http://www.ewenger.com/pub/pubusfedcioreport.doc

Communities of practice in government:Communities of practice in government:the case for sponsorshipthe case for sponsorship

An executive memowith exhibits

William M. SnyderEtienne Wenger

etienne@ewenger.com wsnyder@socialcapital.com

December 2003

Executive summaryGovernment today faces unprecedented challenges, from rising citizen expectations to an expanding breadth and complexity of problems to address. These challenges require an

increased capability for learning and innovation as well as a scope of coordination that are not afforded by current structures. Creating large consolidated departments such as Homeland Security is only applicable in a few cases; and even in this case, addressing the problem of security still requires learning and coordination with other entities such as local governments and the private sector.

This study, sponsored by the Council of CIOs, reveals that promising new structural approaches already exist in government. They are peer-to-peer networks of practitioners that we call “communities of practice.” These communities cross formal boundaries to bring together practitioners who are facing a common challenge—to learn from each other, to develop new solutions to problems, to find synergies across organizations, and to coordinate efforts. We argue that it is important to learn to recognize these communities, legitimize their work, and cultivate them more intentionally and systematically.

We describe several communities of practice in the federal government. They bring practitioners together within and across agencies, as well as across government levels. And they produce results. The Rumble Strip community has spurred the widespread adoption of highway safety devices that have saved lives and taxpayer money. The E-Regulation community has accelerated the implementation of a cross-agency effort to reduce paperwork. SafeCities has created new partnerships across a range of disciplines to reduce gun violence on the streets. And CompanyCommand has helped Army company commanders take on the challenges of leadership.

We could have described many more, but we thought these four examples would make the point. They illustrate what these communities of practice are, how they work, what value they produce, and what it takes to make them thrive. But the main result of our study—and our main argument here—is the urgent need for executive sponsorship. This was a theme that pervaded all our conversations with community members. Practitioners unfailingly value the opportunity to learn and coordinate with peers, but they believe much greater results are possible with increased support from the hierarchy. The problem is especially acute when a community crosses agency or government-level boundaries because much of the value of such cross-boundary connections shows up outside the purview of local managers. The importance of executive sponsorship for communities of practice in government parallels the experience of leading organizations in the private sector.

This memo is addressed to leaders in agencies, in the Administration, in Congress, and in cross-sector advocacy groups. Community sponsorship should become an essential role of government executives. Sponsorship for community initiatives should be built into legislative mandates and the management strategies of the executive branch. The purpose of this memo is to make the case that results require sponsorship: We need committed leadership to cultivate strong, vital communities of practice; and we need such communities to build and apply the capabilities required now to get results.

IntroductionA recent press report summarized the results of a joint committee of the House and Senate intelligence panels, which found that “the September 11 attacks were preventable, but the plot went undetected because of communications lapses between the F.B.I. and C.I.A., which failed to share intelligence….”1 The report itself went even further, arguing that underlying structural conditions of the various agencies that constitute the “intelligence community” (including NSC, CIA, FBI, DOD, CTC, NSA, and the US Military services) left it unprepared for terrorists attacks. Not only were there communication lapses, but the cross-agency community of intelligence personnel was “fragmented, under-resourced, and under-skilled.” (In fact, the report also found that between federal agencies and local authorities, and even within agencies, major gaps in information-sharing and coordination created significant security vulnerabilities.) The report’s recommendations included eliminating “obsolete barriers to coordination among agencies,” improved terrorism analysis, and better training for agency staff.

The 9/11 disaster demonstrates in stark terms the price we can pay for organizational structures and processes that are not keeping up with the complexity and urgency of the problems we face now. Today’s messy civic problems combined with increasing performance expectations are forcing leaders and citizens alike to challenge basic assumptions about organizing government to achieve mission-driven objectives. The current model just doesn’t seem ready to address the step-change in performance demands we must tackle in the 21st century.

Large and complex problems such as homeland security, education, or the environment require widespread collaboration among various governmental and non-governmental bodies. Citizens expect to have these problems addressed in their terms, not in terms of the division of labor among agencies. The efficient use of taxpayer funds can only be achieved by leveraging synergies and sharing practices across the entire government. New issues arise that require adaptive responses. Addressing these complex problems requires an unprecedented level of learning, innovation, and coordination among large and disparate groups of practitioners. Such collaboration must happen along several dimensions: within agencies and departments whose employees are distributed throughout the nation; across agencies and department boundaries; across levels of government, local, state, and federal; and in partnership with private-sector and non-profit organizations.

The purpose of this executive memo is to outline how network structures that we call “communities of practice” can be used to tackle urgent, complex civic problems. We focus especially on what federal executives can do now to increase the number and influence of such communities in order to strengthen our government’s ability to fulfill its mission.

Structure of the argumentOur argument for executive-level sponsorship for community-based knowledge networks can be summarized as follows:1 Johnston, David, “9/11 Study Faults F.B.I.-C.I.A. Lapses,” New York Times, July 24, 2003, p. A1.

1

The complexity of problems faced by governments today requires an increased capacity for learning, innovation and professional development—at levels beyond what traditional structures alone can provide.

Indeed, no formal structure can fully address problems that are too complex to predict or standardize. Moreover, these problems invariably require a configuration of disciplines and resources that are rarely contained in any one agency, level, or sector. This calls for the explicit cultivation of knowledge-based, boundary-crossing structures such as communities of practice to complement formal agency and program structures.

These communities contribute to the work of agencies in multiple ways, by developing capabilities at both individual and organizational levels, and by coordinating activities across boundaries.

Despite their clear benefits, cultivating communities in bureaucratic organizations can be difficult because the investments of time and effort and the value produced do not fall cleanly within organizational structures.

One of the main lessons learned in a decade of intentional initiatives to cultivate communities of practice in leading organizations is the need for executive sponsorship. Executives need to understand the role of sponsorship and the forms it can take.

Community sponsorship should become an institutionalized dimension of government leadership roles. Sponsorship for community initiatives has to be built explicitly and systematically into the work of government, both in the design of legislative mandates and in the management strategies of White House and agency leaders.

Intended audienceWithout strong leadership from senior government officials, community initiatives will not take root or fulfill their potential for strategic impact. This memo thus targets a number of audiences, including congressional leaders, agency executives, the White House, the Office of Management and Budget, the President’s Management Council, and citizens who ultimately pay the price—as both investors and customers—of government services and legislative policies. These various leadership audiences can be categorized at three levels corresponding to their primary areas of responsibility: those whose focus is agency-specific performance; those with a cross-agency mandate; and those whose interests cross agency, governmental, and sector boundaries. Agency leaders should review their top strategic objectives and identify areas where

stronger cross-boundary collaboration and knowledge sharing—within and across agency boundaries—could significantly enhance results.

Congressional and Administration leaders need to consider how programs will be administered without limiting their assumptions to specific agencies or departments. Formal approaches as well as informal, peer-to-peer mechanisms should be considered to foster coordination across agencies, government levels, and sectors.

Citizens and advocacy groups need to learn more about how organizing approaches—not merely changes in laws, policies, and funding levels—can increase their capacity to share the influence and responsibilities of civic stewardship.

2

It is important for stakeholders in various roles to see the connection between our organizational capacity and our ability to achieve the results we aspire to—whether within, across, or beyond the boundaries of federal government agencies. It is especially important for senior executives to recognize these opportunities and understand their crucial role as sponsors of these initiatives.

Organization of the memoThis memo is organized into five sections addressing the following topics:

1. Four representative communities of practice in the federal government2. The value that such communities can bring to organizations3. The leadership structure required to support communities4. The crucial role of executive sponsors of community-based initiatives5. Recommendations for next steps

To keep this memo short, we have focused on the argument for sponsorship and put more detailed information on cases and concepts in separate exhibits. These exhibits are referenced in relevant sections of the memo and are also listed on page 16.

A community strategy for building government capabilityThe knowledge-based, network structures that government needs today are not new. To the contrary, they have always existed on an informal level wherever practitioners—whether farmers, artists, or engineers—have gathered to swap stories, solve problems, or just hang out together. In recent years, however, these structures have been deployed aggressively and systematically in a growing number of leading global organizations (such as Proctor & Gamble, Shell Oil, McKinsey & Company, the World Bank, and DaimlerChrysler). They are getting particular attention in the private sector where organizations depend on such strategic community initiatives to compete in a hyper-competitive global economy. For them “knowledge capital”—skills, methodologies, and innovation capabilities—drives results. We call these knowledge-based networks “communities of practice” to emphasize the role of practitioners in taking charge of knowledge issues and we believe they provide an essential new structural tool that can dramatically increase government’s capacity to fulfill its civic mission on a number of fronts.

Examples of communities of practice can be found throughout the federal government today. Some are thriving and some are struggling. Some are based in one agency, others span several agencies, and some even cut across levels and sectors, involving state and local governments as well as private-sector and non-profit organizations. Here is a brief description of four representative communities. More extensive descriptions and analysis of each community can be found in the Exhibits section.

An agency-based community: the Rumble Strips CommunityThe Federal Highway Administration (FHWA) had set a strategic goal to reduce highway fatalities by 20 per cent nationwide in the 10 years between 1998 and 2008. Leaders of FHWA’s Knowledge Management (KM) group worked with agency executives to design an initiative to accelerate the diffusion of rumble strips—a proven road-design innovation that

3

significantly reduces run-off-road crashes. Their challenge was to go beyond disseminating information, which was having little influence, towards more compelling ways to engage state safety engineers and policy makers to implement rumble strips as a smart investment that saves lives as well as tax dollars.

Rumble StripsSponsor Federal Highways Administration

Domain Installing rumble strips to prevent highway crashes and fatalities

Members 100+ federal and state agents nationwide, industry, civic groups

Activities On-line access to research, vetted written and video descriptions of methods, directory of members; coordinator for Q&A and expert referrals

Outcomes Acceleration of adoption across states; reduction in highway fatalities See Exhibit 1, page 17, for the full story of the Rumble Strips Community.

An inter-agency community: the E-Regulation CommunityA number of regulatory agencies had to respond to a legislative mandate to offer online access to compliance forms for their customers. Bill Bennett, the person leading this initiative at the Federal Energy Regulatory Commission (FERC), decided to seek out his counterparts at other agencies to explore best practices for meeting that mandate. The Federal CIO Council was aware of his activities and through its Knowledge Management Working Group invited Bennett to coordinate a pilot community of practice involving a number of federal regulatory organizations. This community would serve as a forum for learning from each other, finding synergies, establishing standards, and coordinating with key entities such as the National Archives and Records Administration (NARA).

E-Regulations/E-RecordsSponsor Federal CIO Council

Domain Developing electronic compliance and records management systems

Members Professionals in IT, KM, and Records Management from: FERC, DOD, GSA, SEC, NRC, IRS, NARA, Interior, Agriculture, Transportation

Activities Face-to-face meetings, sharing agency approaches, joint project

Outcomes Faster learning about methods across agencies; phase one of an initiative to establish on-line access for citizens; new community to design e-records system

See Exhibit 2, page 20, for the full story of the E-Reg community.An inter-level, inter-sector community: SafeCitiesReducing gun violence on the streets of America is a goal that is shared by many governmental and non-governmental organizations in cities across the nation. The National Partnership for Reinventing Government (NPR) office responded to a report from the Attorney General’s office that highlighted an opportunity for cities to learn from

4

each other about how to reduce gun violence. A core group of managers from several federal agencies convened a community of practice that included practitioners from federal agencies, mayoral offices, and local law enforcement agencies, as well as citizen groups, faith leaders, school administrators, business executives, social workers, and others. The SafeCities community fostered learning across its member cities as well as collaborations among players at local, state, and federal levels.

SafeCitiesSponsor Vice President’s Office/NPR, Departments of Justice and Treasury, White

House Office of Management and Budget

Domain Reduce gun violence nationwide

Members Justice (mostly the Office of Community Policing), Treasury (mostly ATF), local law enforcement, social services, faith leaders, etc.

Activities Teleconferences, listerv, website, visits, face-to-face conference, projects

Outcomes Joint efforts between police and faith leaders that spread innovative policies and practices to new cities; and increased collaboration among ATF, FBI, US Attorneys, and local law enforcement

See Exhibit 3, page 25, for the full story of SafeCities.

An online distributed community: CompanyCommand.comNothing can fully prepare new company commanders for the responsibilities they must take on during their two-year command. When two commanders who were friends and neighbors realized how much they benefited from their frequent conversations about their challenges, they decided to find out if others would benefit just as much. They created a forum for new and seasoned commanders to share their insights and experiences; lessons learned; and tools and methods. The result was one of the most successful online communities in military history.

Company CommandSponsor Self-sponsored initially; now sponsored by the United States Military

Academy at West Point

Domain Professional development for company commanders

Members Thousands of company commanders worldwide, (including future and former as well as current commanders)

Activities On-line access to tools, stories, videos; coordinator team for newsletter, book program, local forums, and Q&A

Outcomes accelerate “time-to-talent” for isolated leaders in demanding roles See Exhibit 4, page 29, for the full story of Company Command.

Understanding the value of communities of practiceThese four examples show how communities of practice complement formal units to help organizations weave critical connections across formal groups and leverage knowledge

5

for performance. Through shared commitment to a domain of knowledge that members care about, a community of practice generates “social capital” that enables new levels of collaboration: for creating collective knowledge, for developing members’ skills, and for coordinating actions across boundaries.

Building, sharing, and applying capabilitiesIn an era where problems are complex, urgent, and constantly changing, organizations cannot rest on laurels of size, market share, or institutional monopoly. Sustained performance is driven by organizational capability—including expertise, methodologies, and professional relationships. Communities of practice help organizations essentially by building, sharing, and applying capabilities. Indeed, strategic capabilities can only be kept up to date and disseminated if there are appropriate learning mechanisms in place—and these need be predominantly informal, as true expertise cannot be codified in a formal class or procedural manual. The structures that support such informal learning are ones that connect practitioners who share passion for an area of expertise—not necessarily people in their same team or organization. Community members are connected by trust and reciprocity, which promote mutual commitment to informal, peer-to-peer learning.

Communities build or adapt the methodologies, skills, and relationships that enable organizations to solve problems. Practitioners talk together about problems and innovative approaches and in the process develop new ones. They may organize projects to build a new tool—as did SafeCities practitioners regarding a tool for assessing a city’s capability to reduce gun violence.

Communities of practice share capabilities by connecting practitioners and engaging them in peer-to-peer action-learning activities. For members of the E-Reg community, this kind of sharing was a key to accelerating the implementation of their mandates and avoiding learning the same lessons over and over. Professionals these days are generally overwhelmed with the amount of information produced in their area of expertise. Member relationships provide a network for finding out quickly which information is most important to know about and where to get the knowledge you need when you need it.

Community members apply their practice day in and day out. Collectively, they are a living repository for ideas, information, best practices, directories of experts and resources, and the rest of the requisite repertoire that civic leaders will need. As illustrated by the Rumble Strips community, this knowledge is not simply stored in manuals; it lives among actual practitioners, who are engaged in applying its components to solve problems and explore possibilities that can be adapted to their own situation.

See Exhibit 5, page 35, for a description of the structure of communities of practice that enables them to act as capability-building entities.

See Exhibit 6, page 37, for a description of the ecology of activities typical of the learning of communities

See Exhibit 7, page 40, for a description of the stages of development of communities of practice.

6

Attracting, retaining, and developing talent Communities also build the organization’s capability by providing a context for practitioners to develop new skills and relationships with peers. Every agency in the federal government—like organizations nationwide—is faced with a demographic time bomb that threatens to de-commission nearly half their employees between now and 2010. Again communities of practice are called to play a key role. An informal sense of belonging among practitioners and associated opportunities for professional development are the hallmarks of organizations that attract, retain, and develop top talent. In governmental institutions, participating in a community can nurture or reawaken the sense of personal calling that inspired many employees to choose a career in public service. As CompanyCommand.com demonstrates, communities provide an ideal context for experienced practitioners to pass on their expertise before they leave—or to stay in the loop as alumni who provide “on call” support when the community’s active members need help.

Enabling coordinated action across boundariesLast but not least, communities of practice provide a mechanism for coordination across boundaries in ways that complement formal structures. When you can put a human face on agency bureaucrats and experts who have joined your community, you are more likely to collaborate with them, discuss the problems you face, and find ways to act in concert. For instance, the E-Regulation community discusses standards that should be adopted by all agencies so they can coordinate software purchases as well as exchange and archive records of transactions. In a community, the impetus for coordinating actions is rooted in members’ passion for their domain. They seek to act in concert because they care about what they do, not merely to comply with a top-down directive. This produces informal acts of trust and reciprocity that are at the heart of any well-functioning bureaucracy—and which cannot possibly be specified in systems and procedures or extrinsically enforced by authority. There is of course a place for formal, hierarchical coordination mechanisms. But informal, peer-to-peer collaboration among capable, responsible practitioners is extremely efficient. Moreover, it provides much-needed flexibility to deal with unique or urgent situations that demand on-the-spot decision-making and creativity. Unlike a policy that forces all localities to comply with one way of doing things, a community of practice enables members to compare, learn from, and adapt what they do to fit specific situations while providing coordination among locales.

See Exhibit 8, page 42, for a discussion of ways of assessing the value of communities

See Exhibit 9, page 45, for a discussion of how to “go to scale” to increase impact

7

A key to success: multi-level leadership structure The representative cases we described provide striking illustrations of how communities can bridge agency, level, and sector boundaries to get results not otherwise possible. Yet, like any innovation—organizational or otherwise—there are few willing to be pioneers, and there are as many failures as successes in the early going. In fact, in several of these cases and others that we have documented, the communities have closed down or significantly reduced their level or scope of activity. If communities are so useful, then why do even successful ones struggle to survive? What are barriers to sustainability and effectiveness—and what can we do about them? Several characteristics of the government environment make the cultivation of communities particularly problematic.

Administration and agency leadership turns over at regular intervals, and as a result, organizational development investments are limited. Federal employees are therefore somewhat skeptical about the long-term viability of initiatives like communities of practice.

Agencies tend to act as silos and avoid investments of staff time to help others. The silo mentality—reinforced by formal systems, budgets, and rewards—makes it difficult to see the opportunities that cross-agency collaborations create, much less make commitments to act on them

Bottom-line financial savings or increased value created for citizens are simply not compelling when they are only measurable by adding up the impact across agency lines—which means that community members working horizontally often get little visibility or management support.

In a highly politicized context, agencies may pay lip service to compliance and collaboration, but remain reluctant to revealing what they are really doing.

Innovations that do not correspond to a legislative mandate or executive order tend to be ignored. For instance, a community about “performance-based contracting” was highly successful because it was aligned with a mandate requiring that agencies use this approach in a growing percentage of their contracts. By contrast, a similar community about “reverse auction” contracting never really took off because there was no mandate requiring agencies to use reverse auctions to select vendors.

The key to addressing these obstacles is an appropriate leadership infrastructure that can guide, support, and renew the community initiative over time. In every case we are familiar with, the critical success factor for community participation and effectiveness is leadership. There are several key leadership roles—at both community and organizational levels. Three of these are especially crucial: community coordinators for each community, a support team for a community-based initiative, and an overall executive sponsor. It will be easier to understand the central role of the executive if we consider this role in the context of its relationship to the community coordinator and support team roles.

8

Community coordinatorsAt the community level, the “community coordinator” can make or break the success of a community. Community coordinators nurture the community by orchestrating activities, connecting members, shepherding initiatives, and helping to solve problems. Ideally, for an active community in a strategic area, the coordinator is half-time or more. Michael Seelman, coordinator of the SafeCities community, was nearly full-time, and moderating the start-up phase of the Rumble Strips community took more than half a person’s time. Bill Bennett had less time to devote to the E-Regulation community, but his availability and dedication were nevertheless instrumental to the community’s progress and accomplishments. It is not only the time and commitment of coordinators that are crucial, but their level of technical and interpersonal competence. As one cross-agency community member said:

“There has to be someone whose sole responsibility is to keep the network going, to bring the issues to the table…. It has to be someone with connections, with access to decision makers who will return their calls and respond to requests that come to the table.”2

It is not necessary to be an expert in the field, but it helps to know enough in order to appreciate who should be involved, who should talk to whom, and to have legitimacy with members who feel it is important to know the business.

See Exhibit 10, page 48 for the portrait of a community coordinator.

The support teamAnother important role is the “support team.” It supports the launch and development of communities of practice in an organization by providing services at an overall initiative level (which may include anywhere from a few to hundreds of active communities). Support team functions include educational activities; initiative planning and coordination; coaching for community leaders; managing infrastructure (especially technology); and acting as a liaison among communities and with sponsors to facilitate ongoing learning and alignment. In all organizations where there is a significant community-of-practice initiative that is getting results, inevitably there is also a strong support team. Mike Burk and colleagues at the FHWA provided the support-team function out of their Knowledge Management office, and they have helped launch a number of communities beyond the Rumble Strips Community. The E-Regulation community and another one on Performance-Based Contracting received support from the GSA team, and SafeCities was supported by a strong support team at the NPR office, which also served several other communities. Community-of-practice initiatives are still very new for organizations, and there is much to learn in the early going. Support team staff can help the organization start further up the learning curve and improve the success rate of communities that produce benefits for both members and the organization.

See Exhibit 11, page 51 for the portrait of a support team leader.

2 A community member, as reported in Snyder, W.M. & Briggs, X. de S., 2003. “Communities of Practice: A New Tool for Managers.” Arlington, VA: IBM Center for the Business of Government, http://www.businessofgovernment.org .

9

Executive sponsorshipOur consulting experience and case research over that last ten years has repeatedly reinforced the importance of executive-level sponsorship—particularly when the initiative is designed to address strategic objectives or cross conventional organizational boundaries. Executive sponsors are generally required for communities to have the requisite staff time and organizational legitimacy to fulfill their potential. Sponsors must also be careful to staff and support those who act as support team professionals. The support team role requires expertise in a number of disciplines and enough savvy to liaise among sponsors, communities, and line managers.

Communities of practitioners exist in all organizations. The question for sponsors is how important it is to cultivate key communities to address strategic objectives that cannot be achieved otherwise. The sponsor’s investments, guidance, and legitimacy are crucial to both support team and coordinator roles—and therefore to the success of the community initiative overall. Sponsors are even more important in a government context where the communities’ contributions to bottom-line objectives may be less visible because there are no reliable cross-agency performance metrics—and yet citizens rely more than ever on cross-agency coordination for good government.

See Exhibit 12, page 54 for the portrait of an executive sponsor.

Setting up a stewardship structureBefore launching a strategic community-of-practice initiative, agency leaders should organize a sponsorship and support structure to steward the overall community initiative. The first step is for agency executives to consider how a community-based initiative can contribute to their strategic objectives. This often means convening formal and informal conversations to talk about how communities of practice can build capabilities, and about which capabilities the organization needs to achieve strategic objectives.

Steps for establishing the strategic context and stewardship structure include: 1. Organization and education (informal as well as formal) of a sponsor board that

includes a high-level sponsor and steering committee members who will be instrumental to implementation.

2. Identify where to focus the community initiative—through an executive review of strategic priorities or by engaging a broader group of stakeholders in collective conversations to identify hot issues.

3. Establish key roles and leaders for sponsoring and supporting the initiative:o Identify sponsor board functionso Staff support team to coach community leaders and liaise with sponsor boardo Recruit and develop community coordinators (one or several for a community,

depending on size and intensity of activity and respective time allocation).It is of course not necessary to have all these roles formally established to get started. But when these initiatives are successful, they typically rely on influential and skilled people actively filling community coordinator, support team, and sponsorship roles.3

3 Of course, there are many communities of practice that meet member needs and never show up on the radar screen—and often do not want to. The focus here is on strategic communities of practice taking on important agency and national challenges.

10

The role of executive sponsorSponsorship may be provided by an individual executive, by a sponsorship board, or by a specific executive role such as a Chief Knowledge Officer (CKO). Sponsors play a key role in integrating communities of practice in the organization while preserving their independence as self-governed groups of practitioners. In this sense, sponsorship is quite different from traditional management. Communities of practice do not report to their sponsors in the traditional sense. Rather the role of the sponsor is to connect and protect—connect communities to strategic organization goals and protect them from bureaucratic red tape. It is to give them space to operate and a voice to contribute. The sponsor provides strategic direction, allocates resources, aligns stakeholder support, and fosters the institutional legitimacy of communities. Sponsorship functions include:

Visioning a community-based approach for building capabilities that drive performance and achieve strategic objectives. Such a vision may include a constellation of communities addressing an array of capability areas—as in FHWA where multiple communities are operating today, or in organizations such as Daimler-Chrysler and the World Bank that have over 100 active communities.

Providing policy direction and ongoing reviews (generally annual or semi-annual) to assess community progress and developmental needs—types and levels of participation, accomplishments and results, development issues, and proposals for further development

Encouraging participation among staff

Allocating staff and financial support, including setting funding levels and facilitating investments among multiple agency executives (or via a corporate entity) for communities that cross formal unit boundaries

Championing the initiative among stakeholders, especially direct managers of community members, to build support and foster implementation of community contributions

Ensuring alignment of organizational systems and policies, such as staffing policies, performance reviews, and IT investments

Providing both formal and informal opportunities for managers at all levels to get educated about communities of practice and their role in the organization

Where should sponsorship come from?For cross-agency communities, it appears that a high-status, powerful source of sponsorship is important. This may be an artifact of the Federal Government culture, which generally impedes collaboration across agency silos. Presidential orders and legislative mandates seem to be what’s required to spur initiatives across agencies. The mandate that requires agencies to offer electronic compliance, for example, has helped increase interest and participation in the E-Regulation Community. But that community does not have nearly the level of support that SafeCities had—and if it did, members say they could be making much more progress. For the SafeCities community, the support of the Vice President’s office was extremely important for both symbolic and practical reasons. When an agency official—federal, state, or local—gets a call from the Vice President’s office, they generally

11

respond promptly. Moreover, such a high-level office has greater capacity to facilitate cross agency collaboration and is less likely to take a perspective that is limited by the parameters of agency-specific programs. As one member said:

“The Vice President’s office is better than any agency as a convener, because it’s more flexible to represent people versus low-level agency sponsorship, where it may become focused on their policies and programs; this is more open and has a cross-agency perspective.”4

High-level executive sponsorship makes it easier for managers at all levels to support the participation of their employees. As Mike Burk remarks in a report he wrote about his work at FHWA:

“An even bigger stumbling block is failing to win the support of managers to whom community members report. Usually that lack of support is manifested in the manager’s unwillingness to give employees time to engage in sharing ideas in a community of practice. When workers are busy, management tends to say, ‘We have a lot to do. Just go on to the next item.’ [I]nstead of doing that, managers [should] say ‘You have to make that success available to a multitude of people.’”

There is, however, a downside to high-level support from politicians and appointees—it lasts only as long as the sponsor stays in office. And even when newly appointed officials are sympathetic to the means and purposes of these initiatives, they often do not have sufficient background or experience to continue supporting them. The FHWA case suggests that such initiatives may gain the capacity for sustained influence when sponsorship and support processes become institutionalized in the agency.

Where to go from here?We’ve tried to make a strong case that government leaders can dramatically increase their ability to achieve results by complementing traditional structures with community-based network forms.

The purpose of this memo is three-fold: First, make the case that communities of practice are crucial tools for managing for results in today’s environment; second, describe the role of sponsors and in brief, how to start a community initiative. Third, and perhaps most important, we want to leave senior managers with an interest in learning more—both more about where communities could make a difference and about how to sponsor them successfully. We have kept this memo short in hopes it will be more likely to be read (the exhibits, of course, are quite extensive). But in truth, there is not much point in going into more detail here—the case for communities and the how-to has been extensively described elsewhere.5 More important, the only way to demonstrate the potential of community initiatives, and to learn how, is to do it in practice.

4 Snyder & Briggs, 2003, op cit.5 See Wenger, E., McDermott, R., and Snyder W.M., “Cultivating Communities of Practice,” Boston: Harvard Business School Press, 2002. For an analysis of government-sponsored community initiatives, see Snyder & Briggs, 2003, op cit.

12

Unlike recommendations for massive restructuring, launching a community of practice initiative can start modestly, focusing on two or three strategic domains as a start, and building from there. The up-front investments are minimal—a small number of FTE to provide community leadership and professional support—with several hours a month of attention and support by senior officials (and depending on the situation, there will be travel costs, staffing allocations for active members, and website-development expenses).

Long-term steps: institutionalizing sponsorshipOver time, as communities prove their value and need deeper levels of integration in the organization, it is important to institutionalize sponsorship. This provides symbolic as well as substantive support to knowledge initiatives and ensures their sustainability when current leaders move on. The bottom-line is that strategic communities need serious sponsorship and support and that such sponsorship and support need to be baked into the institution of government:

Legislative sponsorship. When the legislature announces strategic mandates, it should include explicit provisions for sponsorship and resources to support the communities that can build the capabilities and cross-boundary connections the mandate requires.

Executive sponsorship. Agency leaders in charge of implementing mandates also need to provide sponsorship. Their focus is on translating legislative intent into intended program results. Therefore, agency leaders must look to develop communities with sufficient scope and participation—including ones that operate within the agency, across agencies, and in partnership with other government levels and non-government constituencies. It may be necessary to organize a cross-agency sponsor board to provide support and guidance for strategic community initiatives that are focused on cross-agency opportunities—such as the implementation of the E-Government mandate. A way to institutionalize sponsorship inside an agency is to establish a knowledge management office.

Centralized support. A corporate office in the federal government needs to build strong supportive structures to enable the development of communities of practice across the federal government—and particularly for ones that cross agency and governmental boundaries. This includes process and technology support from an office that can serve both community- and initiative-level leaders in all agencies. This corporate support team should also convene a community of community leaders (a “meta” community) so they can learn from each other about the process of cultivating communities.

Legitimate participation. Participation in communities of practice should become institutionalized as an aspect of the job of every federal employee. The point is not to make such participation compulsory, which would be counter-productive, but to create a context in which roles, time commitments, and reasonable expectations of results are well understood and community contributions are recognized.

Visible results. There need to be institutionalized ways to make the results of community work visible, especially when it is not primarily located within one agency. This will enable subsequent initiatives to leverage early successes and spur a more general evolution from strategic pilots to system-wide implementation.

13

Immediate next steps: gathering momentumWe recommend the following set of immediate next steps to provide an initial impetus for cultivating communities of practice in the federal government:

Convene a gathering of agency leaders and other senior executives to discuss this memo—perhaps including members of the President’s Management Council or a similar group.

Educate a core group of senior executives about the approach (with background conversations before the meeting), and create a learning agenda as a sponsor community.

Conduct scenarios or case-study presentations in which a community-based approach is combined or contrasted with traditional project-management and formal restructuring approaches. For example, ask representatives of the Homeland Security Office to lead a case discussion that outlines their organizational challenges and initiatives—and explore opportunities for complementary community-based strategies.

Imagine where such an approach could lead in the futureo Create scenarios for urgent issues such as homeland security and

education.o Help agency leaders think through an application for their agency

(something like what FHWA did), or one that will cross agency boundaries (like the E-Regulation initiative)

o Develop sample project plans, funding, and assessment strategies that agency leaders can use to ballpark what such an initiative in their agency would entail

o As a vision exercise to provide a long-term context for immediate work, project the possibilities for scaling and institutionalizing such initiatives to address an array of relevant issues; within and across agencies; as well as across government levels and private and public sectors.

Facilitate partnerships among attendees to begin working together on real initiatives—and make plans as a group to provide an ongoing context for continued learning together as a community of practice on executive-sponsorship.

See Exhibit 13, page 57, for an analysis of how the E-Regulation community complemented a project.

See Exhibit 14, page 62, for a scenario on how this approach could be applied to Education to complement a program.

See Exhibit 15, page 66, for a scenario on how this approach could be applied to Homeland Security to complement a formal structure.

14

Coda: a vision of governance, and a place to startThe boundary-crossing organizational structures that we describe here serve not only to accomplish agency missions better. In the longer term, they also provide a foundation for a new kind of national governance model that emphasizes participation, inquiry, and collaboration. Imagine a nation of cities and regions in which each city has a thriving constellation of cross-sector groups that is stewarding its own “whole round” of civic disciplines—including education, economic development, health, housing, public safety, and others. Imagine that each local discipline-based group also participates in a cross-city community of practice in an area related to its domain. SafeCities was such a community. Each member coalition was a cross-sector community focused on reducing gun violence at the local level, while SafeCities was a national community that included coalitions from 13 cities; all were focused on learning from each other about gun-violence-reduction strategies. We have seen other national-level communities composed of local communities on topics such as workforce development, school improvement, children’s health, and economic development among others. We have seen international communities composed of issue-based coalitions from cities in various countries—on topics such as reducing risks of natural disasters, juvenile violence, and urban slum revitalization.

Such a vision of community-based, inquiry-oriented governance is not so speculative when we consider that the elements—communities at local, national, and even international levels—have already been developed. Moreover, intentional, systematic efforts to cultivate them have become much more prevalent and influential in just the last five years as their visibility has increased.

The federal government can become an important catalyst for a governance model that envisions the nation as a civic learning system. It can do this by pioneering community-based applications within and across agencies, and by helping to convene intergovernmental and inter-sectoral communities. Again, such an aspiration is not merely hypothetical—the groundwork has been laid; successful pilot efforts have proven the “theory of the case”; and active executives, support team professionals, and community coordinators and members exist in the federal government today. (And we know of similar initiatives in other federal governments—in Canada, Australia, the UK, Singapore, and other countries).

In any case, we have no choice but to discover some combination of new organizing strategies to deal with the problems we face today. We need a step-change in our organizational capacity for learning and collaboration. The complexity of civic problems, combined with increasing performance demands, is likely to become only more urgent in the years ahead.

This memo has outlined an argument for cultivating communities. It has provided a high-level look at how to get started and included exhibits that provide illustrative cases and pointers to useful tools and frameworks. But its essential purpose is not to provide a definitive set of recommendations or an implementation plan. Rather, we hope it will attract a group of leaders with a shared interest in the issues and opportunities it highlights, and spark a conversation about what might be possible and where to start.

15

Exhibits:Exhibits:Table of contentsTable of contents

Detailed information about ideas introduced in the memo can be found in the following exhibits. They are listed under the sections of the memo where they are referenced in the text.

Section Title Page

Examples of communities of practice in the government 3

Exhibit 1 An agency-focused community: Rumble Strips 17

Exhibit 2 A cross-agency community: E-Reg 20

Exhibit 3 A cross-level, cross-sector community: SafeCities 25

Exhibit 4 An online distributed community: CompanyCommand.com 29

Understanding the value of communities of practice 6

Exhibit 5 The structure of communities of practice 35

Exhibit 6 How communities learn: typical activities 37

Exhibit 7 Community lifecycle: stages of development 40

Exhibit 8 Going to scale with communities 42

Exhibit 9 Assessing the value of communities 45

A key to success: multi-level leadership structure 8

Exhibit 10 Portrait of a community coordinator: William Bennett 48

Exhibit 11 Portrait of a support team leader: Mike Burk 51

Exhibit 12 Portrait of an executive sponsor: Shereen Remez 54

Where to go from here 12

Exhibit 13 Communities complementing projects: the case of E-Reg 57

Exhibit 14 Communities complementing programs: the case of education 62

Exhibit 15 Communities complementing formal organizations: the case of Homeland Security

66

16

Exhibit 1 Exhibit 1 An agency-focused community: An agency-focused community:

Rumble StripsRumble Strips

The Federal Highway Administration (FHWA) community initiative was started when Mike Burk, leader of a Knowledge Management (KM) department at the agency, began talking with agency executives about new approaches for achieving strategic objectives. Executives were looking for ways to leverage technology to facilitate dissemination of tools and methods for reducing traffic fatalities. Experience showed, Mike argued, that such technologies are rarely used unless there is an organized community of practitioners to moderate the site, manage the resources, recruit participants, and encourage contributions. Although communities of practice were unfamiliar as intentional organizing approaches, FHWA executives recognized the phenomenon itself—collegial networks wherein professionals keep up with advances and find help for solving problems. Mike got the go-ahead to launch a pilot on an issue where there was an opportunity to make a clear and relatively short-term impact on results.

The community purposeThe FHWA initiative provides a model for how an agency can apply a community-of-practice approach to address its strategic objectives—in this case the goal to reduce highway fatalities by 20% in the ten-year period between 1998 and 2008. The case illustrates how a highly distributed agency can leverage widespread knowledge resources to accelerate the dissemination and application of methodologies that help the agency better fulfill its mission.6 Mike Burk and a team at the FHWA launched a community to help the agency diffuse a proven innovation across all 50 states that had been effectively implemented so far in only 15 and was spreading very slowly to others—despite its compelling record for reducing injuries and saving lives. The community is called the Rumble Strips Community because its domain is about issues related to proposing, planning, deploying, and documenting the beneficial results of rumble strips that alert drivers when they are in danger of running off the road. (Rumble strips are the washboard-like indentations at the side of the road that create a loud noise and palpable vibrations when run over by a car.) The agency had stacks of reports and brochures documenting the research and explaining why, when, where, and how to plan and implement rumble strip installations. But safety engineers at FHWA felt helpless to communicate the information in ways both comprehensive and compelling enough to motivate state-level engineers and decision-makers to act.

Who they were and what they didThe target members of the highway safety community that Burk and his team wanted to reach were a complex constituency. It included state DOT staff engineers from all 50

6 This review of the Rumble Strips Community closely follows a report by Michael Burk, who was one of the principal organizers and who is a member of the Knowledge Management office, which has served as the convener and support team for a number of communities of practice. His report is available at: www.km.gov.

17

states, but also federal safety engineers nationwide as well as highway contractors, governmental decision-makers, researchers, and citizens. The core group of the community consists of federal engineers from the Safety Business Unit of FHWA. Members participated in the community largely on-line. They discussed topics in the Rumble Strip online forum, and they contributed and accessed relevant reports, brochures, and presentations in a shared repository. A directory of community members from government and industry provided information about members’ areas of interest and expertise. Members could keep in touch either by going to the website or by subscribing to the repository and discussions via email.

While the online facility certainly made it easier to access information—the information in the repository was not all that different than the paper brochures and reports the agency had been sending out. The crucial difference was the allocation of a dedicated community coordinator—Jim Growney, a federal safety engineer who was passionate about this topic. Jim facilitated the interactive on-line exchanges, worked with federal engineers on outreach strategies to engage the state-level engineers, and moderated the website to ensure it was up-to-date and useful to members.

Participation by constituencies across sectors and disciplines was encouraged in order to promote broader support for deploying rumble strips, and in hopes of spurring innovative approaches. In one case, a sub-community of cycling enthusiasts became quite active and challenged the highway engineers and policy makers to find ways to make rumble strips less troublesome to cyclists who ride primarily on the sides of the road where rumble strips are located. The exchanges were sometimes heated and it took a while before members on different sides built trust. But eventually the input of cyclists did lead to a new appreciation by engineers of the unintended consequences of the strips. Together they developed ideas for innovations to their design that would mitigate the problems for cyclists.

Results they achievedThe Rumble Strips community accomplished a great deal and the FHWA staff was careful to document results. The number of states routinely using rumble strips has more than doubled (from 15 to 37) since the launch of the community. FHWA staff showed how increased use of rumble strips has reduced the number of highway crashes. They cited research from states such as Wyoming and Pennsylvania to show that rumble strips prevent approximately 100 crashes a year in a typical state, which equals a reduction of about 50 injuries and .66 fatalities. The team cut these efficacy numbers in half to be conservative and then and estimated the benefits in financial terms. The socio-economic savings of implementing rumble strips one year earlier than expected (due to the activities of the Rumble Strips Community) would represent a savings of 2.6 million dollars in associated costs. (Such dollar estimates, of course, cannot adequately measure the costs of human suffering.) This compares to the community’s cost of less than one FTE per year to coordinate all community activity.

Stories from members help illustrate causal links from community activities to results. For example, an engineer with the Wisconsin Department of Transportation explained

18

how the community helped him get better information on implementing rumble strips: “We’d call state by state and see what they were doing. We were also not sure who to contact. This way, the contact information is on the Web site.” According to Burk’s report, an Ohio DOT administrator said “the rumble strip portal helped her finalize plans for a number of specialty highway safety applications. [She stated:] ‘We also used data from the site to help develop Ohio’s current policy on rumble strip use.’”

What kind of sponsorship did they receive?Although the FHWA effort includes an array of players who are not federal employees, the agency was the sponsor and convener for the community. The Rumble Strips Community had strong executive sponsorship from the beginning. Important elements of this sponsorship included staffing allocation for a coordinator, funds to develop a well-designed on-line forum (which was especially important for a group that mostly interacted virtually), and a commitment to FHWA’s Knowledge Management office, which provided coaching, design guidance, and IT support. The KM group also facilitated measurements of the results achieved by the community—documented in Burk’s public report on the community. Making the community’s results visible reinforced the commitment of agency sponsors, who subsequently decided to support an expanded initiative to launch several more communities on related topics such as environment and planning, air quality, high performance concrete, and transportation asset management.

19

Exhibit 2Exhibit 2 A cross-agency community: A cross-agency community:

E-RegE-Reg

One of the mandates associated with the federal “E-Government” initiative was that regulatory agencies offer online compliance processes to their constituencies. William Bennett, Chief Knowledge Officer of the Federal Energy Regulatory Commission (FERC), was charged to fulfill this mandate for FERC. He took the initiative to seek out colleagues at other agencies who were facing the same challenge to learn best practices associated with an agency’s transition to electronic filing. The CIO Council was aware of his efforts and chose to sponsor the E-Regulation Community (or “E-Reg” as it is known to members). E-Reg was selected a a cross-agency pilot community because it addressed a strategic concern that aligned with a new government-wide legislative mandate, the Government Paperwork Elimination Act, which required agencies to streamline processes and reduce paperwork by October, 2003 This mandate, as is often the case, required agencies to build and expand related capabilities in areas such as electronic form design, records management, IT, and knowledge management. Agencies with intensive regulatory record-keeping requirements—such as FERC, Trademark and Patents, Interior, and others—would need to build extensive capabilities quickly. The CIO sponsors wanted to test the hypothesis that a cross-agency community could help agencies both learn from each other and improve inter-agency coordination to reduce costs, align standards, and meet deadlines.

The community purposeThe E-Reg Community of Practice was launched in Fall, 2001 to bring together professionals from various regulatory agencies charged with the mandate to develop electronic compliance systems for their agency’s customers. The members wanted to share experiences from their efforts to implement the mandate, learn from each other, discuss issues, coordinate efforts, and establish standards when necessary.

Issues faced by community members ranged from basic implementation problems, to selection of software and application systems, to advanced questions about leveraging the value of digital records to provide new types of services and analysis. Additional issues included user-friendly design, format standards, and security. Quotes from members (from several agencies, including SEC, FDA, FDIC, etc.) illustrate the concerns they brought to the community:

Internal resistance: “We’ve got a cubic mile of records in boxes, and we’re trying to create a cradle-to-grave electronic system. But we face internal agency resistance, along the lines of ‘if god wanted the Internet, it would be in the 1934 law.’ But our clients are banks, and they are way out ahead of us. We need to catch up if we’re going to achieve our vision of serving clients for ‘all things banking.’”7

7 All quotes cited in the E-Reg case are from interviews conducted for this study.

20

Setting standards: “We’re really in dire straits now and moving to commit to an agency platform. But we’re still not sure what the National Archives and Records Administration (NARA) will require. If they set a different standard than the one we choose now, it could be a catastrophe times 500.”

Security issues: “Security issues could be very dangerous. The worst case scenario is that we don’t have enough data security and this shuts us down from December to March. It’s important to learn much more about security needs and options when digitizing records.”

Adding value: “We have 390 forms that we regulate. Our issues include document format, methods for digitizing, and user interface. We have finished modernization, but now we need to find ways to add value to documents; do trend analysis, and improve our services.”

Who they were and what they didThe community included approximately 30 professionals in various disciplines, including Information Technology, Knowledge Management, Legal, and Records Management. They represented nearly 20 agencies, including FERC, DOD, IRS, GSA, SEC, NRC, NAR, PTO, Interior, Agriculture, DOT, and others. William Bennett was the lead coordinator for the group, which met every two months (with various informal exchanges between) to share experiences, demonstrate tools, talk about standards, and work together on projects to accelerate the implementation of the e-government mandate.

The group met primarily face-to-face because nearly all its members were based in Washington D.C. Typically 15-20 members would come to the meetings. Participation was voluntary and in some cases members came even when their managers discouraged participation—which made the commitment of participants all-the-more noteworthy. In some cases, agency managers were not sympathetic with the notion of spending time sharing ideas and methodologies with a cross-agency group. But participation was high because members found that meetings enabled them solve problems faster and avoid reinventing the wheel or repeating a mistake someone else had already made.

In community gatherings, members talked among themselves and with external stakeholders about critical issues in their domain. One such issue, for instance, concerned the various ways to record a compliance transaction in specific digital formats, and more importantly the digital standards that the National Archive Records Agency (NARA) would require agencies to meet. It was a waste of effort to design an electronic compliance system that would produce documents NARA could not process because they did not meet its standards. Bill Bennett decided to arrange a meeting with some NARA leaders to discuss the issue.

The meeting with NARA leaders was a typical E-Reg community activity. A core-group planned the meeting. They believed colleagues would be interested in coming to such an event—and they were right. They sent invitations to nearly 60 agencies and departments and got a 50% response rate. “50-60 people showed up at our first meeting, representing

21

30-35 agencies.” The focus of the meeting would be to find out more about digital standards that NARA would require agencies to meet. Members had much to gain by finding out what NARA was thinking—and even influencing its deliberations with their practical, implementation-oriented perspective. At that point, there was no concerted cross-agency initiative to address this issue, and no established forum wherein NARA officials could learn from agency practitioners about the complexity of the challenge from the agency perspective.

The records format issue is typical of very prosaic issues (such as the Y2K problem, until just a couple years before the year 2000) that often get overlooked by executives, yet can cause tremendous inconvenience and expense unless they are addressed appropriately (Typically, records management was low on the executive agenda, even in regulatory agencies.) As expected, records management and IT practitioners appreciated how troublesome the issue could become long before agency executives. Moreover, they were in a unique position to address the issues effectively—if given a chance. In fact, it was striking how much initiative the group took with so little formal support or legitimacy. In the end, however, such sponsorship would be essential if the insights gained from the meeting were going to be put into practice.

The community made a particularly strong investment in a project initiative, which they chose strategically as a quick-win initiative that would make the value of the community more visible to executives—and help garner increased sponsor support, in terms of staff time, legitimacy, and influence. The project’s goal was to make it easier for citizens and organizations to access various forms associated with all federal agencies through one central source—a one-stop portal for all federal regulatory forms that citizens could access from the perspective of their problem, without worrying about which agency was in charge of which form.

A core-group of members from seven agencies represented in the community committed to work with GSA—the convener of the FirstGov website—to determine project requirements. Issues included getting OMB approval for forms, creating “meta-data” tags to identify and classify forms, identifying e-form file formats that match the FirstGov search engine, etc. Although the project plan was never implemented, members gained a number of useful insights that were helpful in other contexts: the value of using forms that can be filled out online versus conventional static forms, design principles to make forms easy to fill out, and linking forms that relate to typical customer problems—such as a merger of local utilities that requires filings with multiple federal agencies such as FTC, DOJ, SEC, and FERC as well as state agencies.

As a way to establish the community’s legitimacy, the project presented a chicken-or-egg problem that never got solved: They had hoped that this project to provide online forms via a government portal would demonstrate the community’s value in concrete terms that would attract sponsor support. “If we had only been able to pull off the project, we might have gained the enthusiasm and momentum to keep going.” As it turns out, the project foundered, even after careful preparation by a dedicated group of members, because there was no sponsor to step in to hold GSA to its agreement to provide technical staff to implement the design.

22

Results they achievedMembers gained much value from participation—and could imagine much more with a modicum of dedicated senior-level sponsorship for the community as a whole.

Generally, the greatest benefits of participation were about learning from others, building relationships, and gaining voice in policy matters.

Learning from others ATF: “We learned from Agriculture about using websites—learned what to ask

and not ask. We also learned from what NRC is doing.” “I saw something on document management from the Patent and Trademark

Office, and that was a huge benefit.” GSA: “I find others with similar problems and concerns, and even they haven’t

solved them yet, this helps. I find out a lot about what is going on that I didn't know about.”

“One benefit of the community is learning ‘anti-best practices,’ the ones people talk about but which actually don’t work in practice.”

“Electronic filing is really about records management—and therefore requires learning from others in multiple disciplines.”

Building relationships with peers: GSA: “Another benefit is the contacts we make that could help us in the future

—congressional aides, the CIO Council, NARA staff, peers in other agencies, etc.

SEC: “The federal government is so huge you need groups like this to find peers with information and knowledge.”

“You always see people at these meetings trading their business cards at the breaks.”

Gaining voice (with vendors, partners (NARA), and agency decision-makers): NRC: “The community has created a forum across agencies to bring issues to

the attention of NARA, for example—what standards we need to set, what our cross-agency records-management strategy should be, etc.”

PTO: “It helps to share our experiences with vendors—including the negative ones.

PTO: “It also helps to see where we agree on key records management issues—this helps me be stronger and more credible advocate for workable approaches in my agency.”

Another achievement of the E-Reg community was the spawning of a “sibling” community on E-Records. The question of E-Records management became so salient in many aspects of E-Regulation work that a number of members decided to start a separate though overlapping community on the topic.

23

What kind of sponsorship did they receive?The E-Regulation initiative is a distinctive and relatively rare example of a cross-agency community that was originally sponsored by a group of executives from multiple agencies. The initiative was sponsored as a demonstration project by the CIO Council—consisting of CIOs from various federal agencies—and guided by the Council’s Knowledge Management Working Group, led by Shereen Remez from GSA and Alex Bennet from the Navy. (Support staff members who worked for Shereen—including Wendy Stoner and Jon Desenberg at GSA—also provided support for the initiative.) Unfortunately, just months after the E-Regulation Community was launched, it lost the sponsorship of Remez, who was the primary champion for the community. (She moved on to another position when the Administration changed.) The community thus relied heavily on its coordinator, Bill Bennett, who at the time had the blessing of the head of FERC to devote a substantial amount of time to knowledge sharing and other knowledge-management activities associated with the community. About a year later, after a change in agency leadership, Bennett was staffed by his agency on full-time projects that prevented him from continuing to coordinate the community and no one else was able to provide sufficient leadership to keep it going.

24

Exhibit 3 Exhibit 3 A cross-level, cross-sector community: A cross-level, cross-sector community:

SafeCitiesSafeCities

The SafeCities Network was initiated in June 1999 by Vice President Al Gore’s Reinventing Government initiative to reduce gun violence in the United States. The announcement of the SafeCities community coincided with publication of the FBI’s crime-rate statistics, which showed significant variation across cities in injuries and fatalities caused by gun violence. Senior executives in the National Partnership for Reinventing Government (NPR) office began by convening officials from relevant agencies and developing a shared vision for what the network would be about and how they would work together. They sent out an invitation to cities and regions nationwide and selected ten coalitions to participate in the SafeCities community—based on criteria that included multi-sector collaboration, a track record of innovation, and commitment to improved results. These local coalitions provided stewardship for public safety issues in their cities and regions. A striking characteristic of the initiative was that it offered participants no funding—the value of participation was to get connected, to learn, and to enhance the capacity to reduce gun violence. The scale of the initiative was also distinctive—connecting civic coalitions from across the nation for the purpose of sharing ideas, collaborating on innovation initiatives, and helping to shape policy at local, state, and federal levels.8

The SafeCities community illustrates how a cross-agency, cross-sector, and cross-level community of practitioners can work together to address a national priority—at both local and national levels. SafeCities was a bi-partisan initiative that spanned both Democratic and Republican administrations and produced remarkable outcomes—in terms of increased capabilities, cross-boundary partnerships, and promising results.

Purpose of communityEach of the coalition members was focused on issues related to the area of public safety. Their specific domain targeted a sub-topic within this domain—defined as reducing injuries and fatalities due to gun violence. The specificity of this domain was crucial for coalescing a community with overlapping interests, focusing its learning activities, and attracting sponsors. Specific issues that the community addressed included prosecutorial strategies, gun-tracing methods, community policing, after-school programs, crime mapping tools and methods, and how to involve faith leaders.

The vision of community members was to show that by increasing public-safety capabilities at the local level, cities could achieve significant, lasting reductions in gun violence. The aspiration was also to build relationships with local, state, and federal agencies—as well as private and non-profit organizations. Finally, members at federal

8 See “Communities of Practice: A New Tool for Managers,” William M. Snyder and Xavier de Sousa Briggs (IBM Center for the Business of Government, www.businessofgovernment.org/), for a detailed case analysis of the SafeCities community and others sponsored by the NPR office and other federal agencies.

25

and local levels alike envisioned this community as a pilot for a scaled-up model that would include hundreds of cities in the United States and potentially, cities around the world. Sponsors and members saw SafeCities as a pioneering initiative on two levels: 1) building new capacity to reduce gun violence; and 2) demonstrating how cross-level, cross-sector, knowledge-based governance structures help solve civic problems and create new opportunities.

Who they were and what they didThe community was composed of members at local and national levels and from various disciplines and constituencies. At the national level it included officials from the Department of Justice Office of Community-Oriented Policing Services (COPS), the Bureau of Alcohol, Tobacco and Firearms, and other federal agencies; at the local level, there were mayors, police chiefs, faith leaders, hospital and social workers, after-school staff, neighborhood activists, and prosecutors. At federal, state, and local levels, the combination of cross-sector stakeholders was a unique and important factor in SafeCities’ success. For example, by building trust between police departments and faith communities—often groups with conflicting mindsets and approaches to inner-city violence—SafeCities members were able achieve breakthroughs in community policing programs.

The practice of SafeCities members included research, methodologies, tools, stories and other elements of a typical community’s repertoire. In this case, the practice elements were about coalition-building forums, community policing, after-school programs, crime-mapping methods, prosecutorial strategies, the design of local gun-possession laws, and ways to improve the interaction between at-risk youth and law-enforcement professionals. Members developed their own capabilities and the community’s collective repertoire through an ecology of learning activities; by organizing teleconferences on key issues such as community policing or gun tracing methods; arranging visits from federal officials to support local initiatives; conducting peer-to-peer visits to help spread best practices; and moderating a public website to publicize lessons learned and relevant resources related to methods for reducing gun violence.

Results they achievedSafeCities members include senior officials such as mayors, police chiefs, US attorneys with huge demands on their time, yet they found time to participate in SafeCities activities—one of the best measures there is of a community’s value. (This is particularly true in the case of communities because participation is voluntary—if you’re there it’s because you’re finding value for your time.) Typical benefits included the following:

o Miami police department strengthens relationships with federal partners in ATF and US and State Attorneys’ offices;

o Ft. Wayne champion visit brings press attention and galvanizes local coalition;

o Member visits to Springfield and Inkster help identify new strategies in areas of gun tracing and community courts

26

o A gathering of leaders from Miami and Seattle with peers in Ft. Worth highlights opportunities for involving the business community and influencing public opinion to reduce gun violence

o The King County Sheriff partners with a Safe Cities federal champion to convene a coalition-building forum that brings together mayors and police chiefs from over 30 municipalities in King County; fostering collaboration and spreading innovative practices across municipal boundaries at the local level

A more detailed look at one example helps to show the causal relationship between community participation and results. During an initial face-to-face gathering of the community, the police chief from Highpoint, North Carolina had presented a case study of a local program to integrate criminal offenders back into the community—including those considered “the baddest of the bad.” After hearing the Highpoint police chief talk about his success, groups from Ft. Wayne, Indiana and Inkster, Michigan—including police chiefs, mayors, and faith leaders from both cities—visited Highpoint and observed Highpoint’s program in action. Both coalitions then adapted the Highpoint model for their own locales with coaching from Highpoint. Later, a SafeCities member from Michigan reported that the visit had been instrumental to their implementation of a similar program. She said that the visit “added ideas and motivation to an initiative that we had been planning for a year. Once our mayor visited, he wanted to do it.”

Another important accomplishment of the network was the development of relationships among members across sectors and governmental levels—and across federal agencies themselves. The police chief in Highpoint, for example, had developed strong relationships with ATF and FBI officials, which in turn encouraged his peers to work more closely with federal agents. This level of trust and cooperation between federal agents and local public safety officials often does not exist—yet it has become increasingly urgent as the threat of terrorism becomes a way of life in cities nationwide. Similarly, participating in SafeCities fostered cross-agency collaboration—particularly between the Treasury Department (especially ATF) and various bureaus within the Department of Justice (such as the FBI and US Attorneys). Inter-agency collaboration was driven by urgent requests and a compelling readiness to act among local coalition leaders. The “pull” of local members galvanized officials across federal boundaries to combine efforts. We often underestimate the power of external stakeholders—local government, businesses, national non-profits—to foster such collaboration.

What kind of sponsorship did they receive?SafeCities operated successfully from June 1999 until June 2002, spanning the transition from a Democratic to a Republican administration. Political appointees from both parties, as well as senior civil servants in the Justice and Treasury departments (where the sponsorship was primarily based) believed in the cross-level, cross-sector approach that SafeCities embodied. Sponsors were impressed to see such active participation on the part of local civic leaders, even though they received no government funding for participating. These local leaders felt strongly about the value SafeCities provided—in

27

terms of ideas, access to expertise, and opportunities for national visibility and influence based on local success.

Agency sponsors ultimately transitioned resources and attention away from the SafeCities community in order to focus on a new national program. The decision confused many of the participants, given the minimal federal costs associated with the initiative, principally the cost of funding the community’s full-time coordinator, Michael Seelman (a mid-level staff person) and intermittent attention by agency champions. Michael’s role as the community coordinator was particularly important—facilitating steering committee meetings, arranging speakers for teleconferences, documenting insights on the Web site, arranging peer-to-peer visits, and coordinating with state and federal officials. The loss of the coordinator and agency attention was a fatal blow to the community. In its place, the Justice Department enacted a new program, called Project Safe Neighborhoods, which provided funding for local initiatives in areas such as crime analysis and prosecution. The program managers intended to build on the SafeCities foundation, but they did not fully appreciate the distinctive characteristics of the community—active federal-agency champions and a high-value opportunities for peer-to-peer learning and collaboration across cities, sectors, and levels of government. While SafeCities members were glad that the government was providing new funds to support local initiatives, they passionately argued that such funding could never substitute or compensate for the value of the SafeCities community.

The SafeCities story thus validates the power of cross-city communities of practice while highlighting a key challenge: How help political and agency leaders see the power of this approach more clearly and build on its success? Ironically, initiatives such as SafeCities often have trouble getting visibility and support because they cost so little—perhaps a few-hundred thousand dollars, vs. tens of millions or more for a typical program. Another difficulty is appreciating what it takes for government officials to cultivate peer-to-peer learning and cross-boundary collaboration. Cross-boundary leadership is not as easy to operationalize as are funding allocations.

But this story is not over yet. SafeCities alumni area are still in touch; they continue to explore ways to incorporate what they learned about cross-agency, cross-city collaboration to get better public-safety results.

28

Exhibit 4Exhibit 4 An on-line distributed community:An on-line distributed community:

CompanyCommand.comCompanyCommand.com

Ironically, a community best-known for an online facility that promotes learning and connecting among thousands of members started out as a series of incidental conversations between two rookie commanders in the US Army—Tony Burgess and Nate Allen, the founders of CompanyCommand.com. Tony and Nate had just started their stint as commanders and spent countless evenings talking about work on the front porch of their neighboring houses on a base in Hawaii. They found they learned a lot from their informal talks by swapping stories and helping each other think through problems they faced—ones they wished they had known about before they took command. At some point they realized they couldn’t be the only ones who could benefit from the kind of informal learning and collegial support that they were experiencing. They’d been educated at West Point and received the best training the Army could provide, but nothing in the training courses or manuals could prepare them for commanding a unit with over 120 men and a budget of several hundred-thousand dollars or more—nothing, that is, like experience and the opportunity to swap stories, ideas, and lessons learned with peers as well as mentors who have been there before.

Tony and Nate were driven by their own experience as well as a passionate commitment to help their fellow commanders. They wrote a very popular book on lessons learned from their own experience in command and—together with a team of like-minded peers–began developing an idea for an online site that could provide a more interactive and up-to-date context for commanders to ask questions, talk about hot topics, and find resources. As founders, Tony and Nate began with a vision based on the belief that commanders were savvy enough to assess new ideas and figure out which would work in their context. They recruited an initial team of 10 Army officers who contributed long hours, on their own time, in addition to performing their day jobs (to the consternation of their spouses!).9 The community grew rapidly and now engages nearly 4,000 company commanders worldwide.

The issue of connecting practitioners “virtually” is a salient one because nearly all large organizations today are distributed along various dimensions: by business units, geographies, time zones, and others which make it difficult to connect in conventional ways. All the communities featured here applied technology for distributed learning—and all except the E-records community used online facilities. But CompanyCommand.com relies most extensively and operates most intensively on a virtual platform that promotes learning and connecting.

Too many organizations focus on computer-based mechanisms—such as interactive websites with chat rooms, bulletin boards, and document repositories—and then are

9 The original team of Army officers included: Tony Burgess, Nate Allen, Steve Schweitzer, Pete Kilner, Steve Delvauz. Chris Engen, Tom Woodie, and Craig Whiteside.

29

dismayed to find that no one shows up to use them (a result that occurs more than half the time in our experience). Although technology is a powerful tool in the hands of skilled community leaders and motivated members, it is most often overemphasized, badly designed, and incompetently applied. Therefore, it is worth considering a community that relies heavily on technology for connecting practitioners and facilitating peer-to-peer learning. The headline in this case, as in any successful technology application, is that the community has integrated a “high-touch” approach with an evolving “mid-tech” infrastructure (which started as low-tech and is gradually incorporating high-tech features) that is user-friendly and highly accessible.

There is not room here to detail the design of CompanyCommand.com, but it is very important to understand that the on-line facility would never have succeeded without a dedicated group of coordinators to make it a compelling place to learn and connect for members.

Purpose of the communityThe focus of CompanyCommand.com is to promote professional development for company commanders. The community’s vision is to involve every company commander in the Army in an ongoing, vibrant conversation with peers and other experts about leading and building combat-ready teams.

The learning agenda for the community includes issues related to a specific practice: what it takes to lead a company effectively. Sample issues include: getting a company through the initial training period; organizing a command and getting it ready under battlefield conditions; dealing with interpersonal conflicts; negotiating with and monitoring the performance of key suppliers; interpreting and applying Army regulations to fuzzy situations; and developing and managing a budget under uncertain conditions.

Key objectives for the community include: • Lateral sharing of knowledge across Army• Interactive, ongoing communication• Bringing best practices to the surface• Creating community• Linking the hierarchical silos of knowledge throughout the Army to the field

Who they are and what they doThe CompanyCommand.com10 community is a particularly striking example of a distributed community. Its 2000 members, primarily consisting of company commanders in the Army, are located in Army bases distributed worldwide. Participants also include former commanders who have advanced to higher positions or have retired. And many participants are platoon leaders and others who are getting ready to become commanders. But the site is public and can be visited by anyone with access to the web.

The founding team was determined to find a way to accelerate the learning process and provide encouragement and support for leaders in an extremely challenging and crucial

10 http://www.companycommand.com/

30

role. They developed a richly variegated website that would be easy to access worldwide, and that would not be limited by formal constraints against peer-to-peer knowledge sharing (hence the public accessibility). Members use the various features of the website to get access to vetted tools; hear (and tell) stories about their experiences; watch video-clip interviews with commanders in the field or senior officers, read updates on development opportunities and new resources; keep up with member blogs; participate in a book-reading club, and find others to get help or share interests. Community coordinators have been able to leverage the expertise of former commanders, connect “silos of excellence” among members, and accelerate the “time to talent” of isolated leaders in demanding roles. The power of community participation is particularly visible as members support their peers when they are serving in active battlefield conditions. Members who make particularly strong contributions get recognition on the site as “heavy hitters” and receive mugs and accolades—highlighting both their valuable contributions and reinforcing the values of the community—such as grassroots innovation and “positive voice with a focus on solutions.”

The success of the CompanyCommand website compared to the failure of most on-line initiatives is the exception that proves the rule: high-tech only works in combination with high-touch. The vitality and usefulness of the CompanyCommand.com website depends on a team of nearly 10 passionate core-group members who write up newsletter items; field questions; connect members to peers for problem solving; upload documents, video clips, photographs and other material; recruit participants; cultivate support of senior officials; and finally, tweak the functions and design of the technology to make it easier to use. The community’s core group help to personalize the site, make it a comfortable place to browse and contribute—while also applying a rigorous discipline to make sure the content is fresh, well presented, and useful.

The founding team believes the next step in this community’s development—a step they have already begun to pilot—is to hold local and regional face-to-face meetings. These can build on the connections members have made online and cement the inchoate feel of belonging and mutual trust that is an essential platform for the intense levels of helpfulness and encouragement that members share in the virtual space.

In part because of its unanticipated success, the community faces a number of interesting challenges:

Small groups. Whether or not the face-to-face meetings flourish, the community has started to generate a substantial number of small groups that focus on specific subtopics. This has given rise to the new challenge of keeping a balance between small group involvement and participation in activities of general interest.

New generations. A characteristic of the membership of CompanyCommand is that members are in the practice of command for about two years, which means that membership is always rotating. A corresponding turnover in community coordinators is appropriate and needed in order to stay current and socially connected. The founding team is now putting significant effort into recruiting, training, and coaching new generations of volunteers. The idea is that current commanders—or those fresh

31

out of command—have the relevant knowledge and the social capital necessary to continue delivering world-class value to community members.

Current content. A related challenge is dealing with the accumulation of content over time. The team is considering a process by which old content would be archived away every few years to make room for new content contributed by the current generation of participants.

Integrity and integration. The community’s leaders have been careful to maintain their independence from the Army’s formal bureaucracy in order to preserve the integrity of the community as an open exchange of information by its members for its members. A new challenge, now that the community has achieved high visibility, is to consider what level of integration with existing Army training and learning initiatives will benefit both the community and the organization. Admittedly, the community’s open and public exchange of information among peers is both attractive and discomforting to an organization that faces an unprecedented need for accelerated learning yet has well-established "ways of doing things"—a sophisticated system of schools, security provisions, doctrine-production procedures, and best-practice analysis. While existing institutions recognize the potential value of the community’s rapid and unimpeded information flows, it is essential to acknowledge their ambivalence about it: Is the community the way of the future? Is it a competitor? Will it produce reliable knowledge? Should we embrace it, mold it, ignore it, or undermine it?

Such questions call for a broader perspective. It is important to understand the community’s place in the overall learning system, considering the perspective of the system as well as the community. The community does not aim to replace existing institutions but to provide an additional dimension that complements formal schooling, scientific best-practice analysis, and orderly doctrine production. At the same time, it does need to preserve its integrity as a member-driven community; or it risks losing focus, credibility, trust, and mindset—the very characteristics that enable it to function as a unique place of knowledge. Balancing integrity and integration will require sponsorship from well-placed members of the formal hierarchy who understand the subtleties of these tensions; they must be ready to use the legitimacy of their position to address them constructively. To this end, they must articulate a broad vision of the overall learning system to help orchestrate it in such a way that the various components can all contribute their respective value in mutually enhancing fashion.

Results they achievedThe CompanyCommand.com community has achieved outstanding success in the last several years since its start-up. The community’s leaders are persistent about assessing the value of the community for its members. They collect both quantitative statistics on site activity as well as qualitative evidence through member testimonials in order to assess the success of the community, make its achievements visible, and to learn how to serve members better. These data—both qualitative testimonials and quantitative

32

statistics on activity—also serve to reinforce members’ collective perception of the site as “the place to be” in order to connect with colleagues worldwide.

The best way to illustrate the community’s success is to quote from member testimonials because these describe the linkage between what their community forum provides and how they leverage the knowledge base to improve results in their work.

“I am a Rifle Company Commander in the ARNG. I am also an engineer with a large pharmaceutical manufacturer in south Texas. I drive 350 miles (one way) to drill every month. The nearest military post is Ft. Polk, LA - a five-hour drive. Your website enables me to stay on top of things and continue learning about the numerous ways of dealing with the various demands on commanders throughout the military. I do not have the "luxury" of having my fellow CO Cdr's next door, nor can we meet at the O Club weekly to share and discuss TTP's and pick each other's brains. I can email my BN CDR, but it's just not the same as having a pool of peers to draw knowledge from. CompanyCommand.com has become, in effect, my O Club meeting, the other rifle company commander next door, and my never-ending source of reference. How do you measure the value of THAT?“

“I don't have any cool command stories yet, since I have only been in the Army for 4 years, but CompanyCommmand.com has helped me to make a major decision in my life ... As time for promotion to captain and career course drew near I began visiting CompanyCommand.com and I realized that there are A LOT of officers in the Army who really do care about combat readiness. I found that there really are capable leaders who are leading our soldiers and doing great things. I completed the FA Captains Career Course in July, and I have recently reported to the 1st ID in Germany. Thanks for helping me to see the truth.”

"I'm currently in Baghdad awaiting for someone to tell me I can come home.... Our soldiers did a very great job. I'm proud of each and every one of them...I will be happy to sit down with you guys when I get back. I suggest that you guys come to Ft. Stewart and visit all of the commanders. There were a lot of things we have learned. In ending, I would like to say that you guys deserve a Medal of Honor for all the work you have put into this site. I owe you guys, the things I have learned from this website helped me in combat."

“Having been in the Army for such a long time, I have watched as ‘mentoring’ and current leaders failed to train their subordinates to fall in on their position—perhaps believing themselves indispensable or unwilling to relinquish power. This website empowers all of us who have had to pick up the reins of leadership without having those before us show us the way. I have learned more from this website and the 'mentoring' provided by all of you who care so much about our Army to produce this wonderful product. I can assure you without a doubt that the value added of your website is such that without it, I would not have been able to complete many of the myriad tasks that I face on a daily basis without making onerous mistakes that would have required a long time to untangle and fix.”

33

Quantitative data corroborate the evidence from testimonials and give numerical value to the intuitively obvious intensity of on-line participation. For example, in an update in January 2003, the community leaders announced that January “was another record month for companycommand.com.” The community’s site had over 46,400 unique visitors, logged more than 1.8 million hits and downloaded 402,632 pages, 16 GB of data. That month, the most popular items downloaded by members were about support unit deployments to the Persian Gulf and lessons learned from previous Gulf operations, deployment checklists, and advice and tools for unit family-readiness groups. Meanwhile, a sister site for platoon leaders, created by an offshoot of core leaders of companycommand.com was building quickly. Leaders announced that “PlatoonLeader.army.mil has become the place to engage in vibrant conversations on life and leadership as an Army lieutenant.” Between its start-up in September, 2002 and January 2003, traffic on the site grew 300% and the site logged 852,000 hits and 144,000 page views and had a record 1.6 GB of data downloaded in January alone.

What kind of sponsorship did they receive?Again, regarding sponsorship CompanyCommand.com is the exception that proves the rule. The impetus for the community started with the passionate determination of its founders to connect company commanders. The CompanyCommand core team was overwhelmed with the enthusiastic support and intensive level of contributions and appreciation from commanders worldwide, many serving in battlefield conditions in Afghanistan and Iraq. In this case, the core team directly serves colleagues who share a strong cultural identity and the common experience of a well-defined leadership role. Commanders feel an immediate and strong connection with peers. This culture of shared identity and mutual commitment, along with the urgent need for help in a difficult role, has contributed to making the community a success without senior support and official legitimacy. Long-term, however, the community will need such executive support and legitimacy to continue to build and to further leverage opportunities to integrate with established military training development activities. This is especially true as the founders move on and the new leaders find they need time allocations to keep up with the demands of a growing community.

To address this evolving situation, the founding team approached the leadership of the United States Military Academy at West Point. The senior leaders at West Point recognized the value of the project and resourced it with personnel slots and funds "out of hide" to allow the community to continue to grow and develop. The opening of these positions at West Point did not fully resolve the issue of sponsorship and integration for CompanyCommand, which is still under discussion beyond the Academy. But it did mark a new phase for the founding team, which reflected the potential value of their work beyond their focus on CompanyCommand: They were to become a support team that would train and coach new generations of community leaders in the US Army.

34

Exhibit 5 Exhibit 5 The structure of communities of practiceThe structure of communities of practice

It is useful to define the notion of community of practice more formally so that organizations understand better when to turn to this kind of structure, what to expect from them, and what it takes to develop them.

Structural componentsWe have described the key components of communities in terms of three structural dimensions: domain, community, and practice (Wenger, 1998; Wenger, McDermott, and Snyder, 2002). A community’s effectiveness as a social learning system depends on its strength in all three areas. Domain. A community of practice focuses on a specific “domain,” which defines its identity and what it cares about—whether it is implementing transportation safety measures, reducing gun violence, digitizing regulatory records, or developing military officers. Passion for the domain is crucial. Members’ passion for a domain is not an abstract, disinterested experience. It is often a deep part of their personal identity and a means to express what their life’s work is about. Community. The second element is the community itself and the quality of the relationships that bind members. Optimally, the membership mirrors the diversity of perspectives and approaches relevant to leading-edge innovation efforts in the domain. Leadership by an effective “community coordinator” and core group is a key success factor. The feeling of community is essential. It provides a strong foundation for learning and collaboration among diverse members. Practice. Each community develops its practice by sharing and developing the knowledge of practitioners in its domain. Components of a practice include its repertoire of tools, frameworks, methods, and stories—as well as activities related to learning and innovation. Note that a shared practice goes beyond a shared interest: over time, members of a community of practice seek to improve their ability—collectively and individually—to do what they do.

To optimize a community’s ability to act as a living knowledge structure in an organization, it is important to develop all three elements in parallel.

Distinctions: Communities of practice serve the specific function of allowing practitioners to get together to “manage” their own knowledge. They are not meant to replace other structures in organizations. Contrasting communities of practice with more familiar structures can help make this clear.

Teams. Communities are distinct from teams because they focus on building capability rather than on delivering a product or service. The domain of a community consists of key issues to address, while the goals of a team set specific performance expectations. Community members participate voluntarily at various levels of

35

engagement—core, active, or peripheral. Team members have clearly defined roles and are accountable for formally established time and service contributions. Finally, the practice of a community consists of its skills, methods, and stories—and how it learns together. Team members also learn together, of course, but their primary emphasis is on getting the product (or service) out the door.

Networks. Communities are distinct from the common notion of networks, in which members are connected directly or indirectly, sometimes based on work issues but often through personal relationships or chance meetings. Communities, in contrast, are a type of network in which nearly all members know all the other members in the community, and feel a shared commitment to learn together about issues related to a domain they all care about.

Formal units. Communities of practice do not replace more formal organizational structures such as teams and business units. On the one hand, the purpose of formal units is to deliver products and services—whether they are consumer goods, government services, or a project outcome such as neighborhood renewal. Communities of practice, on the other hand, help ensure that learning and innovation activities occur across formal structural boundaries. Indeed, a salient benefit of communities is to bridge established boundaries—among organizations, constituencies, and disciplines—in order to increase collective capabilities and social trust. For example, the brake engineers in the product development organization at DaimlerChrysler have their primary affiliation with the car platform where they design vehicles. Yet they also belong to a community of practice where they share ideas, lessons learned, and tricks of the trade. By belonging to both types of structure, members can bring the learning of their team to the community so that it is shared through the organization, and, conversely, they can apply the learning of their community to the work of their team.

36

Exhibit 6Exhibit 6How communities learn:How communities learn:

typical activitiestypical activities

Projects

Visits

ConversationSpace

Informalinteractions Teleconferences

Face-to-faceconferences

Website

Face-to-face conferences build trust and foster sense of joint enterprise. Members become more willing to:• Engage each other on challenging topics• Initiate one-to-one exchanges • Ask questions and solve problems over the phone• Provide answers on listserv• Contribute to website

Teleconferences (and video) provide for low-cost, interactive problem solving, idea-generation, and education on topics

On-site visits develop personal relationships and foster innovation diffusion

Projects address community and stakeholder knowledge requirements

Virtual space for efficient information-sharing, Q&A, and peripheral learning

Website captures information for easy access to members and broader public. One-to-one

interactions by phone and email help build relationships while getting help and sharing ideas

WorkshopsWorkshops and classes educate members on codified tools, frameworks, and methods

Figure 6.1. An ecology of community learning activities

The members of a community of practice do not usually work together on a day-to-day basis. They engage in various learning activities to connect and learn—formal and informal, public and private, virtual and face-to-face. They organize an ecology of complementary learning activities. In a typical example, teleconferences would be held every two or three weeks. These provide the heartbeat of the community—regular, easily accessible ways to keep up-to-speed on hot topics and hear what others are doing. The website provides a mechanism for catching up on resources or finding contact names or resources mentioned during the calls, while the listserv reminds members of what’s coming up and highlights related news in the field such as new grant opportunities or partner accomplishments. Agency-champion and peer visits as well as informal backchannel conversations are ways for members to get help in areas where they want to learn and innovate—often following up on ideas or experiences they hear about from others.

37

Informal conversations among member peers and with the community coordinator, combined with ongoing coordinator canvassing, help identify future topics for teleconferences and collective projects for the community or project teams to pursue. The face-to-face meetings provide a crucial foundation for all these activities. They provide a forum—“where we can break bread together”—to meet people in person and find connections. These personal relationships weave the community together and help build trust and mutual commitment.

The overall constellation of learning activities forms an ecology of interactions that provide value on both practical and personal levels. (See Figure 1.) They provide concrete mechanisms to promote learning and connections among practitioners from a variety of organizations, disciplines, and locales. And they also increase the community’s “presence” in members’ lives and reinforce the sense of belonging and identity that are the foundation for collective learning and collaborative activities.

A vignette of the SafeCities Community in actionIt is 1:30 pm EST on Wednesday April 12, and 29 members of local coalitions nationwide are joining their monthly teleconference. They are all members of the SafeCities Network (see Exhibit 3). This month the topic is “Faith-based community strategies.”

Faith leaders from Ft. Wayne, Indiana and Highpoint, North Carolina are the featured speakers. They describe how groups of faith leaders in their cities had achieved dramatic reductions in gun violence through collaborations of local faith organizations and police departments. Reverend Jordan from Ft. Wayne describes the evolution of a collaborative effort that began in 1992 after a highly publicized accidental shooting of a teenager participating in activities at a local YMCA. Jordan began holding meetings between gang leaders and city officials to talk together about how to stem the rising tide of violence in the city. These gatherings—which featured gang leaders giving presentations and putting on skits to communicate the importance of recreation alternatives and mentors for teenagers—were very successful. The meetings led to the development of a city-wide “Stop the Madness” program and to later collaborations that included an alliance of churches, the police department, the mayor’s office, and community leaders.

Reverend Fails of Highpoint describes how faith leaders and the police department collaborated to reduce citywide violence by focusing on “the baddest of the bad”—individuals on parole or probation who are most likely to commit violent crimes. Their approach includes businesses, faith leaders, school principals, social services agencies, and community leaders—as well as the chief of police, district attorney, and representatives from the FBI, ATF, and state and federal-level prosecutors. The Highpoint model features a “good cop-bad cop” approach that focuses as much on helping these people find jobs and stabilize their lives as on aggressive law enforcement methods. Speakers field questions regarding how to get faith leaders and their congregations involved. The partner from Los Angeles describes a program in which clergy help gather information on gun violence from congregation members, and a

38

partner from Louisville talks about their clergy’s effort to pass firearm legislation and distribute trigger locks. Partners request the phone numbers of speakers and are encouraged to intensify or begin explorations of ways to involve the faith community to reduce gun violence. That summer, faith leaders from Inkster, Michigan will accompany their mayor and police chief on a visit to Highpoint to learn how to develop their own program.

A follow-up initiative leverages the lessons learned in the April 12 call. In October, a satellite broadcast reaches over 50 sites nationwide and features Reverend Jordan talking about youth violence and a new initiative to develop a training academy for faith leaders on topics such as crisis response and community policing.

* * * “The faith community conversation was really useful. We have many churches in our community, but we haven’t been sure about how to involve the faith community. Fort Wayne talked about how after a tragic experience they got ministers involved. We wanted to know how to involve them without waiting for a tragic experience. The Fort Wayne example encouraged me to dig for more examples.”11

11 Vignette and related analysis is quoted from Snyder & Briggs, 2003, op cit.

39

Exhibit 7 Exhibit 7 Community lifecycle:Community lifecycle:

stages of development stages of development

Like all human systems, communities of practice evolve through a developmental life-cycle. Some evolve further than others, of course, and many never become more than ad-hoc groups of friendly colleagues who see each other now and then. The full community development cycle can be parsed in terms of five basic stages: discovery, coalescing, maturing, stewardship, and transformation (Wenger et al., 2002). This schematic model of typical developmental stages helps to specify the types of OD interventions that can promote community maturity and vitality in both organizational and societal contexts.

Stage 1 – Discovery. A community of practice often begins with informal conversations among people who share interests and passions related to an area of inquiry, a discipline, or a set of pressing problems. At some point, a core group of participants may agree to explore ways to connect on a more regular basis. They begin to reach out to colleagues to connect the nodes in their various networks. In this phase, the inchoate community starts to identify issues and opportunities that members want to pursue together. Potential members are invited to join and sponsors are courted.

Stage 2 – Coalescing. The community begins to coalesce when it has reached a critical mass of members and they have identified compelling issues to explore together. In this stage, members begin meeting to talk about innovative practices, present mini-cases, and consult to each other on problems—as well as imagine how they might steward their collective practice more systematically. They develop a leadership structure, which includes a coordinator (one or several people) to organize community events and to do back-channel recruitment and networking to keep members engaged and the community vital. As in any group, community members may need to address group flare-ups that occur, which may be fueled by competing external interests and framework wars, for example, or due to members’ personal attachments to outdated ideas or “signature skills.”

Stage 3 – Maturing. Once the community has established a track record of creating value for members—knowledge, relationships, reputation—it begins to consider more structured, time-intensive projects. At this point, the community becomes especially dependent on sponsors to fund staff time, travel, and at least minimal technical and logistical infrastructure. Sponsors are also crucial for the visibility and legitimacy they can confer on community projects and knowledge resources. During the maturing stage many communities expand membership—often by creating sub-communities organized by topic, roles, or geography.

Stage 4 – Stewardship. A minority of communities achieve the stewardship stage. Many are content to share ideas and approaches among members at a stage-two level. Stewardship communities, in contrast, seek to provide visible and influential leadership related to the issues in their domain—within and across organizations. Stewardship at its

40

best means that the community grows with the changes in its domain—by addressing the driving issues that worry stakeholders; the needs of experienced members to keep up to speed while mentoring apprentices; and the ongoing challenge to invent state-of-the art tools and frameworks in the field. As the community becomes more strategic, stakeholders raise their expectation levels—in part due to the promotional efforts of the community itself.

Stage 5 – Transformation. The final phase of the development cycle can take different forms. (And a community need not reach maturity to begin transforming itself in one way or another.) The community may morph into a formal structure—a functional department, for example, or a non-profit organization. The domain may transform so radically that various sub-communities tackling second-generation issues take precedence—and the original set of issues and knowledge base become history. Finally, members may simply discontinue their structured activities, while a few continue to connect informally. In some cases, the community may decide to organize a closure ceremony to celebrate its accomplishments and honor distinctive contributions by members.

41

Exhibit 8Exhibit 8Going to scale:Going to scale:

fractal communities of communitiesfractal communities of communities

Once an agency or cross-agency community has been established and has demonstrated its ability to support results, sponsors may consider increasing investments to enable the community to address the full scope and scale of the issues included in its domain. Many of the problems faced by governments are vast and complex. The relevant domains of knowledge are varied and interrelated; the communities tend to be very diverse and distributed over large geographical areas; and the practices involve different disciplines and varied local conditions. One of the great challenges of a community-based approach is to “go to scale” so that the learning system matches the size and complexity of the problem it is meant to address. How do you significantly increase the scale of a community-based learning system without losing core elements of its success—identification with a well-defined domain, close personal relationships, and direct access to practitioners for mutual learning? This dilemma was captured in a comment by one community member:

“We’re very loathe to see bigger as better. When programs expand, they expand into oblivion, so how to expand our sphere of influence without expanding the program? The strength of our program is that it is relatively small, with one-on-one relationships with each other and people in Washington.”

The idea is to grow a “community of communities.” In such a design, communities are nested within one another in “fractal” structure. That is, each level of substructure shares some basic characteristics with the other levels so that it “embodies” the whole in each nested community. Applying such a design principle, it is possible to preserve a small-community feeling while extending a system from the local to the national and even international levels. Groups can subdivide along various dimensions: geography, topic, role, and others. Topical and role-based sub-groups provide a way to connect members of local communities across regions—strengthening the overall fabric and connectedness of the entire network. The transition from small communities to a large-scale learning system is illustrated in Figure 8.1.

The key insight of a fractal structure is that crucial features of communities of practice can be maintained no matter how many participants join—as long as the basic configuration, organizing principles, and opportunities for local engagement are the same. The pay-off for creating such an extensive, multi-tiered system of interrelated communities is that it provides a magnificent learning laboratory to identify emerging issues and test proposals in practice with motivated sites. This natural laboratory leverages innovations and replication efforts through a system of communities ready and willing to share results quickly and convincingly with peers and stakeholders. They can then help to translate these quickly into better policies and improved results. As a result, the learning potential of the overall system and the influence at local levels can increase

42

significantly with scale. The key challenge of a large-scale learning system is not whether people can learn from each other without direct contact but whether they can trust a broader community of communities to serve their local goals as well as a global purpose. This depends on the communities at all levels—local, state, national, and international—to establish a culture of trust, reciprocity, and shared values. Developing this social capital across all levels is the critical success factor for going to scale. The evolution of a learning system must therefore be paced at the time-scale of social relationships, not according to an externally imposed objective to achieve short-term results. Organizers need be careful not to scale up too fast. They need to establish trust and shared values at different levels of aggregation through various mechanisms, including a community of trusted brokers who weave the network across localities.

The federal government is in a unique position to cultivate such systems of communities of practice that achieve impact at both local and global levels. Indeed, there is a crucial role for actors in the position to act as overall conveners of such large-scale systems—something that local actors are often not in a position to do even if they can see the need and are willing to participate. Well-positioned conveners can serve the system in a variety of ways:

Establish a foundation of shared vision and values, including cross-community commitments to social norms and a collective vision of what their overall problem is about.

43

Figure 8.1 Going to scale through a fractal structure

Create a “brand” to legitimize the work of the community and promote membership as a nationally recognized achievement. Many members of cross-agency communities mentioned the power of the community’s brand name and affiliation with high-level executive offices in the federal government as having a positive effect on their credibility and influence at the local level.

Provide logistical support and shared technical infrastructure to help with local event planning, community coordination, and resources. Communities of practice benefit from at least a minimal communications and knowledge-capture infrastructure. Longer-term, it makes sense to use a common set of infrastructure facilities and tools to facilitate interoperability and reduce behavioral hurdles to cross-boundary communication.

Organize events and projects that cut across levels and include communities across regions and at multiple tiers—local, regional, national. They may also be organized to include communities focused on different domains to encourage cross-pollination of ideas and to weave relationships that facilitate the flow of insights and methodologies. Events and projects may be face-to-face or virtual.

Escalate problems. Where problems that can be solved locally, they are. When they need regional or national attention, members can escalate the issues through mechanisms they develop within and across chapter communities. Local communities are likely to have more influence and credibility for issues they raise to the federal level because they have been vetted by a representative group of peers before being escalated.

Orchestrating such large-scale community-based learning systems is a new role for government agencies. It gains its effectiveness, not so much from making policy or grants, but from connecting disconnected communities so they can learn from each other. This new kind of responsibility therefore requires visible sponsorship from well-positioned executives who can give it legitimacy, articulate the strategic opportunities, make resources available, and convince players and stakeholders to participate.12

12 Material on going to scale draws from Snyder, W.M. & Briggs, X. de S., 2003. “Communities of Practice: A New Tool for Managers.” Arlington, VA: IBM Center for the Business of Government, http://www.businessofgovernment.org

44

Exhibit 9 Exhibit 9 Assessing the value of communitiesAssessing the value of communities

The question of whether and how to assess the value of informal structures such as communities of practice has been the subject of some debate, including extreme positions on both sides. On the one hand the purists argue that the real value of a community can never be captured in data and, moreover, that any attempt at measurement will ruin the spirit of a community. On the other hand, the hawks argue that if something cannot be measured in exact bottom-line figures, it does not have any relevance and certainly will never gain legitimacy in organizations. Our experience suggests a more pragmatic position:

It is quite possible to assess the success of a community of practice in a meaningful way if one uses the right mix of qualitative and quantitative data.

Measurement does not come for free. It takes time to gather and process meaningful data. Given that community time is usually scarce and precious, one has to be careful how much energy a community spends justifying its existence.

Unlike traditional accounting, useful attempts at measuring the value of a community do not purport to measure everything, but to give a sufficient account of value creation to satisfy the needs of various stakeholders, including sponsors and members.

Communities that seek funding, legitimacy, and support from sponsors will likely need to make their value visible in terms the organization can understand.

In the end, members need information on a community’s impact as much as sponsors do. Information about value is a source of inspiration. It helps them make decisions about when and how to participate and makes it easier for them to recruit colleagues and gain the support of their own managers for their participation.

Finally, measures help a community identify its strengths and weaknesses. They can provide information for deciding where to focus renewal efforts—including expansion, new projects, or stronger linkages with stakeholders.

One way to think about how to describe the value of communities is to trace the linkages that connect community activities and performance outcomes. These can be described in terms of a simplified model as shown in Figure 8.1. This model posits causal links between community conditions and activities (meetings, discussions, trust), community output (new knowledge, tools, ideas) and effect on performance in terms of improved results (time saved, better services).

It is important to collect stories that will show the causal links between these elements to explain how a community activity led to an improved result. Thus, to evaluate the extent to which communities of practice drive results at the local level, it is necessary to tell the story of two related causal linkages. The first link traces the path from activities—such as

45

a teleconference discussion about after-school programs—to a new capability, such as improved deployment of after-school programs in areas with high levels of juvenile crime. The second link shows the causal relationship between capability and performance; for example, better, more available after-school programs and a reduction in local crime rates. Figure 8.1 illustrates this process with the story of the visit to Highpoint, North Carolina, by SafeCities members from Indiana and Michigan (see Exhibit 3 for details).

Meaningful data of both qualitative and quantitative nature can be gathered at each level. Community activities. Do people participate, come to meetings, and login to the

website? Do they respond to e-mail requests for help from other members? For busy professionals, these measures reflect the health of the community and its value to discriminating participants.

Qualitative: member feedback regarding the quality of meetings and opportunities to meet others with a variety of perspectives (E-reg)

Quantitative: statistics on the number of website visits and document downloads (Rumble Strips, CompanyCommand)

Capabilities produced by the community. New capabilities take the form of fresh ideas, solutions to problems, documents on the web site, the density and extent of community connections, tools and methods produced, etc. Capabilities by themselves do not ensure that they are applied productively, but they do show the potential generated by a community.

Qualitative: reportings of new ideas and connections (E-reg), new coalitions (SafeCities)and comments on quality of website (CompanyCommand)

Quantitative: number of states adopting rumble strips, number of tools posted (CompanyCommand)

46

Community activities

New capabilities

Improvedperformanceproduces produces

Highpoint Police Chief Quijas presents case study at meeting

Indiana and Michigan partners visit Highpoint to see program in action

Goal to achieve significant reduction in gun violence as demonstrated by Highpoint

Attendees get new ideas and interest and know whom to contact

They adapt Highpoint model for local application

Figure 9.1. Causal model of value creation with sample story.

Improved results. Effects such as time saved, assets reused, new business acquired, and errors avoided can all be expressed in terms of bottom-line value. Member testimonials, when showing a steady pattern, can give a good sense of the value that the community is producing.

Qualitative: testimonials about skills gained, time saved, soldiers prepared (CompanyCommand)Quantitative: savings in lives and treasure (Rumble Strips)

With this model, qualitative and quantitative data are not gathered in a vacuum. They are used to corroborate the causal relationships provided by the stories that link activities and results. Together, stories and data can yield a solid picture of the value that a community produces, even assuming that the measurements have not captured everything and that there is a lot more happening of a less tangible nature.

Having a good picture of the value created by a community may be especially important when it cuts across agencies or other budget centers and there is no obvious sponsoring body, as was the case for the E-reg community. In such cases, it will be necessary to show the value of crossing boundaries at two levels. First, showing how the community delivers value to each individual agency involved will encourage agencies to invest in the community and support staff participation. Second, showing how the community contributes to the overall e-gov initiative will encourage sponsors to incorporate a community-of-practice approach in the design of the initiative.

The point is not whether to measure or not to measure, but rather to determine how best to produce a useful picture of what a community is doing. This will enable sponsors to decide what level of investment to make in communities. In most cases, communities do not require huge investments compared to other initiatives. What is the return on investment compared to alternatives? Consider that the expense of a full-time coordinator plus travel for 30 distributed members to an annual event is about $200,000 per year—and this minimal investment engages its members—local mayors, police chiefs, faith leaders, local prosecutors, business leaders and others—to accelerate and improve the implementation of pioneering gun-violence-reduction methods across the country? Such a community investment recruits the energy of 30 influential civic leaders nationwide for the price of one federal employee. In many cases, implementing the community’s ideas for innovation does not add significant new incremental costs. Changes in methodologies often simply leverage current funds or local civic resources in qualitatively better ways–improved utilization of gun-tracing methods, prosecutorial strategies, or programs for integrating parolees back into communities. Activities to improve cooperation among district, state, and US attorneys may cost very little—yet the benefits in prosecuting and preventing crimes could be significant. Linking civic leaders by a teleconference twice a month to share stories such as these and make connections to help each other implement innovations costs a pittance. Compare the cost to what it takes to develop complex policy initiatives, restructure organizations, or fund nationwide programs—yet in many cases a community-based approach is likely to have a commensurate if not greater effect.

47

Exhibit 10Exhibit 10Portrait of a community coordinator: Portrait of a community coordinator:

William BennettWilliam Bennett

William Bennett’s work with the e-regulation and later the e-records communities shows a savvy, skilled community coordinator in action. His case is particularly useful because he, like many in this role, also had a “day job” and could not contribute more than about a third of his time to the role. Nevertheless, he was able to help launch two communities from scratch, each of which required pulling together practitioners from across agency boundaries—no mean feat. We can learn much from Bill’s example in part because the challenge was so difficult and thus illustrates in stark relief how crucial the coordinator’s role is to community success. Indeed, the community coordinator is the most critical factor for success at the community level.

Bill’s work started during what we call the “discovery” phase (see Exhibit 7), when the community is still an inchoate notion shared by a few pioneering colleagues. Bill found a couple partners to work with and they put together a comprehensive list of people in various regulatory agencies who were likely candidates—or who would know ones who were. Bill and his cohort developed an informal interview protocol and encouraged people to come to an initial meeting, which Bill organized and facilitated.

Bill understood—from his professional experience and years in government—that the first event would have to create immediate value or it was unlikely members would return. He was careful to develop an agenda that would produce valuable take-aways the first time, while outlining a vision of where this group could go. He facilitated a discussion with members—approximately 20 in the first meeting—about agenda items for future meetings and later worked with a core group of 4-5 members to help set up regular community meetings.

Managing the logistics of community activities is not a sexy process, but without it events don’t happen or they fail to capture member interest. A key task is recruiting speakers—whether outside experts, influential stakeholders, or members themselves. The coordinator needs to prepare speakers to target their comments to members’ interests and to keep their initial remarks brief in order to leave plenty of time for give and take in an interactive session. Bill actively sought out speakers who would engage community members and help to increase the visibility of the community’s work. For example, he arranged for meetings that included the head of the CIO Council who spoke about the future of e-government; staff from Senator Lieberman’s office who provided a congressional perspective on e-government initiatives; and an important meeting with the National Archives and Records Administration to learn about their plans regarding digital-format standards that they would require agencies to meet. Organizing events such as these is no trivial matter, and the coordinator risks both time and reputation with colleagues if they don’t come off well. In this case, all three of these events were very successful and helped to increase the reputation of the community and build participation.

48

Members will typically find community participation useful as long as competent peers show up with interesting questions and useful lessons learned about pitfalls to avoid or methods to try. Bill worked behind the scenes to encourage members from various agencies with valuable experiences to attend meetings. He helped make the value of participation visible by documenting the insights from sessions and sending them around to all the people who had expressed interest in the community. (This number was about double the number of people who typically showed up at a meeting. As coordinator, Bill was working with several member segments—including the core group, the active members who typically came to meetings, as well as a larger peripheral group that included a number of influential and very competent people who were important to the overall community’s vitality.)

Bill wanted to push the community’s development beyond a forum for talking about agency problems and approaches. Although these discussions were extremely helpful for members, he believed, and other core-group members agreed, that the community was likely to lose steam unless it could accomplish a project together. They wanted to establish the community as a more visible, influential forum for addressing e-regulation issues. This would help to demonstrate their value to sponsors in a concrete way and attract the type of skilled and influential members that the E-regulation core group wanted to engage. Such a project would also help to build member relationships and new skills and insights.

Bill led a team of members from seven different agencies to provide one-stop access to regulatory forms for citizens through the public “FirstGov” portal. The team defined the value for citizens and participating agencies, outlined the tasks required to set a common standard for formatting the records, and began negotiations with GSA, which managed the FirstGov site. About this time GSA was making leadership changes and redirected its staff away from the E-Regulation project. This was a great disappointment to the group’s project team and took the steam out of the community’s momentum.

One of Bill’s functions was to take a “meta” perspective on the community. He was disappointed that GSA withheld its support for the initiative and began talking with community members about how they should respond. He sought out perspectives from other coordinators in the federal government as well as external experts to get ideas. Ultimately, he found that his dilemma was not unusual—without some level of sponsorship or extreme urgency among practitioners to connect, it would be difficult to sustain participation in a cross-agency community. As it happened, about this time the agency leadership changed under the new Administration and Bill lost agency support for his time allocation to the E-Regulation initiative. (He is trained as a lawyer and was assigned to a large regulatory case.)

Bill continued to advocate for the community, nevertheless. He encouraged others in the core group to try to take up the slack, but ultimately no one was able to do it. Members had remarked since the beginning that Bill’s role had been crucial. Bill has continued to

49

connect with colleagues from the E-Regulation community on an informal basis, and with other coordinators and external colleagues through occasional gatherings of the “Knowledge Management” community within the federal government. Although he is committed full-time now on other duties, he remains passionate about what cross-agency communities such as E-Regulation can accomplish.

This review of Bill’s activities as coordinator suggests the level of expertise, credibility, and persistence this role requires—often under conditions where there is little credit given to acknowledge its contributions. After all, the work is often done outside the explicit scope of the coordinator’s “day job” in the organization. Furthermore, the value of the community is distributed across agencies—and thus generally gets little recognition from executives who do not consider results associated with interagency collaboration.

Key skills for a successful coordinator are illustrated in this profile, including: interpersonal skills—promote the opportunity, recruit peers to contribute,

connect people institutional savvy—liaise with sponsors and stakeholders, invite the right

people to keynote at meetings, understand the politics of agencies group facilitation skills to keep meetings on track and address conflicts

among members—as a group and “back-channel” passion and motivation to make the community a success—strong personal

interest in the topic and in community leadership as a professional development opportunity

credibility among peers as leader and someone with relevant professional expertise

a motivator and coach for others—recognizing others contributions and importance to the community

Finally, it bears emphasis that the coordinator’s success—and thus the community’s—depends on more than natural ability and passion for the work. As this case shows, it is crucial for sponsors to provide sufficient staff time and funding to support a community coordinator and to provide sufficient legitimacy and influence so the community can take on substantive projects with the expectation that results will be implemented. And sustained sponsorship is a particular challenge in a government context where many leaders are politically appointed. Unfortunately for the E-Reg community, Bill Bennett lost management sponsorship for his participation when agency leadership changed with the new Administration.

A member succinctly stated the linkages between community leadership, sponsorship, and success:

“If Bill Bennett can’t devote energy to it, it will probably die. I can be on the core group, but it’s been on my own time; without management providing me time, then I can’t give enough time to coordinate the community.”

50

Exhibit 11Exhibit 11Portrait of an support-team leader: Portrait of an support-team leader:

Mike BurkMike Burk

Mike Burk plays an instrumental role with the Federal Highway Administration (FHWA) in enabling the development of communities. At the strategic level, he works with senior line managers to outline a vision for a knowledge initiative. In running the knowledge initiative, he plays a leading support-team role to help communities get up and running—coaching coordinators; providing tools and templates to guide community activities; and measuring results. Finally at the infrastructural level, he suggests organizational design changes and coordinates the development of a technology infrastructure to support knowledge sharing and capture—a particularly important piece for the FHWA communities.

At the strategic level, Mike worked with senior management at FHWA to think through how a community-based knowledge initiative could address key strategic objectives in the organization—such as reducing highway fatalities by 20% over a ten-year period (from 1998-2008). Mike advised the senior management on where to focus an initial effort to demonstrate how communities of practice could help. They selected a domain where there was a clear opportunity for early, measurable success (in this case, related to the diffusion of rumble strip applications across the United States). In fact, one of his roles was to help form a senior management sponsor team for the initiative; then advise them of their leadership role and outline a vision of a broad-based community infrastructure that could support strategic capabilities in a number of areas. Part of this recruitment effort was about educating sponsors—by providing case examples from external applications, outlining a vision of how it could work at FHWA, and arranging for sponsors to talk to peers in other organizations about what the process would entail and what it could achieve.

Mike outlined how the community would work—who would be involved (both FHWA employees as well the state-level DOT safety engineers and other stakeholders. He helped executives understand the typical development process for a community like this one, where there was a pre-existing informal network—but no systematic effort to share ideas and approaches across states in order to accelerate implementation. Mike developed a project plan to document what a roll-out of several communities would mean in terms of staff time, infrastructure and support team expenses, and the potential for strategic impact. He also explained that given the nature of communities, it was likely that bottom-up grassroots communities would emerge in addition to the strategic communities that were identified at the executive level.

At the community-support level, Mike organized efforts to recruit, train, and provide ongoing coaching to community coordinators—and lobbied with the agency sponsors to set aside sufficient time for the coordinators to focus on their community (as much as

51

three-quarters time during the early stages of the community’s development in the case of the Rumble Strip community).

Mike leveraged his direct work with communities to codify templates and tools for future communities wherever possible. For instance, the design of the Rumble Strip website—its various functions, the look and feel of the design, and pointers about how to use the site—were codified and then rolled out to new communities as they came on board. He developed tip sheets for community coordinators to help them structure and facilitate member meetings, as well as training them on strategies for moderating the website interactions and repository so they were most useful.

A key function that Mike played was to help communities document their achievements and make them visible—for current and potential members, as well as for sponsors and other stakeholders who could help renew or expand the initiative over time. For example, Mike and his team interviewed participants both within FHWA and in the state agencies in the Rumble Strip community to capture stories about how the community had created value for members—saving time on designing a plan and specifications for implementation, for example, or providing evidence to persuade policy makers. He also leveraged available research on Rumble Strips to make the causal links between the community’s contribution to faster implementation and hard numbers on subsequent reductions in actual fatalities—as well as associated economic costs to citizens and states.

Finally, at the infrastructural level, Mike planned the development of an IT platform to support the work of communities. On this platform, members of his team designed a web-enabled space that would allow community members across government levels and sectors—including private citizens and university researchers and policy makers—to participate. After the Rumble Strip community had provided a “proof of concept,” Mike and his team planned and implemented an agency-wide platform. This technical infrastructure facilitated rapid and low-cost launches of multiple communities—each of which used an established communication and knowledge-management template built into the system. The common platform helped to integrate communities by enabling members to move easily from one community to another.

With regard to organizational infrastructure, Mike began helping leaders think through implications for changes in formal organizational design elements to help scaffold community success in the long term. There are a number of formal design elements that typically need adjustments to account for the unique nature of communities. For example, appraisal systems often do not include—or prioritize—peer-to-peer contributions or contributions to the organization’s knowledge base. Often employees are reinforced for hoarding knowledge that gives them a performance advantage over peers. As mentioned, Mike already had begun to create new types of measures to document a community’s contributions. This evidence would be important for dedicating staff time to coordinate and participate in community activities. Creating a sponsor board—called a Knowledge Management advisory group at FHWA—is in itself a design intervention establishes a knowledge-focused management structure than complements traditional decision-making bodies focused on more immediate service-delivery

52

functions. Putting community activity and contributions on the management radar screen ultimately can produce new insights about how to get the work done faster and better—while helping executives see emerging strategic opportunities and take advantage of them.

As in the case of the community coordinator, key skills for a support-team leader are interpersonal skills for liaising and educating sponsors and stakeholders. It is also crucial for the support team to have broader strategic and organization design competencies in order to lead the development of an overall knowledge initiative that includes the roll-out of a number of communities over time.

The support team leader is not only a professional expert, but also plays an important role as a persuasive advocate for community-based initiatives to build capability and collaboration. Community-based initiatives are still rare in government and are generally unfamiliar to managers and members alike. Organizations will need articulate evangelists who can promulgate what communities can do (and are doing)—as well as provide expert advice on how to get started. Participants and sponsors will then need handholding and encouragement along the way to get through the inevitable glitches, or in cases where communities simply fail to thrive despite best intentions. The support team leader should have a balanced perspective that includes a passion for what communities can accomplish as well as a grounded appreciation for the challenges of such an ambitious initiative in the context of a typical traditional organization with a normal distribution of skeptics, pioneers, and majority in the middle.

Finally, as is the case for coordinators, the support team cannot succeed unless the leader and other members have sufficient time to do their job, access to external expertise, and top-level legitimacy for their work—including influence with line managers. Ultimately, line managers determine whether practitioners will have time to lead or participate in ongoing activities or special projects—and whether ideas and practices that come out of the community will be used to improve results. Here again, we see that a successful community initiative requires an ecology of leadership in various roles and that adequate executive sponsorship is the “keystone” species that makes this leadership ecology vital and effective.

53

Exhibit 12 Exhibit 12 Portrait of an executive sponsor:Portrait of an executive sponsor:

Shereen RemezShereen RemezDr. Shereen Remez, formerly the director of Communications for the General Services Administration (GSA) and later GSA’s CIO, was appointed to the first Chief Knowledge Officer (CKO) position in the federal government in 1999. As Shereen acknowledged, neither she nor her boss—GSA’s CEO—knew precisely what the role should be about, but as she wrote later: “I was given the gift of freedom to explore and experiment with this new idea called Knowledge, and the responsibility to help craft GSA’s future.” 13 Shereen was somewhat isolated in the federal government; a survey of federal agency CIOs that year did not even mention knowledge management in its top ten issues. She refers to her experience launching the CKO office at GSA as “an adventure in extreme learning.” (Of course, since 1999 Knowledge Management has become a buzzword nearly as much in government as in the private sector, where it is a multi-billion-dollar industry. And yet, CKOs in organizations are still in the early stages of figuring out what the role should be about.)

Shereen recognized from the start that nothing significant in the area of knowledge management would be accomplished without a broad base of management buy-in and willingness to invest time and attention to lead substantive projects. Her strategy was not to build a KM empire in headquarters, but rather to “light…little sparklers among interest groups across GSA’s geography that would one day become multi-colored and elaborate displays in the night sky, and elicit a few ‘ooh’s and aah’s’ from management.”

Like communities, initiatives go through stages and the role of executive sponsors evolves as communities mature and as the broader initiative unfolds. In the early phases of a community initiative, the key is to identify key strategic opportunities that communities can address and put an infrastructure in place to get the initiative going. Getting from zero to thirty is the hardest part—so it is particularly important for senior managers to pay attention at this stage.

1. Typical sponsoring activities in setting up an initiative: Convene potential sponsors and champions and review initiative purpose, process,

and benefits Identify strategic areas appropriate for early communities to build capacity Establish key roles and recruit leaders: sponsor board, agency champions, support

staff, and community coordinators

Shereen focused her attention on helping managers lead pilot initiatives that had a clear business purpose and that included a commitment to measure results. For example, one pilot focused on sharing best practices among a community of building managers across the country and resulted in significant increases in customer satisfaction nationwide. She 13 This quote and others in this Exhibit are taken from: Remez, S., 2000, “The GSA story: Swimming with Dolphins.”

54

also established a small team to support communities throughout GSA and sent them to conferences to learn from the experiences of other organizations in the private sector. The team provided consulting and logistical support to managers and helped document and publicize the success of several pilots that were led by early adopters among business-unit leaders.

2. Typical sponsoring activities in the early formative and pilot stages Help community identify strategic capability issues related to a targeted domain Encourage members to lead and participate Support development of initial community design—issues to address, member

participation and roles, and initial learning activities Support the development of technological infrastructure

Shereen’s office helped launch several communities of practice in core competence areas such as procurement, IT, customers, workforce development, and KM. In addition, because of Shereen’s role in the KM working group of the CIO Council the team also supported a number of inter-agency pilots, among them the E-reg community (see Exhibit 2). They organized seminars for community leaders to meet and learn from each other and from experts about the art of community development in organizations. Her office also coordinated the development of a technology infrastructure that would enable these and future communities to extend their reach and impact.

3. Typical sponsoring activities as a community initiative matures Help communities take on active stewardship of knowledge in their domains Systematize the use of communities as organizing units Establish measures and budgets for resources needed for community activities and

project proposals More generally, consider organizational changes required to support communities’

development and institutionalize their role in the organization

In her function as CKO, Shereen became an advocate for communities of practice in the organization (in GSA and in the federal government more generally via the CIO Council). She saw these communities as key to a vision of GSA as a knowledge organization whose mission was increasingly about providing knowledge and other value-added services related to building management, not merely providing physical space to tenants. Her role then was both to remove obstacles to the development of communities of practice and to listen to communities and engage them in strategic thinking about the place of their knowledge in the evolution of GSA services. Indeed, in later stages communities themselves can discover strategic opportunities and practical ways to take advantage of them. At this point, they help generate a virtuous cycle—increased strategic value, leading to more sponsorship attention and resources, creating new opportunities—and so on. The greatest contribution of a senior-level champion is getting the community initiative off the ground so it can achieve this self-generating state of strategic value creation.

55

4. Typical sponsoring activities to sustain a community initiative long-term Establish and recognize community participation as part of work Leverage and institutionalize the role of communities to build, share, and apply

capabilities Identify where to cultivate new communities and how much to invest Ongoing leadership to adjust organization design elements that affect community’s

ability to get results, such as how staff are recruited, hired, trained, promoted, and rewarded

Negotiate with community on strategic development and on transition: formalization, sub-division or dissolution

The role of executive sponsor is one that entails a steep learning curve for most managers. In a paper she wrote on the subject, Shereen shared a number of lessons learned from her experience (and thus modeled her commitment to knowledge transfer for her own domain). Several of these points reinforce her strategy as CKO to be a catalyst that inspires and enables others to take up the knowledge management challenge. Here’s a sample:

- “You are courting failure if your top people don’t get it.”- “Light fires, sparklers, and spread the word—don’t mandate.”- “There has to be a business reason to share, and it helps immensely if you can

measure it.”- “Scale up the promising idea that works with rather than railing against the

prevailing culture.”

Shereen concluded from her experience that leading a knowledge-organization transformation is not a lonely individual effort to push a rock uphill, but more like “lassoing the future and pulling yourself and your colleagues toward it.” Her story provides a compelling model of a CKO dealing with a difficult combination of much uncertainty and high aspirations. She worked hard to generate a constituency of knowledgeable leaders throughout GSA by building on energy where she found it and linking initiatives to business results. She leveraged the expertise of external experts both inside and outside the federal government to accelerate her own learning curve and those of her internal clients.

Shereen compared her foray into uncharted waters as akin to “swimming with dolphins”—an appropriate metaphor, given dolphins’ distinctive intelligence, sociability, and exceptional ability to flow amidst turbulent waters.

56

Exhibit 13Exhibit 13Communities complementing projects: Communities complementing projects:

The E-government caseThe E-government caseThe E-Reg and E-Records communities described in Exhibit 2 illustrate how community structures can complement conventional structures in organizations—in this case the project structure. The federal government’s E-government mandate has been taken up by a formal “E-governance” project structure led by an OMB Associate Director. (See Figure 12.1.) The E-governance project structure consists of four major steering committees (in addition to the inter-agency CIO Council). Each of these committees has responsibility for one of several portfolios of E-governance projects (e.g., the “Government to Citizen” portfolio). Each project cluster, in turn, includes several E-government initiatives, which are led by “integrated project teams” (IPTS) whose purpose is to develop and implement new E-governance approaches across agencies. The “Government to Business” steering committee oversees projects related to the domain of the E-Reg community, and the “Internal Effectiveness & Efficiency” committee includes a focus on “E-Records Management.” The E-Governance project scope thus overlaps with the domains of both the E-Reg and E-Records communities of practice.

E-GOVERNANCE STRUCTUREPRESIDENT'S BUDGET

FINAL FUNDING APPROVAL -- OMB DIRECTOR

OMB Associate Director ForInformation Technology and E-Government

GOVERNMENT TO BUSINESS

PORTFOLIO MANAGER

GOVERNMENTTO GOVERNMENT

PORTFOLIO MANAGER

GOVERNMENT TO CITIZEN

PORTFOLIO MANAGER

INTERNAL EFFECTIVENESS &

EFFICIENCY PORTFOLIO MANAGER

IPTS FOR:•GEOSPATIAL INFORMATION ONE-STOP

•DISASTER ASSISTANCE•E-GRANT•WIRELESS•E-VITAL

IPTS FOR:•RECRUITMENT ONE-STOP•ENTERPRISE HR INTEGRATION

•INTEGRATED ACQUISITION SYSTEM

•E-RECORDS MANAGEMENT•E-TRAINING•PAYROLL INITIATIVE•E-TRAVEL

PORTFOLIO MANAGEMENT OFFICEMANAGEMENT OF TRANSFORMATION

CIO COUNCIL

STEERING COMMITTEE

STEERING COMMITTEE

STEERING COMMITTEE

STEERING COMMITTEE

IPTS FOR:•E-AUTHENTICATION•ENTERPRISE ARCHITECTURE ASSESSMENT

IPTS FOR:•USA SERVICE•EZ TAX FILING•ONLINE ACCESS FOR LOANS

•RECREATION ONE-STOP

•ELIBIBILITY ASSISTANCE ONLINE

OIRAIPT

LEADERS

IPTS FOR:• FEDERAL ASSETS SALES• ONLINE RULEMAKING MANAGEMENT

• SIMPLIFIED & UNIFIED TAX & WAGE REPORTING

• ONE STOP BUSINESS COMPETITION

• INTERNATIONAL TRADE PROCESS

• (CONSOLIDATED HEALTH INFORMATICS)

IPTS = Integrated Project Teams Source: “E-government governance policy” overview document

Figure 13.1. Formal E-Governance project structure for U.S. federal agencies

57

The overall E-governance structure (like Homeland Security and various interagency councils, is an example of a formal structure created to foster inter-agency collaboration) sets rigorous criteria for agencies to participate in various capital programming initiatives. Design elements of these project-level initiatives include highly specified workplans, staff assignments, and well-defined targets, which are subject to OMB review.

The project structure is an appropriate response to the dynamic nature of the E-government challenge. Indeed, project structures have proliferated in the last 15 years for good reasons, such as increased agility, flexibility, and innovativeness—to the point that nearly half of employees in many organizations are staffed on projects at any given time. But the strengths of projects—consisting of temporary teams tightly focused on achieving a particular outcome—create new challenges. Each project team tends to be so focused on its own deliverable that broader issues of knowledge sharing and coordination are not addressed optimally.

The E-Reg community was convened to tackle the knowledge and coordination issues that the E-governance project team and internal agency units were unlikely to address effectively on their own. The E-Reg community complements the E-governance project teams in a number of ways, including (a) identifying emergent issues, (b) broadly sharing and applying tacit knowledge, and (c) providing coordination and support across diverse agencies.

(a) Identifying and addressing emergent issues: How keep up with the dynamic range of issues that are emerging, including those that are particular to a sub-set of agencies or which manifest differently under diverse agency conditions?

Some agencies, for example, are far ahead of their peers and more interested in advanced issues such as data-display, trend analysis, and security (“We have finished modernization, but now need to find ways to add value to documents”). Meanwhile, others are simply trying to get the data from paper or word-perfect format into XML format; and security issues loom large as documents are converted and stored as digital files (“A security problem here would shut us down from December to March”). Finally, some agencies have only begun to understand the implications of the mandate. It is hard to determine top-down specifications when individual agencies in many cases have not got a handle on their own problems (“Our agency has lots of different internal programs to consolidate—and we have so many internal IT shops it’s overwhelming and hard to prioritize”). The advantage of a sponsored project team is its ability to tackle a well-specified problem whose solution has broad applicability. But for any serious problem, there remain hundreds of issues that are less tractable and do not lend themselves to one-size-fits-all solutions but can profit from collective thinking as enabled by a community. (“I find others with similar problems and concerns, and even if not solved yet, this helps; find out a lot about what is going on that I didn't know about.”) For example, several agencies may be saddled with legacy systems whose data is particularly hard to convert; or the barriers to implementation may have more to do

58

with skittish managers who would rather avoid the issue (“We want to create a cradle-to-grave electronic system, but there’s internal agency resistance”). Issues such as these are pervasive in any organization and create significant drag. Yet these are issues that a broader community can address, without necessarily making a project of them—thus complementing formal initiatives to address issues that call for a project-based approach.

(b) Broadly sharing and applying tacit knowledge: How capture practitioners’ tacit expertise—which cannot be codified in a standard or procedure—in one agency so it is available to peers in other agencies?

Regular interactions among project-team members can build a shared understanding of the solutions they develop, but practitioners from various agencies who are not actively involved will not understand the nuances of the project’s outcomes. When it comes to adapting tools and methods to local conditions there are no shortcuts to peer-to-peer conversations with practitioners in the field. Furthermore, much of the knowledge practitioners need is not project-type knowledge, but simply tips to help avoid pitfalls and get started in the right direction. (“We learned from Agriculture about using websites—learned what to ask and not ask.”)

A community provides an open network for any practitioner in the field to join in order to keep up with a project team’s work and to find colleagues who can help implement new approaches. The members of the E-Reg community share a set of issues that provide a background to their exchanges even though they do not work on the same project. Because of this common background, which is largely unspoken and assumed, they can exchange ideas and tips in very efficient ways. When someone comes in with a problem and asks the community for help, it only takes a few minutes for members to grasp the context and go to work. Because community participation is voluntary, members “vote with their feet” based on the value-for-time they experience. Participation is not required; nor is it restricted (as when a project manager must limit staff time to control expenses). This way, knowledge can circulate widely, beyond the reach of any project team. (“The federal government is so huge you need groups like this to find peers with information and knowledge.”)

(c) Coordinating and supporting implementation: How foster ongoing coordination and learning after project teams are disbanded and implementation is underway—and as the original issues morph and require continuous adjustments under varied conditions across agencies?

Initially, there was no forum for coordinating E-Records management across agencies. The E-governance project team on records management has provided a context to work on specific issues, but it does not provide an open-network forum for agency participants to coordinate efforts—such as working with vendors on contracts, educating managers about implementation issues, identifying opportunities to work together.

59

Meanwhile, nobody at the table is quite sure which format a key player in the field—the National Archives and Records Administration—will require. And it’s not clear that the NARA can—or should—make this decision without much more informed and inclusive participation with agencies, including those in this community. (“NARA has not been willing to take the lead. Agencies going ahead and making best guesses, but there is no consistency. Then NARA comes out after the fact and says you guessed wrong—‘you need to have a higher level of resolution.’ But this slows things down for agencies. We need to create a forum to work these things out.”)

The example of the E-Reg community has emphasized how communities can complement projects. More generally, when planning project initiatives it is important to ask up front: what types of structure are best suited to the various elements of an effective, sustainable solution to this problem? How much of the problem lends itself to an established unit structure (accountability for implementation); a project structure (solving problems, creating a new approach), or community structure (stewarding the tacit knowledge and cultivating peer-to-peer networks that will facilitate effective implementation)? Communities, for their own part, conduct a corresponding deliberation: is this a problem to be raised with a line unit—or something we can address ourselves? Given how widespread and urgent it is, should we launch a project to work on this issue—and what sponsorship will we need? Who should we be coordinating with?

Both formal project-based and informal community-based approaches have strengths and weaknesses. Skilled managers today are already well-versed in conventional formal approaches, but they are generally not so aware of informal mechanisms for spurring innovation and knowledge-sharing. This is particularly important because community approaches work best for types of management problems that are becoming more common—that is, problems that are messy, dynamic, and decentralized. The point is not “which approach is better,” but rather “when to use which approach, and why?” And what combination of approaches will leverage fully the complementarity between them?

Formally managed projects work best wheno problems can be clearly defined o reliable, quantifiable measurements are establishedo an authority structure is in place to ensure that project results get

implemented. Communities are most effective when

o problems are complex and dynamic or very situation-specific o measures require stories to link cause and effect o authority is decentralized and depends more on professionals’ intrinsic

commitment to getting results (versus extrinsic appraisals and incentives).

Communities and project teams, when combined to complement each other, provide a powerful structural response to vexing problems that don’t lend themselves to standard or stable solutions. While communities contribute new problem-solving skills and boundary-crossing relationships, the project structure provides legitimacy and senior

60

sponsorship. For instance, the E-Reg community has fallen on hard times—it lost its coordinator and an energizing project stalled when the sponsoring agency backed out—the E-governance structure may provide reason to hope for its revival. One member noted that:

“Everybody is now looking at President’s E-gov initiative; the OMB is going to force us to pool resources, and agencies are just waiting till then to move.”

But he argued that such an approach is not likely to resolve the types of issues covered here—addressing emergent issues, sharing tacit knowledge, and fostering broad-based, peer-to-peer coordination. He concluded with a recommendation:

“What the E-Governance structure should do is serve as convener for communities of practice on each of the E-government topics they want to address.”

Ultimately, the point is to consider a full range of organizing options—including departments, project teams, and communities of practice—to tackle the complex problems we face today.

61

Exhibit 14Exhibit 14Communities complementing programs: Communities complementing programs:

the education casethe education case

Communities of practice provide new opportunities to improve the results of legislative programs, whether they are new ones such as “No Child Left Behind” or established ones such as Head Start. Staff expenses for agency-administered programs generally pay for oversight, assessment, and technical assistance. What if they focused more on building learning networks among grantees as well as among non-grantees with shared needs and capabilities? Communities of practice leverage the latent practitioner networks that these programs enact—networks that are now generally overlooked or under-developed. A program vision should not be limited to specified outcomes such as the number of students passing the grade on standardized tests. Rather, the vision should include a capacity-building goal to build a learning system that will continue to generate new levels of achievement long after federal funds have been reduced or cut off.

To show how a community-based approach could enhance program effectiveness, consider an illustrative program designed to improve student achievement by fostering the professional development of school leaders—in particular, urban-school principals.

It is generally accepted that successful schools, like any organization, rely on strong leadership. School principals are particularly important because they have much influence over the professional development of teachers. They also set the tone for relationships with stakeholders (parents, district officials, etc.) and for the overall culture of the school.

There are 90,000 principals in the nation, and approximately 40,000 will be eligible to retire in 2005. Moreover, the schools most in need of effective principals, urban schools in poor districts, are ones that suffer most from highest levels of turnover and under-developed leaders (because schools of education generally do not prepare graduates sufficiently for the distinctive challenges of leading urban schools). While formal education and training is essential, the most reliable methods for developing highly effective urban principals—as is true for any professional—are informal learning activities that occur during internships and on the job. These include coaching and counseling by mentors and co-consulting, visits, and mutual encouragement among peers and colleagues.14

The application of a community-based approach starts in the program-design phase. The design process begins by organizing a series of gatherings that include leading researchers, practitioners, educators, school districts, and leaders of non-profit initiatives who are doing pioneering work in the area. Let us call this the “program design

14 See the development strategy described by a leading non-profit whose mission is to develop successful urban school principals—briefly outlined on the homepage of their website: http://www.nlns.org. See also: Fink, E. & Resnick, L.B., Developing Principals as Instructional Leaders,” Phi Delta Kappan, April, 2001.

62

community.” The purpose of organizing the group is to consolidate an understanding of the range of initiatives underway today. But this is only a start. The expectation is to build a national community of practitioners in the field of professional development for school leaders. Program design is seen here as a continuous process, with expected renewal along the way, under the stewardship of an ongoing community of leaders in the field.

Communities are also important during the pilot phase of program implementation. Let us assume that 12 districts are chosen as initial program sites to focus on dramatically improving school-leadership capabilities (and related student-achievement results). Representative leaders from each district are invited to organize as a “program implementation community.” Federal agency staff provide sponsorship and coordinating roles. The implementation community includes a cross section of district stakeholders: principals, district professional development staff, teacher leaders, parents, city officials, and others. The purpose of the implementation community is to help members build, share, and apply what they learn through experience in the program. Collective insights influence the actions of both agency staff and local participants.

During the start-up phase, community members meet together face-to-face and virtually to compare how they are developing leaders now, talk about options they are considering, and consulting with each other about how best to leverage program dollars and federal resources. The group negotiates program changes and adjustments when appropriate. Regular forums provide for members of both the “program design” and “program implementation” communities to share ideas and incorporate these into the program as it evolves over time.

Once the pilot initiatives have been launched, community members continue to meet on a regular basis to talk about what results they are getting, what questions and problems they are facing, lessons learned, and ideas they have for advancing their work or helping others. Again, at both national and district levels, the objective is not only to achieve specific program results—better student achievement through professional development of urban principals—but also to build a sustained learning system to continue the work and expand it to a national scale.

Over time, the program becomes more about sponsoring a nationally distributed community on the topic of school leadership. The community itself would grow along the lines of a typical association (such as the confederation of rotary clubs)—with local chapters of about 12 districts, each of which is connected through regional and national networks. Community activities are similar at various levels—local, regional, or national. They include monthly teleconferences, listserv announcements, a website repository with an interactive component, site visits, projects, and annual face-to-face gatherings.15

15 See Exhibit 6, “How communities learn.”

63

Ultimately, the community model might be applied at levels that get closer and closer to students themselves. Districts might promote leadership development by cultivating communities of principals to build skills, identify school-design changes, gain influence, etc. These may lead, in turn, to the development of teacher communities—led by “teacher leaders” and focused on what it takes to help students succeed. Finally, as some teachers are doing now, students themselves may be encouraged to participate in communities during the school year that focus on particular topics—from mathematics to art history—and apply both formal and informal learning methods to make the learning process more effective and engaging. Finally, other relevant communities include superintendent and parent communities as well as ones that combine various constituencies—parents, school administrators, city officials, business leaders, etc.—to take on overall stewardship for the domain of education as a civic discipline at the local level.16

The role of the government—and in this case the agencies that are running the program—is to provide overall sponsorship for the initiative: providing funds for community coordinators, a common communications infrastructure, influence to catalyze support from stakeholders, and legitimacy and visibility for what these communities are aspiring to achieve. In complementing the formal program, one purpose of the community is to fully leverage the funding allocated to specific districts involved in the program. Participation in the community as a capacity-building learning system should be a criterion for program participation.

The community-based program approach does not merely fund a particular solution to a social problem. Rather, it builds the capacity of a learning system to design and implement solutions over time—and tailor approaches to fit unique conditions in the local context. This approach integrates a community-building dimension into the program from the start. In this case, participating school districts are not viewed as isolated agents who receive funds and advice—but rather are in continuous contact with peers and other experts to accelerate learning and results.

This approach can be applied to any program whose intention is to improve a socio-economic outcome by helping distributed clients build new skills and capabilities. As the cases in this paper illustrate, the objectives can be as varied as highway safety, e-government, reducing gun violence, or leadership development. In all cases, specific solutions become less important as the community becomes more developed—and becomes mature enough to generate and diffuse innovations more quickly, better adapted, and more fully supported than any central entity could do. And as the SafeCities case illustrates, local participants can find participation in a community of practice as valuable as direct allocations of federal funds.

The SafeCities community featured in this paper (see Exhibit 3) provides a model of what a national community of practice looks like. In the SafeCities case, local coalitions (i.e.,

16 For example of such communities at principal, teacher, and student levels, see: (principals), Fink & Resnik, op. cit.; (teachers) [need a citation here—K. Belancik?]; (students), Guernsey, L., “A Young Writers’ Roundtable, via the Web,” New York Times, August 14, 2003.

64

city-based communities of practice), have organized to reduce gun violence at the local level. Each of these local coalitions, in turn, participates in a national community of practice. If SafeCities had received continued sponsorship, its plan was to spawn additional communities of about ten cities each—and to keep these connected as part of a national community on the topic of reducing gun violence (see figure 8.1 in Exhibit 8 on going to scale).

The SafeCities initiative, in fact, is a good example of a program that was designed from the beginning as an initiative primarily about building a learning system, rather than focusing on a particular solution, such as after-school programs, community policing, or gun-tracing methods. All of these activities were considered important, but the goal was to provide a forum for members to learn and connect in order to build capabilities in all of these areas and more—tailored to the local needs and opportunities.

Ironically, SafeCities was closed down in favor of a more traditional program, called “Project Safe Neighborhoods,” which provides funding for local initiatives in areas such as crime analysis and prosecution. While members of SafeCities would not turn down funding, they argued heatedly with agency officials that they were making a mistake. Agency leaders were closing down a remarkable forum that cost the government little more than one FTE to leverage the knowledge resources and leadership of high-level innovators from ten major U.S. cities to make a difference in the nation’s ability to reduce gun violence. (Granted, agency leaders were planning for Safe Neighborhoods to continue to integrate a community-based approach, but thus far they have not succeeded because they have not given the community dimension enough attention.)

The hypothetical education program described here and related insights from the SafeCities case illustrate how a community-of-practice approach can complement conventional programs. The community approach makes a number of compelling contributions: Program-affiliated communities of practice improve program design; leverage learning across grantees; accelerate implementation of innovations and achievement of results; and provide a platform for continuous innovation and roll-out on a national scale.

65

Exhibit 15 Exhibit 15 Communities complementing formal organizations: Communities complementing formal organizations:

the Homeland Security casethe Homeland Security case “It takes a network to fight a network.”17

Introduction: new threats require a stronger communityThe Homeland Security challenge is a combination of an increased external threat and a corresponding lack of organizational readiness to respond.18 The US intelligence community was not prepared for Al Qaeda on September 11—and still is not. Enemies of the US, in this case state-less terrorists, had increased their levels of skills, determination, and organizational sophistication beyond the intelligence community’s ability to protect the nation; and apparently the community itself was unaware how much ground it had lost to emerging terrorist organizations.

Mismatched performance requirement and internal capability:“There is consensus within the government on two issues: the terrorist attacks were brilliantly planned and executed, and the intelligence community was in no way prepared to stop them.”19

Increased performance requirement:“The relative immunity from international terrorism that America had enjoyed for many years was gone. Al Qaeda was also unusual in its dedication, size, organizational structure, and mission…became more skilled and attracted more adherents …”20

Insufficient internal capability:“The attacks of September 11 could have been prevented if the right combination of skill, cooperation, creativity and some good luck had been brought to the task.”21

* * *The 9/11 joint committee report provides insights into the barriers to cross-agency collaboration, and their effect on coordination and knowledge-sharing, and ultimately on the ability of government to fulfill its mission. The case highlights key issues that appear

17 John Arquilla, Consultant, Rand Institute; Professor, Naval Postgraduate School, as quoted in “America’s Secret Weapon,” T. A. Stewart, IssueBusiness 2.0, December, 200118 This is a classic case where an organization’s environment changes drastically, but only incrementally over time, and therefore the changes are not recognized until a tipping point at which disaster (in business, generally a bankruptcy or an unwanted acquisition) strikes. This is known in business strategy circles as the “boiled-frog syndrome”—where the water in the pail is slowly heated until the frog has begun to boil, at which point he is too feeble to leap out and gets cooked.19 Hersh, S.M., “Annals of National Security: What Went Wrong,” New Yorker, October 8, 2001: 34.20 Cited from declassified material in the July 24, 2003 report by the Congressional intelligence committees; from excerpts reprinted in the New York Times, July 25, 2003: A12-13.21 Senator Bob Graham, as quoted by D. Johnston in “Reports of 9/11 Panel Cites Lapses by C.I.A. and F.B.I.” New York Times, July 25, 2003: A1.

66

in many of the ad hoc or intentional government communities we have studied—including ones focused on E-government, acquisition, public safety, workforce development and others. But here the stakes are particularly high—and it is urgent to begin responding now, and leveraging new tools to do so. There is much to learn government-wide from a high-profile, concerted, community-based initiative to build world-class cross-agency capability and coordination in the US government.

The Homeland Security case raises the visibility of the organizational challenge facing all agencies—indeed, all organizations. The challenge of managing in the new era is about managing across dynamic boundaries of various types—organizational, geographical, disciplinary, and cultural—and it is about managing knowledge as well as business processes. But the structural implications of these new management realities are only now becoming clear—and the challenge to respond is a tough one. Given the typical inertia of organizations, it is rare that they are able to achieve transformative changes, even when their survival is at stake. Homeland Security is a case that demands a dramatic increase in capability—a breakthrough that could serve as a compelling model for organizing a 21st Century government.

From an organizational perspective, the homeland security problem is not only about defending the United States from terrorist attacks. It is also about building the organizational capacity to learn and innovate in order to address threats we cannot now imagine. As in the Challenger disasters of 1986 and 2001, we need to address—and leaders are now appropriately focusing on this—the fundamental organizational capacities that have created these lapses. Indeed, these organizational capabilities are crucial for addressing our national priorities across the board.

Current analysis and related responses to the homeland security problem include reorganizations, training, staffing levels, and oversight. While these will certainly have impact, they are not nearly enough. We need to open up new degrees of freedom regarding both the analysis and options for action. In particular, we need to better appreciate the importance of informal phenomena and related organizing approaches to improve cross-boundary collaboration and collective knowledge stewardship.

There is a long tradition of research in organization studies about “informal” phenomena, but the practical implications of this research is often neglected because it is harder to put into practice than corresponding formal solutions. For example, organizations pay less attention to leadership development than to conventional management training; to group conflict vs. goal setting; and to cultural issues in a merger vs. the numbers on increased market share. Informal phenomena associated with communities of practice include: professional identity and passion; trust, reciprocity, and belonging (“social capital”); and the meaning and artistry in expertise (that which cannot be standardized or codified). These cannot be “managed” in the same way that we manage information systems or customer service procedures, but there is nevertheless much we can do to influence informal phenomena. They are crucial not only because formal changes won’t work without related changes in the informal dimension—but also because informal changes alone, if handled well, can influence performance far more than formal changes.

67

This brief review of the Homeland Security case begins by describing the nature of the challenge in terms of several organizational issues: departmental fragmentation; skill gaps; and missing resources. Then it compares formal and informal responses. The current formal response addresses these mainly in terms of the presenting problems themselves—creating new structures that remove formal boundaries, offering new rounds of training, and increasing staff allocations. But there are also complementary and powerful community-based approaches that have not received as much attention: connecting across boundaries rather than eliminating them (many, after all, cannot be eliminated); building learning systems for capabilities that training alone can’t teach; and addressing underlying reasons why resources get misallocated. We refer to these latter approaches as “informal” because community structures depend more on members’ professional passion, relationships, and desire to learn than on any formal-system specifications. This analysis argues that we must use both “hard” and “soft” lenses to achieve an optimal organization design.

Caveat: This analysis is based on a limited knowledge of the topic, primarily gleaned from news accounts of the story and excerpts from the July 24, 2003 report of the Congressional intelligence committees. The objective of this analysis is to illustrate the potential of an “informal” community-based approach to supplement more typical formal responses. The ideas and implicit recommendations outlined here are thus speculative rather than conclusive arguments or proposals.

Organizational weaknesses—fragmentation, skills, and resourcesAn analysis of the homeland security problem suggests several key organizational issues in the US intelligence community that need to be addressed: fragmentation, skills, and resources. The deficiencies in these areas provide basic specifications for the organizational structures we need now. While we consider each of these issues separately, clearly they are related. Fragmentation prevents an organization from leveraging both its skill base and resource allocations; insufficient resources can limit access to and development of skills; and deficient skills mean botched problem solving and increased rework, thus draining resources. We take up each in turn—noting key issues, outlining formal responses, and describing complementary informal, community-based approaches that should be considered. The point here is not to say that formal methods are wrong, only that they are insufficient; in any case, their influence can be much strengthened when combined with complementary attention to the informal organization.

Fragmentation: among agencies, levels, and disciplinesFragmentation is defined in this case as the result of real or perceived boundaries that have hindered intelligence community members from sharing information and collaborating in order to identify terrorists and ward off attacks before they happen. In fact, boundaries are endemic to organizations—creating boundaries is what “organizing” is about. The question is where to draw the boundaries—and how to manage them. The challenge of operating in today’s fluid markets—whether providing consumer products or homeland defense—is to shift structures to match demands. While matrix-type structures have not panned out, project-based structures have proliferated because they provide

68

temporary boundaries to solve immediate problems. Communities of practice, meanwhile, provide yet another structural option—an informal one that complements formal boundaries without creating the dual-accountability problems of matrix structures.

There are several types of fragmentation—and compensatory boundary-crossing approaches—that a homeland security organization must address: inter-agency, inter-level, and inter-disciplinary

Inter- (and intra-) agency boundariesBoundaries among agencies—particularly the CIA and FBI, but also including nearly a dozen other agencies in the “intelligence community”—have received the most attention, and have been a prime focus of the reorganization effort. Intra-agency boundaries—both across divisions and levels of authority—are nearly as thorny. (These have not got as much attention from Congress or the press, perhaps because the problems with them are more familiar and messier to address.)

The US “intelligence community” includes a broad range of agencies that collect, interpret, and act on foreign and domestic intelligence. These include the NSC, CIA, FBI, DOD, CTC, NSA, US Military services, and others. (Ultimately, of course, the community also includes state and local authorities nationwide and government agencies in countries around the world—as discussed in “inter-level” section below.) Among these organizations, there have been long-standing institutional and cultural barriers that have hindered communication, cooperation, and knowledge sharing—and which have been blamed for our inability to prevent the 9/11 terrorist attack. The Congressional Report on 9/11 found that

“[t]he CIA and FBI did not talk to one another at critical junctures, threats and warnings were sometimes ignored, intercepted conversations between terrorist suspects often went untranslated, and American officials missed changes to ‘unravel the plot’ before it occurred.”22

In fact, it was only months after the incident that the intelligence community realized it had the knowledge it needed to prevent the disaster. As one senior law official said:

“As the information was gathered over time—like a collection of puzzle pieces, by a number of agencies over a period of time—no one person or agency had the complete picture on September 11 we have now.”23

The fragmentation across agencies is exacerbated by fragmentation within agencies, an issue that has not received as much pubic attention, but which may be just as responsible for missed opportunities to prevent the attacks. Geographic and cultural boundaries within the FBI meant that suspicious observations of pilots learning how to take off but not land were not heard in headquarters; or if they were, they were ignored until it was

22 Quoted from a news analysis of the Congressional intelligence committee report, by E. Lichtblau, “On Terror, Doubts Anew,” New York Times, July 25, 2003: A1423 D. Johnston in “Reports of 9/11 Panel Cites Lapses by C.I.A. and F.B.I.” New York Times, July 25, 2003: A13.

69

too late.24 Furthermore, geographic and organizational boundaries among distributed offices also cause problems, and are perhaps simply endemic to any distributed organization. After the original 1993 World Trade Center bombing,

“…many other FBI offices around the country were unaware of the magnitude of the threat….Richard Clarke contended that FBI field offices, except New York, were ‘clueless’ about counterterrorism and Al Qaeda and did not make them priorities.”25

Inter-level (and inter-sectoral) boundariesGovernment levels—at local, state, national, and international levels—are always difficult to bridge. While these also have received less attention than inter-agency boundaries, learning to manage them better may in the end be much more important to our national security. Although we are the enemy of terrorists as a nation, the effect of an attack is ultimately local. In fact, the definition of “homeland” in homeland security must be interpreted at several levels. The hole in lower Manhattan is a global concern, a national wound, and a local reality.

Furthermore, as we address the problem at different levels we see more clearly that it is not only a governmental problem, but also one that crosses sectors as well as levels. This is particularly clear at the local level, where it is impossible to plan for security without including an array of key stakeholders—businesses, hospitals, police, fire, schools, city and state agencies, neighborhood groups, non-profits, and others. The nature of “security” is not only about police or military response, but also about the capacity of local hospitals, businesses, schools and others to coordinate rapidly and effectively under great pressure—and of course, their collective job is much more about preparing and preventing terrorism than it is about responding after the fact.

There are over 11,000 cities and 3,000 counties nationwide who must share responsibility for homeland security. Local cities and counties must now establish the capacity to prepare, prevent, and respond to unfamiliar threats of unprecedented danger—bio-terrorism, dirty bombs, suicide attacks, etc. How can we build local security capabilities all across the country quickly and effectively? How can we combine and coordinate the multitude of disciplines and organizations—businesses, agencies, schools, hospitals, fire and police, universities, and others—and also connect across local, state, and federal levels where needed? How can we build a sufficient multi-level capability that includes new databases, protocols, technologies, simulations, standards, case studies, and research? A recent study commissioned by the Centers for Disease Control and Prevention and the RAND Corporation found that most emergency workers in the 40 cities and towns that they surveyed

24 This intra-agency, line-management communication problem, of course, was also fatal in the two Challenger disasters, which have received much public attention for the underlying organizational failures associated with them. In the Challenger disasters—1986 and 2001—issues raised by front-line engineers were not heard—regarding cracked O-rings and falling bricks of foam insulation respectively. 25 July 24, 2003 Report by Congressional intelligence committees, op cit., A12.

70

“…feel vastly under-prepared and under-protected for the consequences of chemical, biological or radiological terrorist attacks.”26

Bob Herbert, in a New York Times editorial on the topic, goes on to quote Dr. Redlener, the founding director of the National Center for Disaster Preparedness:

“My biggest concern….is that now, nearly two years after 9/11, the hospitals and public health systems are absolutely unprepared for another major act of terrorism. There's been very little improvement from two years ago. No one's really even defined what we mean by preparedness.”27

Finally, while the effect of terrorism is ultimately local, the scope of “homeland” security must be international as well. The dirty bomb that threatens Boston will most likely come from a country like Uzbekistan; the security mechanisms need to start there to prevent a disaster here. The intelligence on a worldwide “state-less” organization that transcends country boundaries can only come from an equivalent network of agents working inside and across countries around the world. As John Arquilla, a professor at the Naval Postgraduate School, stated: “It takes a network to fight a network.”28

Inter-disciplinary fragmentationFinally, there is also a disciplinary fragmentation. Disciplinary boundaries, like ones defined by geography and levels—as well as the preponderance of organizational boundaries—are inevitable and even essential. We cannot focus research and practice development without creating boundaries somewhere. (While many physicists and historians have interests in common (hence the “history of science” discipline), they each have a distinct repertoire of skills and methodologies that take years to learn. Without specialization, the core problems in each field could not be solved, nor new ones discovered.) The challenge then is not to eliminate the boundaries, but to learn to bridge them. As in the “netwar” strategy used in Iraq, which created extraordinary and nearly seamless connections across the military services—we need to find ways to connect across boundaries of all types. After all, the point of teamwork, or diversity, is not that we all play the same position or bring the same mindset to a problem—but rather that we find collaborative ways to leverage our distinctive differences.

Responding to fragmentation—a formal approachThe current formal response to the fragmentation problem can be characterized on two levels—increased oversight and merged organizational units. (Oversight is defined here in terms of accountability for budgets, classification rules, committee assignments, etc.)

Oversight“Congress should maintain vigorous, informed, and constructive oversight of the intelligence community.”

Oversight is one way to handle the fragmentation problem—by creating a structure that oversees the whole operation and presumably sees across unit lines. One problem with it 26 B. Herbert, “Ready or Not,” Op-Ed, New York Times, July 25, 2003: A21.27 Ibid.28 Quoted in T. Stewart, 2001, Op Cit.

71

is the difficulty of adjudicating problems without deep knowledge of operating unit perspectives—and many related unintended but common consequences: slower decision-making; weaker internal accountability for coordination; increased politics. Intensive oversight often simply does not work. In fact, Congress and the OMB have had oversight authority over all the elements of the intelligence community, and yet this has not avoided problems such as unmet resource requirements. The Director of the CIA is quoted in the Congressional intelligence committees’ report stating that before 9/11 Congress did not allocate the resources required:

“Senator Kyl once asked me ‘How much money are you short?’ ‘I’m short $900 million to $1 billion every year for the next five years’ is what I answered.”29

Also, if increased oversight is wielded in a top-down fashion it is just as likely to reduce cooperation as it is to increase it. The informal organization knows only too well how to evade autocratic commands to “share knowledge” and “be cooperative.” Increasing traditional control-oriented oversight can undermine internal initiative and honest appraisals of what it would take to foster real collaboration.

Merging formal units “The administration should review…what progress has been made in reducing the inappropriate and obsolete barriers among intelligence and law enforcement agencies engaged in counterterrorism, what remains to be done to reduce those barriers, and what legislative actions may be advisable in that regard.”30

This recommendation to reduce “inappropriate and obsolete barriers” suggests that these barriers are palpable entities, like an overgrown hedge or vestigial barbed wire left long after the cattle were gone. In truth, few formal boundaries have any substance to them that is not reinforced by informal social norms and communication patterns. Moreover, changes in formal boundaries—as is true in this case—merely move boundaries, they do not remove boundaries. Certainly the FBI and CIA have been admonished for years to cooperate better, but the differences in culture, craft, and professional guilds between the agencies have undermined the best intentions.31 Organization systems tend to override isolated changes and snap back to the status quo unless a complement of changes is made to establish a new configuration. Moreover, decades of research and experience regarding large-scale change efforts (and this is one) tell us that no formal system change alone is likely to achieve any significant changes in behavior, no matter what controls are put in place; and especially when these are applied to independent-minded professionals. This is not to say that formal levers do not have influence—changes in reporting relationships, role authority, appraisals, rewards, and other design elements can have much influence when they are aligned. But the point is that they need be aligned, and not

29 July 24, 2003 Report by Congressional intelligence committees, op cit., A12.30 Ibid., A13.31 Current cultural differences and collaboration difficulties have a legitimate historical source in the 1949 law that differentiated the two agency missions and separated the work of international espionage (CIA) and national law enforcement (FBI)--with the idea that the U.S. Government should not spy on its own citizens and that espionage and law enforcement played by different rules.

72

only with each other, but also with informal elements such as group norms, shared values, and a sense of common mission.

Responding to fragmentation: a community approach“It was not clear that the structural changes put in place by the White House would do enough to reverse the longstanding problems in communication and cooperation identified by Congressional investigators....The question is whether the administration is prepared to do the kind of bureaucratic head-banging that’s needed to force everyone to the work together. (So far, I’d have to give the changes a pretty low grade.)”32

We agree, of course, that formal structural changes alone are not likely to work, but is “head-banging” the solution? We have reviewed two formal approaches to addressing fragmentation—oversight and restructuring. On both counts, we have argued that informal issues—cultural norms, professional identity, political issues, and others—are likely to get in the way unless they are explicitly addressed. Community-based approaches provide a strong mechanism for addressing these informal issues and others that will be critical to success.

Oversight, for example, can be reframed in the context of a community-based initiative as “sponsorship.” In this case, the purpose of sponsorship would be to foster a constellation of knowledge-based communities that weave together members of the broader intelligence community. A sponsor board—consisting of members from Congress, OMB, agency leaders and others could commission the development of a set of communities of practice to take responsibility for building up strategic capabilities—within and across agency boundaries. Such capabilities might include ones that require contributions from agents across agencies who specialize in various disciplines—and even officials from other governmental or non-governmental organizations where appropriate. Topics might include (and these are merely a layman’s speculations) different types of terrorism—dirty bombs, bio-terrorism, infrastructure attacks (e.g., the Internet or a power grid); or specific applications such as a city plan for terrorist preparation, prevention, and response; or methodologies such as social network analysis.

A community-based learning system that consists of a constellation of topic-based sub-communities can facilitate problem solving and knowledge sharing across boundaries—without reducing or changing established boundaries that serve useful functions (such as defining areas of accountability or a set of operational objectives). Members participate in these sub-communities for the purposes of professional development and building the collective capability of the agency in targeted areas. The relationships and mutual understanding that are developed in this context create a reservoir of “social capital”—shared values, trust, and reciprocity—that members from different operational units can depend on to facilitate connections across formal boundaries. Their shared community experience provides a “trust fund” for cross-boundary coordination while working on operational issues. Professional relationships can help members of the intelligence

32 David Benjamin, a former National Security Council aide on terrorism, as quoted in E. Lichtblau, 2001, op cit.

73

community across discipline and department boundaries manage difficult issues such as confidentiality, resource allocations and standards. These are easier to manage when officers know each other or have colleagues in common—which is more likely when members are participating in knowledge-based communities as well as in their formal assignments. (Figure 14.1 illustrates how formal accountability units and informal knowledge-based communities can complement each other.)

Such a system provides a foundation for addressing fragmentation issues from a different angle than outlined in the Congressional report. A community-based approach to both oversight and linkages among units complements formal approaches. The oversight structure, reframed here as a sponsor board for a community-based learning system, by itself fosters linkages across agencies. Sponsor-board members deliberate and negotiate together with various communities about what types of problems need to be addressed, what capabilities are required, and what types of cross-boundary work is need to make this happen. Meanwhile, this collaboration at the sponsor level demonstrates in concrete ways the type of collaboration needed by staff and field officers; while providing direct guidance and support—time for staff participation, for example, or funds for a project to hire an outside expert, hold a conference, or build up an online knowledge base.

This discussion has focused on inter-agency boundaries. When we consider boundaries across governmental levels, sectors, and disciplines—the need for boundary spanning mechanisms that complement formal structures becomes even more obvious. The SafeCities community (whose purpose was reducing gun violence nationwide, as described in Exhibit 3) was an example of a structure that crossed all these boundaries: members came from federal, state, and local agencies; private, public, and non-profit sectors; and represented an array of disciplines—including city mayors, police chiefs, faith leaders, social workers and many others.

74

Businesses

Hospitals

City & state government

International agencies

Intelligence agencies

Homeland SecurityBioterrorism “Dirty bombs” Infrastructure attacks Preparedness Prevention Response

Figure 15.1. Complementary structures for accountability (horizontal) and knowledge-sharing (vertical)

The skill gap—dimensions, disciplines, and dynamismThere are several dimensions to the counterterrorism skill gap. First, it is important to understand that the nature of “skills” is more than facility with an analytical technique. Second, the range of disciplines associated with counterterrorism is rich and complex. And third, these counterterrorism capabilities are highly dynamic—with continuous changes driven by technology, innovation, and local adaptations, as well as the need to match the learning of terrorist groups; along with changes in the numbers, tenure, and locations of staff who need the skills.

Skill dimensions and typesAs typical with new or heightened performance demands, organizations experience a skills gap. But counterterrorism is not a simple skill set that can be learned through training, however rigorous. In fact, the “skill gap” consists of a variety of dimensions, both tacit and explicit, and a vast array of skill types. The dimensions of the skill gap include, in addition to conventional skills, such elements as: tools, methods, frameworks, technologies, stories, hunches, trouble-shooting techniques, and the artistry associated with expert espionage and law enforcement. All of these contribute to what it takes to identify a terrorist’s location amidst mounds of noisy data or catch a suspect without incident.

The skill set associated with counterterrorism is vast. A sample of skills include “native fluency in the languages of the Middle East and South Asia, combined with policy, military, business, technical, or academic experience”—and the ability to take on “counterterrorist assignments in hostile environments.”33

This mélange of various dimensions and disciplines of counterterrorism are known within the community as “tradecraft.”

“Talk to intelligence professionals about their work, and you will hear them bat around this term: tradecraft. It’s the combination of skills, procedures and, especially, the culture that guides them in their jobs. … Its tradecraft is imbedded in everything from its training manuals to its computers.”34

This tradecraft includes everything from the ability to “schmooze with local Rotarians” in Michigan35 to “going back to deep, hard dirty work, with tough people going down dark alleys with good instincts.”36

In fact, the nature of the tradecraft in this context is somewhat controversial, especially now as agency staff and skills are being combined to address security threats that many have not seen before. There are significant differences between what CIA and FBI officers consider relevant tradecraft. One CIA officer gave an example of the distinction: “The bureau is wonderful in solving crimes after they’re committed, but it’s not good at penetration. We’ve go to do it.” But the old-style CIA espionage methods may not apply 33 July 24, 2003 Report by Congressional intelligence committees, op cit., A12.34 Bruce Berkowitz, a research fellow at the Hoover Institution and an analyst at the Rand Corporation, in “A Fresh Start Against Terror,” Op-Ed, New York Times, August 4, 2003: A17.35 Ibid.36 A senior general in the U.S. military, quoted by Hersh, S.M., 2001, op cit., p. 35.

75

so well to homeland security assignments that include schmoozing with Rotarians and local faith leaders; and FBI surveillance methods may not go over so well if applied too broadly. Bruce Berkowitz, a former CIA officer, suggests that the new demands will require “a fresh start” without undue baggage from either agency—building capabilities from the bottom up. He asserts: “Intelligence organizations tend to fail when they are asked to perform missions outside their tradecraft….Simply put, the C.I.A. is not suited to the mission of homeland security.”37

Whether we agree with Berkowitz or not, his argument points out the dangers of simply combining expertise in the hope that the mix will somehow yield the right configuration for the problem we face now. He points out that the Counter Terrorism Center was established two decades ago—and included officials from a variety of agencies in the intelligence community, including the CIA and FBI—and we still have the problems we have today. Dynamism and DistributionFormal hiring and training strategies have not kept pace with what intelligence tradecraft entails. As one observer put it, “we’ve been hiring kids out of college who are computer geeks….today’s CIA is not up to the job.” Additional challenges for learning tradecraft include keeping up with the dynamism of the field—and the need to distribute skills broadly across organizational and geographic boundaries. The distribution problem is especially severe as large numbers of staff are hired or allocated to counterterrorism—and because the definition of the field itself is still in flux, and many old hands may require new types of skill development. Finally, not only is there an urgent need to bring new people up to speed, but a demographic tipping point (affecting all organizations with baby-boomer majorities) augurs the loss of much staff expertise.

“According to FBI agents, FBI counterterrorism training was extremely limited before Sept. 11.”38

“At the same time, the D.O. has been badly hurt by a series of resignations and retirements among high-level people….”39

Responding to the skill gap: a community approachCommunities of practice are informal structures that consist of practitioners who share a common passion and a desire to work and learn together for their own professional development and to make contributions in the field. They are considered as knowledge-based structures because they build the knowledge of members, colleagues, and stakeholders in their chosen domain. They provide a social context for learning and connecting that may include an array of learning activities that is much more extensive than provided by typical training courses. Moreover, professional development in the context of a community is more continuous, self-directed, available on demand, and performance-oriented. This is not to say that formal training is not important. In fact, mature communities often develop training courses—both on-line and face-to-face. But courses are designed as one component of many that are used intentionally to foster both

37 B. Berkowitz, 2001, op. cit.38 July 24, 2003 Report by Congressional intelligence committees, op cit., A13.39 Hersh, S.M., 2001, op cit., p. 36.

76

professional development and organizational capability, (which includes standards, tools, data-bases, and other elements that can be made accessible across the organization).

The learning—or “practicing”—activities of communities thus include training, coaching, mentoring, conversations, story-telling, conferences, visits, websites, and listservs. These activities provide continuity of learning over time and help to adapt methods to the style and context of the learner and to what they need at a point in time along their professional-development trajectory.The community also provides a strong motivational context for learning. In mature communities you do not learn alone, and the development of your practice is not wholly a private matter. You are expected to build on what others know, and to help them apply what you know. A community provides a social context for building a new understanding of a field and the elements of the capability associated with it. It provides a social vessel in which members can build a sense of professional identity—a passion for the field and a commitment to making a contribution. It can build credibility with stakeholders who may provide resources. The community also gives members opportunities to apply skills, which demonstrates the potential of the practice while solving real problems.Communities respond to the skill-distribution (or dissemination) challenge by building on the social capital—trust and reciprocity—that are characteristic of healthy communities of practice. First, they provide a forum for connecting people who want to learn from colleagues—putting a name to a face and getting an introduction. These informal connections with peers, mentors, and coaches are generally the most powerful source of new skills and understanding. There is another way in which communities help to distribute skills across the organization—by increasing the practitioners’ ability to leverage the skills they have as the community exposes them to a wider set of problems.

There are several other ways in which this foundational set of relationships promotes processes of building, sharing, and applying skills:

Skill itself is “distributed” phenomenon. We do not act alone. Our individual skill is a combination of tools, frameworks, and technologies. And capability includes access to others to whom we can turn—even in the immediate context of practice—for a brief consultation or request for more extensive help. Communities help practitioners fully leverage the “invisible sea of expertise” related to their domain.

A robust skill such as counterterrorism is a conglomerate of sub-skills, many of which can become specialties in a context where a group of professionals agree to share their distinct expertise and collaborate. Without such a context, it is harder for individuals to focus in a particular area, and the community as a whole is less likely to have a distributed access to practitioners with deep expertise

Communities help capture knowledge and make it broadly accessible. They moderate online facilities such as websites that store descriptions of tools, frameworks, case studies, research articles, etc. They may also develop a directory that helps colleagues find others for help or collaboration; and use such a directory to cultivate a

77

group of external experts—including alumni—as part of the overall knowledge network.

From the expert’s perspective, a community helps to leverage skills in ways otherwise not possible—through collective community activities, personal interactions, and ways in which the community works to codify what it can of expertise. The community can help siphon off beginners’ questions so experts can concentrate on higher-order (and more interesting) questions.

The capacity of a community to distribute skills helps to address the problem of dynamism—the rate at which skills and all related components depreciate. The half-life of knowledge in many areas, computer science for example, is less than two years—relentlessly driven by continuous advances in science and technology (and the seeming inexhaustible consumer appetite for new applications). In areas where new knowledge domains are constantly emerging—or especially when a field is in a state of flux—mature communities are particularly important as forums for connecting the best expertise available—in ways that minimize framework wars and promote concerted effort. (Achieving such a culture is often not easy of course, but it is easier to achieve in the context of a community of 30 or so core members than in an organization of 120,000 employees.) Communities can help vet new ideas and the seriousness of emerging problems in ways that only practitioners who have deep understanding of the context can do—thus providing strategic decision-makers with more credible, well-vetted views on difficult questions. (If there had been a stronger legitimacy to the community of engineers at NASA, one wonders if their concerns would have been more forcefully voiced and more likely heard.) Finally, communities can help build, share, and apply capability across government levels and various sectors of civil society—as did the SafeCities community on the topic of reducing gun violence—because membership is based on one’s connection to the domain rather than on a specific organizational affiliation.

The resource gap—staffing, voice, and leveraging what you’ve got“[These] are the kinds of misses that happen when people—even very competent dedicated people such as the CIA officers and FBI agents and analysts involved in all aspects of this story—are simply overwhelmed.”40

The resource gap is clearer now in retrospect as the nature of terrorist threats is better appreciated by Congress, the White House and OMB, as well as those in the intelligence community. Resource allocations are a strategic decision, and of course there were plenty of priorities in the federal budget competing for attention in the years before 9/11. During that time, a number of leaders of the intelligence community had sounded the alarm regarding terrorist threats. In fact, in 1998 the Director of Central Intelligence (DCI) made a declaration of “war” against bin Laden. But this declaration and associated assertions that “no resources or people [be] spared in this effort, either inside [the] CIA, or the community” were not heard by many in the intelligence community, including

40 CIA officer, as quoted in the July 24, 2003 Report by Congressional intelligence committees, op cit., A12.

78

leaders in the FBI. The Congressional Intelligence Committees’ report stated that “The only substantial infusion of personnel to counterterrorism occurred after Sept. 11….”41 The DCI, George Tenet, explained that he had been calling for additional resources, to no avail:

“… [Y]ou get what you pay for in terms of our ability to be as big and robust as people …if you don’t pay at the front end, it ain’t going to be there at the back end.”42

The refusal to the CIA’s request for additional funds to fight terrorism before 9/11 suggests that additional Congressional oversight by itself would not necessarily improve the US counterterrorist capability. Only since 9/11 has there been a rapid, concerted effort to allocate new resources to counterterrorism units. (Soon after, the number of CIA staff in the Counterterrorism Center was nearly doubled from approximately 400 to 800.) But this focus on the number of resources moved or required may obscure a more fundamental point about resource preparedness: Who will listen next time an unfamiliar alarm is sounded—before a new tragedy occurs? Are Congress, the White House, and the OMB ready to listen to voices within the community regarding new types of threats—and therefore new types or numbers of resources that are called for? What has changed organizationally to raise the likelihood that the voice of those who see and understand emerging threats will be heard next time a major shift in resources is required? (This question recalls the Challenger disasters, where the technical problem changed from 1986 to 2001—from O-rings to foam insulation—but organizationally the problem was the same—how ensure that the voice of those who see and understand the problem gets sufficient consideration?)

Community response to the resource gapWhile communities cannot provide resources as directly as they can help span boundaries and build capabilities, they can contribute to the resource allocation problem in three ways:

1) increase credibility and reliability of the intelligence community’s voice regarding how best to allocate resources

2) better leverage the resources that exist3) accelerate and deepen skill development when new resources are brought on

board.

If we see the resource-gap problem as one of “voice,” not only a matter of funds, then it is clear how a community response can help. Communities increase the information available to decision-makers and help integrate information about priorities across boundaries—horizontal and vertical. When communities of practitioners close to problems bring additional credibility and legitimacy and urgency to problems—it is more likely that decision makers will make better decisions sooner. Furthermore, the conversations and negotiations among practitioners and strategic decision makers over time can build stronger connections between practitioners in the field with strategic officials dealing with stakeholders on the Hill and with other strategic partners (such as foreign governments, the embassies, and hundreds of local and state governments and 41 July 24, 2003 Report by Congressional intelligence committees, op cit., A12.42 Ibid, A12-13.

79

cross-sector coalitions). Communities can contribute to the resource-allocation problem, especially to the extent that its causes are largely about coordinating information across boundaries and ensuring that decision-makers are aware of problems and possible solutions.

Communities also contribute to resource allocation, indirectly again, by getting the most from the resources available and building their capability over time. As outlined above in the section on distributing capabilities, communities leverage internal resources—as well as external ones, such as alumni and external experts. Communities help nurture the passion and skills of members, and there is less wheel-spinning and bickering. There is more concerted action and better coordination and less redundancy. One senior CIA officer noted the cost of a frayed community fabric:

“The CTC and two of the other major intelligence centers—dealing with narcotics and nuclear-non-proliferation issues—are so consumed by internecine warfare that the professional analysts find it difficult to do their jobs. ‘They’re all fighting among each other,’ said one senior manager.”43

In addition to leveraging current resources, communities can help reduce outlays by improving the quality of the work and reducing “rework.” This sounds like a trivial term to use in this context, but certainly it is much more efficient to prevent problems than have to clean them up after the fact (not to mention the lives saved and hardships avoided). On a small scale, when communities foster collegial relationships, members solve problems faster—and thus have additional time to take on new ones.

Clearly, of the three elements in the “prepare, prevent, respond” counterterrorism mantra, the prevention component is most crucial. One theme we have heard over and over—in the case of 9/11, the Challenger disasters, and other historical ones such as the Bay of Pigs fiasco—there are almost always people who saw it coming but were afraid to speak up, or did and were not heard. If only we could find ways to pick up the signals of those who see ahead and provide better mechanisms for vetting—vs. ignoring—their voices. For every catastrophe we prevent—however rare—we more than compensate for the investment. And communities cost very little—approximately one FTE for a core group of about 20 (plus the time they spend together in meetings or on projects, maybe 4-8 hours a month).

Finally, it is one thing to hire staff and another to see benefits. In fact, in the short term as resources increase, overall capacity may decrease, because skilled people will be staffed to train and coach rookies. The resource gap cannot be addressed effectively, therefore, without a skill-development strategy. This brings us back to the need to consider community-based approaches for developing skills. Communities help assure that members gain expert-level skills, leverage the potential for peer-to-peer learning; and build a professional identity--all of which help the community as a whole keep up with external demands as well as members’ development needs. Finally, if the intelligence community can better coordinate across its internal boundaries, it can better leverage its

43 Hersh, S.M., 2001, op cit., p. 38.

80

collective resources and increase the strength and legitimacy of its voice with stakeholders.

Conclusion“[Al Qaeda is] hard to target because they have few formal procedures to disrupt and little physical infrastructure to destroy. They are hard to infiltrate because they are held together by close personal ties and intensely shared values.”44

It is ironic, and revealing, that we are at war with a state-less terrorist organization that is held together by passion and designed for boundary-crossing—and apparently doing both better than our own bureaucracies. In fact, it is hard to define who and where this “state-less” enemy is. We cannot find him, literally. We can unseat formal governments as we have in Afghanistan and Iraq, but we cannot pin down Al Qaeda. Like the monstrous many-headed hydra that Hercules faced, it seems to spring up one or more new heads for every one we chop off. This type of enemy poses a qualitatively new challenge, which—no surprise, given the zeitgeist—attacks the Achilles heel of the formal bureaucracy: its immobility, inflexibility, and reluctance to rely on the informal—the passion, trust, and artistry of its people. This challenge cannot be met simply by addressing the symptoms of fragmentation, skills, and resources as understood in conventional terms. We must address the underlying organizational capacity to manage these issues and the unfamiliar ones to come.

The Congressional recommendations focus on formal elements—structural barriers, resources, training, Congressional oversight, and an intelligence czar. And yet the Committee’s record and related reports are full of examples and stories that point to issues that such changes will not address—the cultural boundaries that exist both within and across agencies (and levels); the elements of “tradecraft” that cannot be addressed by training; a resource gap that is as much about voice and skills as anything else.

The mismatch of the evidence available and the conclusions reached may be due to a limited familiarity with relevant analytical frameworks and design options—such as those outlined here (see Figure 14.2). In fact, there is much to learn from recent organizational research and practice, which has validated through a growing number of compelling cases the importance of networks, communities, and related informal phenomena (such as passion, trust, and expertise). These new models address both formal and informal dimensions or organizing; and both operational and knowledge-based phenomena. (The informal and knowledge-based elements are related, because organizing for knowledge must address issues of professional identity, collegial relationships, and the artistry of expertise.) Informal elements, unfortunately, are generally overlooked because they are harder to define and manage in standard ways; yet we cannot deny that issues such as leadership, group dynamics, and intrinsic sources of motivation drive results as much or more than the best-designed structures, systems, and procedures. We now need to add community-related phenomena to this list of influential informal elements. It is essential

44 U.S. Navy Vice Admiral Arthur K. Cebrowski, former head of the Naval War College, quoted in T. Stewart, 2001, op. cit.

81

to address these informal dimensions in any rigorous analysis of what types of changes it will take to accomplish large-scale change in an organization.

ISSUE FORMAL DESIGN RESPONSE COMMUNITY DESIGN RESPONSEFragmentation oversight of budget, intelligence

classifications, committee assignments, and other formal processes

merging departments

sponsor leadership that provides a coordinating role for a constellation of topic-based communities

cross-cutting, knowledge-based networks communities that foster trust across boundaries

Skills formal training reassignments

communities that provide informal context for learning through both formal and informal means

leveraging diversity of skills via better knowledge of who knows what and stronger peer relationships

practitioner stewardship of dynamic knowledge baseResources changes in allocations when

disaster calls attention to need primarily focus on allocation

process

emphasis on informed, credible voice of practitioners and managers to influence allocations—before disaster strikes

find creative ways to leverage resources across boundaries

draw on stronger foundation of skills and motivation

Figure 15.2. Complementary formal and informal design options

The homeland security problem shares characteristics with the gun-violence problem faced by SafeCities. From an organizational perspective, their challenges are quite similar: build capability quickly; leverage the best tools, skills, and experience available at various levels of government; and create mechanisms for innovation, knowledge-sharing, and collaboration among groups within and across agencies, cities, and sectors.

It was a weakness of community, not merely inadequate formal structures, which left the U.S. so vulnerable on September 11. We must pay as much attention to this community dimension as to formal elements in order to prevent future disasters like it, and especially to prevent ones unlike it. Such preparedness will require an organization capability to learn and innovate at a pace, level, and duration that is unprecedented. The emphasis on the term “intelligence community” throughout the Congressional Intelligence Committee’s report and other commentary is adept given the tacit acknowledgement that there is a need for such a community. At this point, unfortunately, the community is dangerously undeveloped along several dimensions: the intelligence services lack a common identity, cross-boundary trust, and collective stewardship for the intelligence enterprise. The good news is that there is a growing body of work to help us analyze the nature of the intelligence community as a community of practice; an analysis which may provide a number of new insights regarding areas for development and how to act on them; and how to integrate such approaches with formal organizational changes, (and how to combine approaches to get the most from each).

The term “community” suggests a stronger shared fate among the intelligence agencies than does the notion of a “network,” which makes it a particularly appropriate term to use

82

in this case.45 The term rightly intimates how important it is for the various intelligence agencies to feel that they “belong” together; and that they see this sense of belonging—an informal sense of mutual commitment—as the true foundation for the kind of knowledge-sharing and coordination needed to prepare, prevent, and respond to terrorist threats on a world-class level. It will require just such a community to take on the proliferation of decentralized, fast-innovating, mission-driven terrorist networks that threaten the security of the United States and nations worldwide.

Congress should incorporate an analysis of informal issues in its Intelligence Committee report, and integrate community-oriented recommendations for action along with those it has proposed thus far. In particular, it should commission an executive-level sponsor board to explore in depth how to combine both formal and informal approaches for building the capability of the intelligence community—including issues such as agency boundaries, skill-development, and resource allocations. This group can consider, for example, the types of sub-communities that might be organized to address particular issues that reach across agency and governmental boundaries.

The sponsor board can lead an ongoing renewal effort aimed at building an intelligence community that is a model for nations worldwide—and in fact, actively builds community connections with peer organizations on a global level. Moreover, the sponsor board provides a context for building trust across agencies among senior leaders, a crucial foundation for genuine collaboration at the operational level. Finally, the sponsor board can provide leadership to strengthen the intelligence community as a learning system—as a true community of practice; as an interwoven constellation of communities that collectively steward world-class expertise to combat terrorism and other threats to national and international security.

* * *

The great achievements of ancient Athens are attributed in part to its citizens’ innovative, impassioned capacity to protect the city-state from persistent external attacks. Historians have argued that the Athenians’ distinctively democratic government was instrumental to the energy and creativity that they displayed in battles on land and sea. Similarly, America’s greatest strength and source of legitimacy—at home and abroad—is our democratic culture and the creativity and civic initiative that are associated with it. Yet, bureaucracies often stifle these cultural strengths. The Athenians used a community-based model to engage all citizens in the governance and protection of the nation. We too need to

45 Formally, we distinguish communities from what are commonly referred to as “networks” (although the terms are often conflated in practice). Briefly, key distinctions include: 1) a community is “about” something, whereas a discursive network may simply exist as a set of relationships, used for multiple purposes but no common intention; 2) community members do more to connect as a whole community to develop collective rapport and trust, whereas many members of a network may never meet, or even know the other exists; and 3) communities of practice, by design, take on the stewardship for a shared practice—whether counterterrorism or a new art form; whereas a network may serve any number of purposes—access to information, job leads, shared interests, etc.—but not necessarily focus on building or sharing elements of a practice repertoire. See also: Wenger, E.C. & Snyder, W.M., “Communities of Practice: The Organizational Frontier,” Harvard Business Review, January-February 2000.

83

use community-based approaches to promote collaborative, creative, and skillful practice to keep up with the challenges we face now, and will face in the century ahead.46

EpilogueSince this illustrative case on Homeland Security was originally written (in September, 2003), the intelligence community has made noticeable advances in collaborating across boundaries. These improvements were highlighted in various news analyses in the wake of the capture of Saddam Hussein. As recently as July, 2003, the Congressional review castigated the community for a failure to connect the dots regarding intelligence on the September 11 terrorists. In contrast, only six months later, the chairman of the Senate Intelligence Committee stated: “Saddam’s capture is a direct result of unprecedented cooperation and joint effort on the part of our intelligence analysts, operators in the field and our military.”47

News reports emphasized that it was “human intelligence,” not data feeds from spy satellites and the like, which was responsible for important breakthroughs: “In recent weeks, the information gathered by the C.I.A., the Defense Intelligence Agency and the intelligence arms of the military services has been closely shared among the agencies through a new corporative arrangement in Baghdad.”48

The intelligence community’s ability to “connect the dots” was literally demonstrated in the application of network-analysis software that “depict[ed] the links between members of Mr. Hussein’s enormous extended family,...[and] helped them visualize non-obvious family relationships, and eventually pinpoint the families in Tikrit who were hiding Mr. Hussein.49 The software allowed them to integrate information from multiple intelligence sources to create a whole picture of the kinship network. Thus, “it takes a network to fight a network”—whether a global one such as Al Qaeda or a more local one like Hussein’s extended family.

Clearly such collaboration is essential, and the intelligence community has proved it can make great progress, under urgent conditions, in a short time. The challenge ahead is to cultivate subcommunities more systematically within the broader intelligence community to address particular areas, such as types of terrorist groups, methods, or regions. These specialized communities can combine perspectives and resources across formal boundaries to create a safer world--as demonstrated by recent successes.

46 Brook Manville and Josh Ober (Company of Citizens, Boston: Harvard Business School Press, 2003) describe in detail the relationship of Athenian democratic organizational structures and the city-state’s capacity for innovation in an array of domains—including the economy, philosophy, and the arts—as well as warfare.47 Jehl., D., “Intelligence: Spy Agencies Vindicated after String of Setbacks,” New York Times, December 15, 2003.48 Ibid.49 Berkowitz, B., “Learning to Break the Rules,” New York Times, p. A35, December 19, 2003

84