From Parameters, Autumn 2000, pp. 66-76.
Go to Autumn issue Table of Contents.
Go to Cumulative Article Index.
The Joint Staff knows why interagency coordination does not occur. According to a Joint Staff memorandum, "in the past it has been extremely difficult to achieve coordinated interdepartmental planning" for two reasons: other agencies of the US government do not understand "systematic planning procedures," and each agency has its own approach to solving problems. The State Department, for example, values flexibility and its ability to respond to daily changes in a situation more than it values planning, while the CIA is reluctant to coordinate for security reasons and the former US Information Agency held Defense and the CIA at arm's length for fear that it would be seen as a mere dispenser of propaganda. If we are to have interagency coordination, the memorandum warns, "these inhibitions of other governmental agencies must in some way be overcome."
Furthermore, the Joint Staff has been aware of the obstacles to interagency coordination for a long time. The memorandum just quoted was written in 1961. This long-standing concern helps explain, no doubt, the worry expressed when the military contemplates the future. A report on Army After Next (AAN) experimentation comments that "the diversity of the interagency, with each agency having its own culture, hierarchy, bias, misperceptions, and unique perspectives, makes unity of effort difficult." These problems, compounded by "low technical and procedural interoperability, and the absence of a common vision," create "formidable obstacles" to interagency coordination. Both Joint Vision 2010 and Joint Vision 2020 make similar points about the interagency process, leading to a similar conclusion: interagency coordination is hard to achieve.
The military is worried about what has become known as "the interagency" partly because of the demands that a Revolution in Military Affairs (RMA) would place on the interagency process. If an RMA occurs in the years to come, it will be led by advances in sensor, communication, and information processing technology that will allow our forces to see and attack the enemy throughout the depth and height of the battlespace. That the enemy will see our forces almost as well as we see theirs will require that we attack from dispersed or remote positions with stunning speed. Knowledge and speed will allow us to achieve "the near simultaneous application of combat power against key elements within the enemy's entire zone of operations." The result will be the disintegration of the enemy's forces. Such disintegration is more likely to occur or to be more valuable strategically and politically if we can enter the theater of operations before the enemy has established himself fully or achieved his objectives. Such rapid strategic maneuver may result in strategic preclusion, denying the enemy the operational opportunity or objective he seeks. If we can demonstrate such a capability, we may even deter conflict.
The stringent demands for coordination that an RMA will place on US government agencies is most obvious at the strategic level. Preclusion will require interagency decisionmaking that can coordinate a rapid response to emerging problems, as well as keep pace with developments in the battlespace. Quick, flexible interagency coordination will also be necessary at the operational level. Rapid simultaneous engagement of the enemy will not always result in the simultaneous cessation of all hostilities. Disintegration may induce some of the enemy's forces to surrender, but others will fight on in isolation as cohesive units, perhaps retreating to nearby urban areas, while others transition to guerrilla warfare. The military, therefore, will be conducting high-intensity operations in one spot, while in other places it mops up, provides humanitarian assistance, takes care of refugees, and implements the transition to a legitimate civilian authority, in these latter cases working closely with other agencies. The traditional sequential approach in which, in one general officer's words, the political guys hand off to the military guys and you "let the military guys do it, and as soon as the military thing is over, immediately turn it back to the political guys" is unlikely to work.
An RMA will increase the importance of interagency coordination for other reasons as well. A force that can disintegrate its opposition will, we hope, discourage enemy forces from taking the field. Our opponents, therefore, will seek to attack us asymmetrically. An RMA will encourage a trend that is emerging already: the likelihood of attacks on America. The RMA quite possibly will join drugs, disasters, and terrorism through mass casualties as reasons why the US military will increase its involvement in contingency operations in support of domestic civil authorities. Success in such operations will require coordination with a host of agencies, as well as state and local authorities. Our knowledge and the speed with which we act may also encourage the use of another asymmetric strategy. Discouraged from taking the field, our opponents may hide in urban areas. In some cases, we may be able and willing to leave them there unmolested or to make sure they do not escape while we take care of more important priorities. Regardless, we will have to fight in the battlespace they choose. Again, this may require that the military execute different military operations (humanitarian assistance, peacekeeping, combat) simultaneously rather than sequentially. Just like combat in the environment we prefer, combat in urban areas will actually consist of a range of operations, most involving interagency coordination at the operational level. Finally, although they are seldom mentioned in discussions of the RMA, operations other than war (OOTW) will continue after the revolution and will remain essentially interagency operations. Since a growing number of US government agencies (currently 30) are operating overseas, a trend that does not seem likely to reverse, the interagency aspects of future OOTW will be at least as difficult as those we are conducting now.
The first conclusion reached when considering the RMA and the interagency is that the operations an RMA will make possible and necessary will not achieve their most potent form unless the interagency process can meet the demands of revolutionized military forces. If it does not, we are likely to get what we got in Panama: an operation dictated by gross failures to coordinate among agencies, which destroyed efforts to oust Noriega without force, followed by a post-operation transition to civilian control so bungled by poor coordination that it almost undid the dearly bought benefits of the military operation. The future may be even worse than Panama suggests, given the greater presence of civilian agencies in military operations after the revolution. This future will be our fate, military discussions of the revolution suggest, unless significant changes are made to a process that is too cumbersome to be efficient and too unaware to be effective. If the interagency process is not reformed, the revolution may come, but knowledge and speed will be undone by ignorance and sloth.
The AAN report suggests ways to move from current interagency problems to future success. The AAN report argues,
An entire network of effort is required in the future to achieve interoperability with respect to planning, organization, communications, standards, procedures, materiel, etc. An adequate physical and virtual infrastructure for interagency planning, training, and readiness activities should be established. . . . A long-term collaborative effort will be required to establish the common doctrine, process, and infrastructure (physical and virtual) needed for effective interagency training. Such an effort may require the development of a National Interagency Training and Readiness Center over the next decade or so.More succinctly, but along the same lines, the standard briefing for JV 2010 requested a "common operational picture; improved doctrine, training and education, leader development programs" as well as other unspecified repairs. The Joint Staff offers the network-centric linking of US government agencies as an important part of the solution to interagency problems. As the AAN report notes, improving the interagency process will place significant demands on DOD resources and time, but the alternative will be a "dangerous period of trial and error" when conflict comes.
Is It Really Ignorance and Sloth?
Clearly, from the military perspective, the interagency is in bad shape. But is this an accurate view? Two cases, countering the terrorism sponsored by Libya's Muammar Qadaffi and ejecting one group of oligarchs from Haiti, suggest that it is not entirely in bad shape and that the interagency in its current form has strengths the military does not always recognize or appreciate.
For over 25 years, the United States has pursued a variety of methods to restrain Qadaffi. It imposed a growing list of sanctions, isolated him diplomatically, bombed his country, and indicted his accomplices. None of this was orchestrated by a centralized or integrated authority. Through administrations of both parties and a constantly changing cast of agency representatives, each method was applied, evaluated, and refined or abandoned and another brought forward and tested through the work of a flat, flexible, adaptive network of agencies. The same interagency process that failed with regard to Panama worked with regard to Libya. Not only did it work, but it worked because the interagency acted like a network, precisely the kind of interagency that JV 2010 hopes we will have someday in the future.
The interagency also has worked under the demands of a crisis. Having learned from the failures of Panama, largely at the urging of the Department of Defense, the interagency put together a deliberate planning process for Haiti. While the ultimate worth of the operation may be questioned, there is little doubt that from the interagency perspective this operation was better planned than Panama. Problems existed, but they were recognized and dealt with. Perhaps the major reason this process worked, however, was not that the interagency was allowed to operate as a network but that its hierarchy was reinforced. Another layer, an Executive Committee, was inserted into the interagency bureaucracy.
This brief review of these two cases suggests three things. It suggests that the interagency process may not be as incompetent as the military believes. It also suggests that neither increased networking, recommended by the Joint Staff and others, nor improved integration, also recommended by the Joint Staff, is necessarily the way to improve the interagency process. Networking worked against Libya, a more integrated hierarchy against the oligarchs. Finally, it suggests that before we attempt to reform the interagency process, we take a step back and make sure we understand what it is.
What is "the Interagency"?
To begin with, there are three characteristics of the interagency process that must be kept in mind: it is a network disguised as a hierarchy; it incorporates different decision modes and speeds; and it has horizontal and vertical dimensions.
Network and Hierarchy
On paper, the interagency is a hierarchy. The President is at the top; Cabinet officials preside below him in a Principals Committee; below the Principals sits the next rung of their department or agency bureaucracies in the Deputies Committee; below the Deputies, an Executive Committee sometimes operates, composed of officials drawn from the next lower rank or two of the bureaucracy; and below the Deputies or, if it exists, the Excom, interagency working groups carry on, composed of officials drawn from the next levels of the bureaucracy. In theory, direction and information flows down the hierarchy and information and policy options ready for decision flow up. In practice, however, there is very little that holds this hierarchy together. Ultimately, only the President has authority, and he has competing priorities.
Beyond the most general policy orientations, which are sometimes given at the top, issues are determined by a slowly established consensus. Participants in the process consult with one another, other components of their agencies, which may have different views, and often with Congress and outside experts, who lobby the agencies. Slowly, agreement and coalitions build. Informal personal contacts are important in all this; sometimes they are decisive. On some issues, decisions are in effect made at the working group level and only ratified, if noticed at all, by higher-ups.
The hierarchical structure of the interagency, then, has always been more apparent than real. It becomes more real than apparent only when the President focuses his attention on a problem or, what often amounts to the same thing, when that problem assumes political significance for the President. Since the number of issues that become politically significant are few compared to the array of issues dealt with, the interagency is for the most part a real network and an apparent hierarchy.
That the interagency process is both a hierarchy and a network may strike
some as an acute example of its incoherence. In fact, however, this dual
nature is an important source of strength, because neither hierarchical
nor network organization is necessarily always best. Each has strengths
and weaknesses, which makes each better or worse for different functions.
Figure 1 provides a rough guide to these strengths and weaknesses.
Hierarchies can make decisions more quickly than networks because authority is centralized. For the same reason, they can implement their decisions more quickly and reliably. This means that they are better at managing risk in the short term than more decentralized organizations, whose members are free to respond to local conditions and thus are hard to guide through difficulties. On the other hand, this freedom to respond makes decentralized networks more sensitive, agile, and adaptive than hierarchies. The members of a network can change individually to meet changing circumstances without waiting for a headquarters to tell them what to do. If some of these changes prove successful, then other members of the network may adopt them if they fit their local circumstances. Such experimentation permits adaptation. Such adaptation, in turn, means that networks tend to manage long-term risk better than hierarchies.
As this summary of strengths and weaknesses makes clear, hierarchies and networks are best suited for different activities. As the example of Haiti suggests, a hierarchy may be best suited for crisis management, while, as the campaign against Qadaffi implies, networks may be best for protracted problems in changing environments. It would be best for our national security if the interagency process maintained both of these organizational characteristics.
Different Decision Modes and Speeds
Thinking of the interagency process in terms of the different functions that hierarchies and networks best perform draws our attention to the fact that the interagency process consists of two different decision modes--crisis management and long-term planning--that have different requirements and occur at different speeds. The demands of crisis management do not allow the deliberate pace of planning, just as the quick decisions of crisis management would be inappropriate for planning, where the slow process of consensus-building can allow new ideas to emerge and prepare the way for effective implementation.
We can think of these different decisionmaking modes and speeds in terms of our current national security strategy, which organizes our efforts into three categories--shaping the world, responding to problems, and preparing for the future. Shaping and preparing are long-term efforts. It is probably best if the interagency's network character dominates in these activities. This is so not only because these aspects of our strategy unfold slowly but because the diversity of the world we live in and the rapidity with which it seems to be changing call for an organization that permits flexible and adaptive responses. Responding, on the other hand, because it most often requires decisiveness and speed, is an activity in which the hierarchical character of the interagency process should dominate. Efforts to reform the interagency process must ensure that the different modes in which the interagency operates are addressed.
Horizontal and Vertical Dimensions
Most discussions of interagency coordination talk about coordination in either Washington or the field. This might be termed horizontal coordination. But the interagency process also incorporates or should incorporate coordination between Washington and the field. This vertical dimension is critical, particularly when managing contingencies. Such vertical coordination is most often thought of as Washington conveying clear direction to the field. But it must also work the other way. Operational possibilities and costs must be conveyed from the field to policymakers, because such considerations should be weighed as policy is made. Those in the field are best positioned to understand these possibilities and costs, and they should engage in a dialogue with decisionmakers in Washington. All of the operations envisioned in discussions of the RMA will require this kind of interaction.
The State Department's vision of future operations includes involving its embassies in formulating policy. This is a step in the right direction, although there is no suggestion in State's vision that this involvement will escape the stovepipe that connects embassies with the Department in Washington. The military, on the other hand, has not accepted this vertical dialogue, at least with civilians, as part of its future operational style. Perhaps it believes that the speed of its operations will make such dialogue impossible. If so, this belief contradicts the central contention of the RMA that speed will be essential for effective operations. If operations are too fast for coordination with policymakers, then they will be ineffective, no matter how successful militarily, because they will unfold before policy can properly shape them. Worse, operations may present policymakers with faits accomplis, and thus determine policy. As it becomes clear that operations have this character, decisionmakers may hesitate or fail to approve operations or may seek ways to slow them down. This already happens when State fears that DOD's planning threatens the flexibility necessary for diplomacy, a fear noted in the Joint Staff memo cited above.
In order to conduct successful operations, the military will need to improve its ability to coordinate vertically. Perhaps the greatest obstacle to such coordination, and for that matter to horizontal interagency coordination as well, is the military's tendency to mask operational capabilities and possibilities from civilian decisionmakers for fear that such knowledge will encourage civilian meddling in operations. If the Joint Staff wants to have effective, integrated, and unified operations in the future, this inhibition will have to be overcome.
Some Questions to Answer, Some Traps to Avoid
The interagency process, much vilified, deserves careful study. Although it has failed on occasion, it is not all bad. On the contrary, it has proven itself flexible, adaptive, and decisive. It was a network before networks were cool, and it has successfully handled hundreds of crises from noncombatant evacuations to war. Because it is both a hierarchy and a network, understanding how it works might even shed some light on how the military might organize its revolutionized forces, which will have to combine both organizational characteristics. If this is too much to hope for, studying the interagency process may at least allow DOD and other agencies to develop a more efficient and effective working relationship, so that the United States, its people, and the other agencies of their government can benefit fully from revolutionized military forces.
Study of the interagency should answer a number of questions. For example, if it is true that the interagency process has the three characteristics distinguished above (a claim requiring further scrutiny), we should ask how we can maximize the advantages and minimize the disadvantages of the interagency's network and hierarchical character. How can we make either network or hierarchical characteristics dominant? Can the communication revolution improve the interagency process in Washington, in the field, and between the two? If the RMA forces initiative to lower echelons, can and should the interagency allow greater initiative to the field? In which activities and operations and to what extent? Should we allow such devolution of policymaking authority? Would it be compatible with the accountability we expect? How would this affect congressional oversight? Should we think in terms of an expeditionary model, where general guidelines are set in Washington but adaptation and implementation are left to the field, or in terms of real-time control exercised from Washington and made increasingly effective through improving communications? Can the interagency community in Washington have the knowledge and experience to exercise real-time control? If the State Department involves Ambassadors in the formulation of policy, would this change how CINCs and their staffs work with Ambassadors and their staffs? In answering these questions, we should remember that the military and interagency need not always exercise the same degree of centralization. For example, the need for greater control of dispersed fire support may actually mean that after the revolution some military operations will be more centralized than they are now. The interagency support for such operations may or may not need to be as centrally controlled.
In addressing these and other questions, there are a number of traps that experience with the interagency process suggests we should avoid both now and as we work to integrate revolutionized forces into the interagency process. When the military discusses the interagency, there is a tendency to see interagency problems as analogous to problems between the services. It is suggested, for example, that just as jointness has replaced service rivalries, so can interagency coordination replace interagency squabbling. The problem with this analogy is that military officers from different services have much more in common than do representatives from different agencies. Compared to the differences between, say, a CIA case officer and a State Department Foreign Service Officer, the differences between an Army officer and a Marine Corps officer are negligible with respect to "culture, bias, misperceptions, unique perspectives," and respect for hierarchy. This means that whatever interagency coordination may be, if it is achieved, it will be unlike what the military has experienced as jointness.
Not only will interagency coordination and jointness be different, they should be. It would be a fatal error to seek interagency coordination by centralizing or homogenizing the interagency process by, for example, creating an interagency national security civil service, establishing interagency "czars," especially if they have budget authority, or pushing for too much unity and integration. Because their functions are different, military officers, spies, diplomats, and lawyers see problems and their solutions differently. Not one of these different approaches is expendable. To succeed, the US government needs them all and needs vigorous advocates for each. The best way to increase interagency coordination will be the one that promotes coordination while respecting these differences and enhancing their forceful expression.
In seeking models for effective coordination that do not overwhelm agency differences, we should consider the interagency effort to combat terrorism. In this case, a small group, a kind of standing Executive Committee, was critical in extracting maximum effectiveness from agency differences. Coupled with the success of more ad hoc Executive Committees, the example of the small counterterrorism group may lead us to think that such small groups are the best way to ensure effective interagency coordination. Such groups do have virtues. A small group of experts, focused on a single issue, meeting frequently enough to develop cohesion and trust, especially if membership remains stable, is likely to be more effective than a large interagency working group that meets infrequently and whose members are responsible for a number of different issues. While small standing groups can be effective, however, they cannot be more than a limited solution to problems of interagency coordination because the critical ingredient of their success is in limited supply. The success of the counterterrorism group varies in direct relation to the support each member of this group enjoys from his Cabinet-level superior. In the mid to late 1980s, for example, this support was high and gave the members of the counterterrorism group clout individually in their agencies and collectively in the interagency. Since the time of senior officials is so limited compared to the number of issues they must contend with, a small-group approach to interagency coordination cannot be used for more than a few issues. Small groups may improve coordination on these issues, but they are not a device that will improve interagency coordination generally.
The military and other operational agencies (the CIA and the FBI) may favor keeping interagency groups small in hopes that this will allow the speedy decisionmaking and security that operations require. It would be a mistake, however, to focus only on the speed of decisionmaking or the need for security. If interagency groups are too small, this may impede the full coordination necessary for success. Crisis may justify a bias for speed, but crisis decisionmaking is only one mode of interagency activity. Even crisis decisionmaking may prove to be different depending on the kind of crisis. Some crises--say, a developing complex contingency operation involving the United Nations and a regional political or military alliance as well as the interagency--may well require a slow decisionmaking process. For the shaping and preparing aspects of our strategy, which are essentially planning exercises, speed will be much less important. In structuring interagency working groups, judgments will have to be made on individual cases, weighing the relative importance of speed, security, and inclusiveness.
Perhaps because of the military bias in favor of speed, military discussions of the interagency process tend to emphasize procedural fixes and technology. As noted above, the AAN report speaks of "low technical and procedural interoperability" as a problem, while JV 2020 laments that the interagency process lacks "extensive liaison cells" and "integrative technology." Such technical and procedural fixes will be useful, but perhaps most fully in implementing rather than in making interagency decisions, and in permitting effective horizontal and vertical interagency operational dialogue. Such fixes will be less useful in making decisions because in this case the critical issue is building consensus. Technological and procedural fixes are likely to be limited in their ability to speed up this essentially political process. In this case, therefore, from the military's point of view, the critical element will remain the political skill of the Secretary of Defense and the Chairman of the Joint Chiefs of Staff and the senior members of their staffs.
As the military is inclined to focus on procedures and technology to fix the interagency, civilians may be inclined to look to ideas drawn from management studies of successful corporations. Something might be learned from such studies, but since the profit motive of commercial enterprises is so essentially different from the purpose of the interagency, the utility of corporate management ideas is likely to be limited. For example, many corporations have tried to push decisionmaking down to the level where information is gathered. If efficiency were the goal, this might be a good model for a governmental bureaucracy. Ambassadors, for example, would be given more decisionmaking authority. But efficiency is not the primary purpose of a governmental bureaucracy. To improve the interagency process, it may be more useful to consider ideas drawn not from commercial practice but from organizations in which the usual incentives do not apply. The study of complex adaptive systems or principal-agent relationships may be relevant.
Whether we look to technology, procedures, recent developments in organizational theory, or other ways to solve interagency coordination problems, we probably should not think there will be one or even just two or three solutions. Given the diversity of the interagency and the activities it undertakes, as well as the uncertain and changing world we inhabit, it would be wise to encourage a decentralized experimental approach to improving the interagency. We should permit the development and implementation of a number of different solutions. In keeping with this spirit, we should resist the temptation to implement a massive structural reorganization, a new National Security Act to update the system put in place in 1947. It would be better to build on the flexibility and adaptability we currently possess. This approach would place the responsibility for reform with the individual agencies, limiting the National Security Council staff to the role of a coordinator.
Finally, while the Defense Department, with its operational and long-term focus, may feel more urgency about improving the interagency process than other agencies do, it should proceed carefully. It should remember the practitioner's ready grasp of agency differences expressed almost 40 years ago in the Joint Staff memorandum. At the same time, it will have to overcome the prejudice of that memorandum, expressed in the emphasis on DOD's strength--systematic planning--and the claim that coordination will improve when other agencies overcome their inhibitions. In this regard, DOD should recognize that if it wants to lead the way to better coordination, it will have to explain how that coordination benefits other agencies and not simply why its revolutionized forces require it.
1. Joint Staff Memorandum, "Organizational Aspects of Special Operations" (Draft), 20 March 1961.
2. US Army Training and Doctrine Command (TRADOC), Knowledge and Speed: Battle Force and the U.S. Army of 2025, The 1998 Annual Report on The Army After Next Project to the Chief of Staff of the Army (Fort Monroe, Va.: TRADOC, 1998), p. 8; Joint Vision 2010 Standard Briefing, Part Three, Slide Three, Internet, http://www.dtic.mil/jv2010/briefings.htm, accessed 5 June 2000; Joint Vision 2020, p. 18, Internet, http://www.dtic.mil/jv2020/jvpub2.htm, accessed 25 July 2000.
3. TRADOC, Knowledge and Speed, p. 11.
4. "General Says U.S. Has Learned from Grenada Mistakes," The Washington Times, 11 August 1994, p. 8.
5. TRADOC, Knowledge and Speed, pp. 2, 8.
6. For the failure to coordinate before Operation Just Cause, see Robert Pear and Neil A. Lewis, "The Noriega Fiasco: What Went Wrong," The New York Times, 30 May 1988, p. 1; and Rebecca Grant, Operation JUST CAUSE and the U.S. Policy Process (Santa Monica, Calif.: RAND, 1991). For the failure to coordinate the transition to civil administration, see Richard H. Shultz, Jr., In the Aftermath of War, US Support for Reconstruction and Nation-Building in Panama Following Just Cause (Maxwell AFB, Ala.: Air Univ. Press, August 1993).
7. TRADOC, Knowledge and Speed, pp. 21, 29; JV 2010 Standard Briefing, Part Three, Slide Three.
8. For example, US Department of State, America's Overseas Presence in the 21st Century, The Report of the Overseas Presence Advisory Panel (Washington: Department of State, November 1999), p. 34.
9. Figure 1 is adapted from Francis Fukuyama and Abram N. Shulsky, The "Virtual Corporation" and Army Organization (Santa Monica, Calif.: RAND, 1997), p. 19.
10. US Department of State, America's Overseas Presence, p. 53.
11. The planning that lead to the dispatch of the Harlan County to Haiti was marred by a lack of coordination between State and DOD. Both agencies bear responsibility for this, but State's failure to coordinate with DOD appears to have been motivated in part by fear that DOD's planning process would establish timetables and requirements that would limit the flexibility that State felt its diplomats needed to make the Governor's Island Agreement work.
12. Fukuyama and Shulsky, pp. 51-53.
13. For a description of proposals to update the 1947 act, see Paul Mann, "Info Technologies and National Security Doctrine," Aviation Week and Space Technology, 23 November 1998, pp. 51-54.
Dr. David Tucker is a Visiting Associate Professor in the Special Operations Academic Group at the Naval Postgraduate School. He previously served in the Office of the Secretary of Defense for Special Operations and Low-Intensity Conflict and as a Foreign Service Officer in Africa and Europe. He holds the Ph.D. from the Claremont Graduate School. His publications include "Combating International Terrorism," US Air Force Academy, Institute for National Security Studies (forthcoming); Skirmishes at the Edge of Empire, the United States and International Terrorism (Praeger, 1997); "Fighting Barbarians," Parameters (Summer 1998); and "Responding to Terrorism," Washington Quarterly (Winter 1998).
Go to Autumn issue Table of Contents.
Go to Cumulative Article Index.
Go to Parameters home page.
Reviewed 16 August 2000. Please send comments or corrections to firstname.lastname@example.org