Collaboration in National Security: Myths and Reality

June 2009 No Comments

Collaboration in the National Security Arena: Myths and Reality – What Science and Experience Can Contribute to its Success.

Author | Editor: O’Connor, J. (DHS), Bienenstock, E., et al. (NSI, Inc).

The purpose of this compendium of white papers is to explore various perspectives on the state of the art in our understanding of collaboration, including insights on the key factors that influence the who, what, when, where, and how of this topic. Collaboration traditionally refers to multiple people or organizations working towards common goals, but there are many other perspectives and definitions. The objective of this compendium is to identify and discuss the issues:

how analytic tradecraft can be enhanced through collaboration
when expansion of access to information take place and if this approach adds value to analysis
how to facilitate collaboration within and across government organizations
who collaborates and how do they collaborate to identify emerging threats
what can be done to improve analysts’ ability to understand and apply social and behavioral science methods and findings.

The basis for all assertions will be given from both scientific and practical bases and areas of dissent and debate will be noted in the papers.

By way of background, this compilation was created after completing a Strategic Multilayer Assessment (SMA) effort during 2008 to develop approaches to anticipate rare events such as the nexus of terrorism and WMD. That effort highlighted the fragility of the models and the need for a multi-disciplinary, multi-agency approach to deal with anticipating/forecasting, detecting and interdicting such events. That effort led to the following:

Publication in November 2008 of a white paper entitled “Anticipating Rare Events: Can Acts of Terror, Use of Weapons of Mass Destruction or Other High Profile Acts Be Anticipated? A Scientific Perspective on Problems, Pitfalls and Prospective Solutions”
Development of a concept for an Inter-Agency Limited Objectives Experiment (IA LOE) as described in the current white paper

This collaboration compendium is published as an adjunct to the aforementioned experiment.

In the months after the tragic events of September 11, 2001, it was discovered that indicators were there which could have led to the prevention of these terrorists’ acts. The 9/11 Commission Report, in looking at this issue, subsequently recommended “Unity of Effort” and a focus on Information Sharing. As we have thought through how best to move from a “need to know” to a “need to share” system, those human issues which contribute to the current “need to know” system have not changed. What has changed, however, is our understanding of human organizing processes and collaboration technologies.

This compendium of papers illustrates that theory, research, and applications are available for enabling collaboration. More importantly, collaboration technologies are now shaping organizing processes – whether our policymakers use them or not. These papers illustrate the breadth of issues involved in institutionalizing the concept of sharing that we now call collaboration. For readers new to this topic, the papers are ordered to minimize the time it will take to gather a working knowledge of the concept of collaboration, what the key constraints and enablers are to collaboration, and what potential paths forward entail.

Section one focuses upon Agency and Operational Perspectives. McIntyre, Palmer and Franks (Section 1.1) quote the President’s Memorandum for Heads of Executive Departments and Agencies on Transparency and Open Government, issued 21 January 2009, which states “Executive departments and agencies should use innovative tools, methods and systems to cooperate among themselves, across all levels of Government, and with nonprofit organizations, businesses, and individuals in the private sector.” McIntyre, et al., bring to our attention the Director of National Intelligence (DNI) Vision 2015 highlighting the need for establishment of “a collaborative foundation of shared services, mission-centric operations, and integrated management…”

The next two papers illustrate the military and law enforcement perspectives on collaboration. Harm and Hunt (1.2) note that collaboration is not a new thing for the military. In fact, Goldwater-Nichols empowered collaboration across defense agencies. The current generation of young military looks forward to their joint assignments. Harm and Hunt focus upon recent advances and evolutions in technology, culture, processes and people driving the current effort to create effective collaboration. Two interesting themes are now starting to emerge: collaboration is defined differently depending upon the culture of the organization, and, there is a need to start small with limited collaboration elements in order to build a functional and effective complex collaboration effort.

Kiernan and Hunt (1.3) point out the nexus between criminality and terrorism. The lessons law enforcement has already learned, as well as the tools applied to defeating social networks of criminals, are also applicable to the military’s fight against terrorism. The authors point out two successful collaboration environments – InfraGard and Defense Knowledge Online (DKO).

The last paper in this section is by Hilton (1.4). He starts with a compelling example of the benefits derived from collaboration enabled by technology during the crisis in the Republic of Georgia. Out of the lessons from this effort arose the concept of Teams of Leaders (ToL). ToLs are high-performing leader-teams whose members are from different organizations, cultures, agencies, or backgrounds and who each bring specific knowledge, skills and attitudes to the cross-culture JIIM leader-team. Components of ToL are Information Management, Knowledge Management and Leader Teams. The synergy amongst these three elements results in high performance. A theme that emerges in this paper, and throughout this compendium, is the idea that the least understood element of collaboration is the human element. The struggle for any organization is not information technology or knowledge management capabilities, but the identification and understanding of the human element in order to effectively apply them.

Section two of this compendium provides a scientifically based understanding of collaboration across multiple disciplines. Bienenstock, Troy and Pfautz (2.1) take on the unwieldy task of providing an overview of perspectives on collaboration. What comes out clearly is that there is a wide range of research, stretching across many disciplines in the area, but almost no overlap. Management and Social Sciences research have primarily investigated social structures and incentives that encourage or discourage collaboration. Computer Science research has focused on teamwork through technology. Additionally, computer design researchers have found that individual, dyadic, and group brainstorming should be encouraged, as well as cognitive conflict.

Research in the military and intelligence communities examine specific physical, virtual, and cultural structures that impact collaboration. The authors identify four critical questions for collaboration: 1) What kind of collaboration is required to meet goals; 2) What barriers exist in status quo; 3) What actions must be taken for facilitation; and, 4) what systems will best enable the actions. Bienenstock, Troy and Pfautz echo Hilton’s discovery: It is the people element that creates the dilemma for effective collaboration.

Next, Kuznar (2.2) notes the anthropological truism that humans are a social species and are interdependent upon one another for goods, services, security, and emotional support. He describes kin-based sodalities (collaborative societies) and non-kin based sodalities. Another theme emerges, which actually runs through all these papers: non-kin based sodalities are often voluntary associations that people create around some purpose. “The fact that voluntary associations are formed around a common purpose indicates that mechanisms of reciprocity are central to uniting a collaborative society…” Quid pro quo is a very old concept and is actually a reasonable way to organize.

Cronk (2.3) adds to the importance of this theme in his paper addressing an evolutionary perspective of collaboration and cooperation. Concepts such as kin recognition systems, cheater detection mechanisms, cooperator detection mechanisms, sensitivity to audiences, reputational concerns, coalitional awareness and theory of the mind suggest that human cognitive abilities may be the product of Darwinian selection in favor of cooperation.

Bienenstock and Troy (2.4) look at collaboration in terms of two basic dilemmas: social traps and social fences. Research is mature on social dilemmas and some findings echo those discussed throughout their paper. For instance, persistence and repeated interaction lead to emergent understanding of a shared fate and, eventually, trust – which contributes to eliminating both social traps and social fences. Also, network structure affects efficiency and promotes feelings of efficacy and a motivation to collaborate.

In the next paper (2.5), Stouder examines organizational studies. The progression of papers in this compendium illustrates that collaboration is studied from many different perspectives and is called many different things. Terminology aside, there is much science has to offer in guiding how information sharing and collaboration can be maximized. Studies have examined interactions and outcomes based on activity at individual, group, organizational, societal, national, and international levels. While the underlying intent of the studies may be to understand how to get people to work together/collaborate, how the research is implemented can result in findings that cannot, or should not, be compared.

Generalizations concerning collaboration must begin with a norming process on the terminologies and definitions. Just because performance on an assembly-line in Michigan increased when lights were added does not mean that it was the lights that increased production (those social scientists among you will recognize the reference to the Hawthrone studies). The problem of the third variable is very real. Empiricists like to get results based on manipulation of facts. However, there are times when the environment in which the empirical assessment is being made changes, and it becomes obvious that what was thought to be causing an outcome was really due to some third variable. Understanding a desired outcome via theory is definitely a more time intensive process, but when the health of entire societies may be on the line, the effort and thought required to test theory is more likely to lead to a consideration involving a rare event such as 9/11.

Stouder’s knowledge of the organizational research literature is a key place to start for the Limited-Objective Experiment (LOE) accompanying this compendium (see Article 4.1). Stouder provides a list of research questions that Bienenstock et. al., began; and authors of other papers add to it. For instance, what is the research seeking to understand – the process of collaboration (type, level, frequency, duration, intensity, variety)? Or, should research focus on the drivers or constraints on collaboration (environmental factors, organizational factors, events, etc.)? The quid pro quo theme emerges again. It appears that asking “What’s in it for me?” is a principle of human behavior as it applies to collaboration.

Heuer, Pherson and Beebe discuss analytic teams, social networks and collaborative behavior in the next paper (2.6). The rising use of Wikis and other collaborative software is building a more transparent and collaborative analytic environment. Hunt (2.7) looks at what can be learned from economics. Economics of engagement indicate that fun, trust and honor are critical components for collaboration success.

The last paper in this section (2.8) presents a Seven-Layer Model for Collaboration. This model is grounded in theory drawn from multiple disciplines. It represents the most comprehensive approach to laying out a means to test concepts of collaboration discovered during our compiling of this paper. Briggs’ Seven Layer Model begins with Goals and moves through Deliverables, Activities, Patterns of Collaboration, Collaboration Techniques, Technology and the Script Layer. Later in the compendium (Articles 4.6 & 5.3) a network architecture is described and it should not go unnoticed that this seven-layer theory and the layers of the mission fabric approach together make a good foundation for future theory development and empirical research.

Section three of this compendium addresses Common Barriers to Collaboration. Rieger (3.1) calls out imbalanced empowerment and accountability as key barriers. Regulatory and legal concerns play roles in making it hard to collaborate. A basic sense of fear of loss also plays a role. Empowerment is determined by whether someone has enough time to do their work, has the training to do it, has the materials and equipment, has open communications, and management support. If a worker puts any of his or her resources into performance, he/she is going to want to know there will be an acceptable form of reciprocity.

Heuer and Beebe examine Small Groups, Collaborative Pitfalls, and Remedies next (3.2). Small groups have been studied extensively across many domains. There are some basic principles of small group behavior which occur regularly (groupthink, polarization, social loafing, etc.). Heuer and Beebe point out that techniques have been developed which stimulate productive group behavior working with tendencies such as those listed above to improve performance. Palmer and McIntyre (3.3) make observations from the trenches about how to build a collaborative culture. The key challenges in building collaborative culture involve processes, technology, and behaviors. Again, the need for incentives for collaboration is noted.

Section four addresses What Applied Research has Learned about Collaboration. Numrich and Chesser (4.1) provide more detail on the Limited Objective Experiment (LOE) mentioned earlier and explain how the effort is embedded in the deeper need to understand and predict rare events. The LOE is designed to enhance existing analytic capability with new collaboration strategies and tools to make the process transparent to strategic decision makers.

The LOE has two parts: a Worldwide Rare Event Network (WREN) experiment and a companion US Air Force Academy (USAFA) experiment. In the WREN experiment a diverse community will attempt to characterize indicators of illicit terrorist activity against the US in a scenario developed by the FBI. Metrics collected during the experiment will include, but not be limited to, where players seek information, to whom they reach out for collaboration and how often, and what tools they tend to use. The USAFA experiment will involve a range of ages and experience (cadets and students plus law enforcement professionals) and the tools used in the second week will permit more visual interaction to measure whether that interaction enhances collaboration. This second experiment will make use of the mission fabric approach described in later papers (4.6 & 5.3).

Pherson and McIntyre (4.2) describe the Intelligence Community’s (IC) experience with operational collaboration. A great example of where new collaborative technologies are embraced is senior leaders who have started their own blogs. Another theme found across papers is perhaps best described here – there are explicit penalties for sharing information too broadly, including loss of employment, but no comparable penalties for sharing insights and information too narrowly. The idea that new collaboration efforts should start with small problems before they are applied to ‘life or death’ projects is brought up by Pherson and McIntyre. Key enablers to successful collaboration identified by the authors include consistent policies, technical and administrative infrastructure, engaged leadership, and use of collaboration cells.

Carls, Hunt and Davis (4.3) discuss Lessons about Collaboration in Army Intelligence. This is the first time that the importance of physical layouts has been specifically noted. It is also the first time that the trend for humans and computers to share reasoning workloads is noted (an element of the dilemma noted earlier involving the human element). Machines require explicit instructions in order to execute tasks involving collaboration. As such, when humans collaborate with machines, one side of a complex collaboration effort is held constant. This type of man/ machine collaboration may provide a base from which collaborative training for man/ man could begin. Further, broader understanding has often emerged from a leisurely stroll around a library or book store. Computers cannot “do” creativity, but humans have workload issues. Collaboration among humans alleviates some of the workload issue, but how do we move to a multi-faceted collaboration where the best of human groups can be brought out?

Meadows, Wulfeck, and Wetzel-Smith (4.4) next explore complexity, competence and collaboration. Identification of factors and collaboation support system design guidelines related to complexity dovetail with Bienenstock’s earlier discussion of the issues involved in researching collaboration. This paper serves as a great means to help merge the Briggs Seven-Layer Model and the framework presented by Pierce for collaboration engineering. The collaboration development process described in the paper is an excellent example of where collaboration started small and how it grew.

Lyons (4.5) provides an excellent review of empirical studies done by the Air Force looking at organizational collaboration and trust in team settings. Four dimensions of collaboration at the organizational level were found using a factor analysis: collaboration culture, technology, enablers (e.g., training), and job characteristics. Other findings relate to structures, processes and reward systems promoting information sharing via IT systems; importance of workspace design (physical layout) in information sharing; importance of individual agreeableness to perception of trust; and, negative communication decreasing performance.

Bergeron and Pierce, in paper 4.6, suggest creation of a means to instantaneously distribute and modulate control of information flow when dealing with security concerns of governmental organizations. Boehm-Davis (4.7) brings a wealth of research from the Human Factors Engineering (HFE) literature to light. HFE has worked to develop safe and effective performance over the last several decades to understand how man and machine have interacted for decades. Boehm-Davis shows there is a flow of knowledge needed to develop effective collaboration where group processes are understood, and also describes how those processes affect work performance and what the nexus is with technology. For example, studies show that if a procedure is put in place, certain aspects of team dynamics can be improved (e.g., checklists used by medical doctors). Boehm-Davis, taking an approach similar to Briggs Seven-Layer Model, also highlights the need to develop both vertical and horizontal models.

Moreno-Jackson (4.8) approaches issues of collaboration from the pragmatic point of view: sometimes the only way to get good collaboration is to involve a non-biased facilitator. MacMillan (4.9) raises the question of whether there can be too much collaboration. She provides findings from a 13-year study conducted by the Navy on Adaptive Architectures for Command and Control (A2C2). One of the key findings is that there is an optimal level of organizational collaboration and coordination for best mission performance – a level sufficient to ensure that mission tasks are accomplished, but not so great as to generate unnecessary workload. Further, she illustrates that it is possible to optimize organizational structures to achieve superior performance even when the number of times humans collaborate decreases. Boehm-Davis notes that the group dynamics literature has found that greater negotiation amongst group members leads to more adaptive groups over the long run.

There are definitions of collaboration throughout this compendium, but most include a need for cooperation towards a goal. The Navy research supports this definition by suggesting that if the goals for collaboration are well-understood and the mission well-specified, only then will there be better mission performance.

Linebarger’s paper (4.10) follows-up on the Navy findings by flipping the thought process around and suggesting that task-focused collaboration can be made more effective, especially if group collaboration patterns are recognized and explicitly supported by the surrounding environment and software system. Linebarger notes that collaboration always occurs if there is dialog. His research suggests that collaboration support improves quality and productivity, especially when the group has some control over how they are supported. Hall and Buckley (4.11) provide a delineated checklist for successful collaboration that has been used in the intelligence community to evaluate collaboration projects.

Section five of this compendium addresses “Potential Enablers for Collaboration.” Bronk (5.1) begins by identifying issues in management and sharing using techno-collaboration. He identifies three core principles for IT in collaborative government work: 1) collaboration tools should be easy to use, 2) collaboration tools should be entirely facilitated by the Web browser, and 3) collaboration solutions should be cheap, or even better, free, as far as users are concerned. Bronk suggests that if talented collaborators are cultivated and rewards systems put in place, appropriate technical tools will be found. He also notes that the quid quo pro in government is tied to the appropriation process and constitutional authority, thus explaining why IT adoption is sometimes difficult.

The next two papers address how IT might overcome some ‘people’ issues. First (5.2), Hunt and Snead discuss “Collaboration in the Federated Environment: The Nexus Federated Collaboration Environment (NFCE).” Essentially, the NFCE serves as a virtual social networking place that transcends the center and edges of its member networks yet facilitates members linking up when they have common specific goals. Interactions among any number of governmental organizations on any number of levels are enabled. The nation must enable people, information and processes to build, explore and exploit a networked federation of diverse organizations so that they can be easily aligned to make timely, transparent and collaborative decisions about adversary goals and behaviors. Again, it is important to put in place mechanisms to reward contributions to collective success.

Pierce (5.3) discusses a multi-dimensionality of collaboration. He describes the interplay between distributed collaboration, security, alignment and provisioning of services. He uses numerous helpful analogies to suggest there should be a paradigm shift in how information is shared and controlled technically. He calls this new approach the mission fabric.

Wagner and Muller (5.4) emphasize that trust occurs between two individuals not an organization. They draw from Gallup’s research to identify eight elements of collaborative success: complementary strengths, common mission, fairness, trust, acceptance, forgiveness, communication, and unselfishness.

Similar to the need sometimes for facilitators, noted by Moreno-Jackson, Pherson (5.5) suggests the use of Transformation Cells to institutionalize collaboration. This approach agrees with several other authors’ observations that there must be a group with skills appropriate for the technologies, processes and behaviors needed, in order to enhance collaboration. Gershon (5.6) examines the scientific research on workspace design and provides blueprints for what might be the most effective designs for collaboration. He also provides illustrations that help the reader understand the key design issues.

Pherson provides another paper (5.7) that illustrates how to develop training courses on collaboration. The approach is comprehensive, noting differences in training requirements dependent upon where a collaborating analyst is in his/her career. He makes a strong argument for joint training because it enables analysts to build teams and networks, develop realistic incentives and metrics, and generate new collection strategies.

Von Lubitz (5.8) discusses the concept and philosophy behind development of Teams of Leaders (ToL) for complex defense and security operations. ToLs were developed because of the need for soldier-leaders who were flexible, adaptable, versatile, and comfortable in operating within the complex setting of Joint Inter-agency, Inter-government, and Multi-national (JIIM) operations.

Finally, Egan and von Lubitz (5.9) discuss the need to include lawyers in groups responding to crises. The Model State Emergency Health Powers Act (MSEHP) helped to organize trans- boundary issues associated with such events as the H1N1 public health emergency, but was more often used to focus on technical challenges. As policymakers worked through crises using MSEHP it became evident that the causes of poor responses were, actually, legal challenges. Three such issues were state sovereignty, definition of response roles, and respect for the federalist process. ToLs, discussed in earlier papers (1.4 and 5.8), which include lawyers, are a means to broaden leadership and to establish a decision-making base that spans the traditional agency and level-of-government boundaries, and generates a collaborative response. The authors emphasize that conflicting laws, jurisdictional domains, and the fear of litigation are present in every decision. As such, it makes sense to add the justice system into all the action already being handled by executive and legislative means.

As described in the 9/11 Commission Report, effective collaboration is a must if we are to prevent other such tragic events. This compendium takes a significant step toward integrating information from many different disciplines and environments in order to develop a field of research on collaboration. Through a more in-depth, empirically-based understanding of the issues, human collaboration can drive the development of new and/or improved technologies and organizational structures/processes.

What can happen if government information holders collaborate? The events of September 11, 2001 illustrate what can happen if they don’t.

Contributing Authors

Jennifer O’Connor (DHS), Chair, Elisa Jayne Bienenstock (NSI), Robert O. Briggs (UNO), Carl “Pappy” Dodd (STRATCOM/GISC), Carl Hunt, (DTI), Kathleen Kiernan (RRTO), Joan McIntyre (ODNI), Randy Pherson (Pherson), Tom Rieger (Gallup), Sarah Miller Beebe (Pherson), Keith Bergeron (USAFA), Elisa Jayne Bienenstock (NSI), Deborah Boehm-Davis (GMU), Robert O. Briggs (UNO), Chris Bronk (Rice), Kerry Buckley (MITRE), Joseph Carls (ret), Nancy Chesser (DTI), Lee Cronk (Rutgers), Bert Davis (ERDC), M. Jude Egan (LSU), Justin Franks (ODNI), Nahum Gershon (MITRE), Tamra Hall (MITRE), Col Craig Harm (NASIC), Richards Heuer, Jr. (Consultant), LTC Brad Hilton (US Army), Carl Hunt (DTI), Kathleen Kiernan (RRTO), Larry Kuznar (NSI), John M. Linebarger (Sandia), Joseph Lyons (AFRL/RHXS), Jean MacMillan (Aptima), Joan McIntyre (ODNI), Brian Meadows (SPAWAR), Victoria Moreno-Jackson, (Nat’l Assoc for Community Mediation), Gale Muller (Gallup), S. K. Numrich (IDA), Jennifer O’Connor (DHS), Douglas Palmer (ODNI), Stacy Lovell Pfautz (NSI), Randy Pherson (Pherson), Terry Pierce (DHS & USAFA), Tom Rieger (Gallup), Ned Snead (IDA), Michael Stouder (GWU), Kevin K. Troy (NSI), Dag von Lubitz (MedSMART), Rodd Wagner (Gallup), Sandy Wetzel-Smith (SPAWAR), Wally Wulfeck (SPAWAR)