely to see ad hoc committees used more and more as time goes on. They should be particularly useful for tackling major issues on which the community is split. I also believe, however, that there will always be a need for standing committees like GMAIC and the SIGINT Committee. I submit that there are a number of positive aspects to these committees which are often overlooked, as follows:
1. A useful mechanism is provided for passing information on particular topics across agency boundaries, and at all levels.
2. A wide variety of problems which range from trivial to moderately important are handled routinely. In general a standing committee provides a more effective way to handle these problems than does the creation of a fresh ad hoc group to deal with each new problem.
3. Ready access is provided for user organizations to air their particular concerns. Conversely, the matters being discussed in committee are routinely reported back to the user organizations who thereby have a chance to object to actions with which they disagree. Both ways, a mechanism exists to keep things from falling between the stools.
4. The USIB committees have an institutional history which provides a useful background against which to view incoming action items. Members of ad hoc groups run the risk of not knowing how related issues were dealt with in the past.
Another way to look at the performance of the USIB committees is to focus on the really egregious problems. From my vantage point, the issue which has plagued the community for years is the problem of getting better collection and processing of radar and optical data from missile reentries in the Pacific. Others have told me that the community's approach to the problems of narcotics and terrorism has been chaotic. The thing these subjects all have in common is that none of them are within the charter of any existing USIB committee. What this says is that however poorly the committees perform, in those areas where there are no committees, things are worse.
[Top of page]
What Should Be Done to Improve the Committees?
I think that the key to more effective work by the USIB committees lies in improving the quality of the representation. Ideally, each agency should pick representatives who are senior enough to have a good rapport with their principal in the USIB, and who are expert in the affairs under study by the committee. If this were done, it would follow that the men on the committees would be representatives in fact as well as in name, and many issues could be resolved in the committees without recourse to the USIB itself.
Also, I believe that wherever possible the members ought to be people whose regular jobs are in line organizations. Obviously, full-time members could devote more time to committee affairs, but this advantage is outweighed (at least in my mind) by the bureaucratization which would take place with a committee composed of members with no other jobs.
It is clear that the future should also see changes in the responsibilities of the committees themselves. It seems logical that the SIGINT Committee and others concerned with collection will evolve into bodies having closer links to the Intelligence Community Staff and the Intelligence Resources Advisory Committee. Similarly, GMAIC and the other committees concerned with substantive intelligence should evolve into bodies more capable of assisting the National Intelligence Officers. Stress should be placed on the word "evolve." I believe it would be a mistake to wipe out the existing committee structure and replace it with an entirely new one. The losses which would follow from ripping out all the present wiring might be a good deal more severe than would be obvious to the casual observer.
[Top of page]