Creating an Innovative Organization

AN INNOVATIVE organization engages everyone throughout the organization in the task of developing and implementing new ways to reach the organization's goals. And everyone indeed includes everyone from the chief executive to frontline workers.^l

Getting the chief executive to be innovative ought not to be too difficult. After all, the chief executive was not repeatedly promoted to more and more sophisticated responsibilities without a few creative ideas along the way. We expect that the chief executive of a business division or a government agency will be innovative (though, all too often, we are disappointed).²

Regardless of how difficult it is to get the chief executive to be innovative it will certainly be more difficult to get middle managers to be innovative, still more difficult to get frontline supervisors to be innovative, and perhaps even more difficult to convince frontline workers that part of their job includes being innovative. This raises important questions: Is it possible to create an innovative organization? Is it possible to persuade every individual in the organization that an important part of his or her responsibility is to develop and implement new ways of achieving the organization's purposes? For most people, I suspect, the answer to these questions is yes. Most of us have observed one or more innovative organizations.

Often these innovative organizations are found in small, suburban communities with homogeneous populations such as Visalia, California (Osborne and Gaebler 1992), or Dakota County, Minnesota (Light 1994). The basic characteristics of these communities seem to foster the trust that appears necessary for innovation to flourish (Behn 1991a). But innovative organizations have also been created in more demanding environments:

If innovative organizations can be found in such non-nurturing environments, they can, perhaps, be created in a wide variety of situations.

If innovative organizations exist, and if we assume that there is some benefit to such organizations, other important questions are raised. How can the leaders of a public agency somehow make it innovative?³ How can these leaders get everyone in the agency to pursue innovative ways of achieving the organization's mission?⁴ How can they get middle managers, frontline supervisors, and frontline workers all to be innovative?

The last question may be the most important for four reasons. In the first place, most organizations have more frontline workers than they have middle managers or frontline supervisors. Also, frontline workers know the most about the actual production of the organization's services. In addition, frontline workers have daily contact with many of the agency's clients and stakeholders, so they are well positioned to figure out how the agency should respond to this key part of its environment. Finally, if a leadership team has some ideas that do inspire frontline workers to be innovative, they may be able to figure out how to get everybody else in the organization to be innovative as well.⁵ Indeed, the 10 hints for getting frontline workers to be innovative can be applied (with a little adaptation) throughout the rest of the organization.

Helping Frontline Workers Become Innovative
Innovative organizations do not miraculously come into existence. Rather, they are created by leaders who establish the conditions necessary to bring out the innovative ideas within everyone.

How can organizational leaders create these conditions? In particular, how can they create conditions that will encourage frontline workers to be innovative? This requires, I believe, that leaders fulfill two major conditions. They must convince frontline workers that the leadership supports the line; and, they must ensure that frontline workers understand the big picture.

In every effective organization, there is some kind of implicit contract between the leadership and the line. The line will produce what the leadership wants; in turn, the leadership produces what the line wants. The organization's leadership wants to make this message as explicit as possible: "You produce for us, and we'll produce for you" (Behn 1991b, 63-64).

This implicit contract is needed by any organization that seeks to become innovative. Frontline workers will not help an organization's leadership do a better job at achieving its mission unless they believe these leaders will help them. This is a simple quid pro quo. If leaders want help from the front line, they had better help the front line. Moreover, leadership has to make the first move. The agency's top leadership needs to go out of its way to make sure that the frontline workers realize that management is on the workers' side. The first two hints are designed to achieve

Condition 1: Frontline workers know that leadership is on their side.
But what should those frontline workers who have decided that being innovative is good for the organization (and good for them) attempt to accomplish? In what direction should they attempt to innovate? What are the constraints? How will an innovation fit within other efforts being made throughout the agency? What is the purpose of the agency and how will any specific innovation help to achieve that purpose? To be effective as innovators, frontline workers must understand what the organization is trying to accomplish, why it is trying to accomplish that, and how it might achieve that goal. "Broader perspectives," according to Rosabeth Moss Kanter, "help stimulate innovation." Innovation is more likely, she writes, "when jobs are defined broadly rather than narrowly, when people have a range of skills to use and tasks to perform to give them a view of the whole organization, and when assignments focus on results to be achieved rather than rules or procedures to be followed" (1988, 179). The last eight hints are designed to achieve

Condition 2: Frontline workers understand the big picture.
Before frontline workers are going to become innovative, they have to believe that the organization's leadership supports them, and they have to understand the big picture.

Hint 1: Be immediately responsive to requests for improved working conditions (or obtain a new photocopier quickly).

When an executive first asks frontline workers or middle managers what should be done to improve the organization's effectiveness, the responses will inevitably focus on working conditions. People will complain about the lack of a soft drink machine, the broken toilet, or the photocopier that barely reproduces the original. At the Bureau of Motor Equipment in the New York City Department of Sanitation, the mechanics' first concerns focused on heating in the winter, cooling in the summer, and cages in which to secure their tools.⁶ All of these are simple complaints; they focus primarily on the working conditions of the workers, though all do relate, directly or indirectly, to the achievement of the organization's mission. Obviously, workers will be more productive if they have the right tool (be it a wrench or a photocopier). The closer workers are to facilities, such as vending machines or rest rooms, the less time lost from the job. Although such requests appear to concern only the convenience of the workers, they also improve organizational effectiveness.

Moreover, such requests are a test. The workers have asked for these improvements many times before, yet management has never obtained a new photocopier. In fact, as far as the workers can tell, no one in management has even tried to order a new photocopier. In the workers' logic, should not the top managers of the organization be able to pull off the simple challenge of getting a new photocopier? (If top management has remodeled its own offices, the workers will know that new equipment can be acquired when management really wants to do so.) Thus, the request for some simple and obvious improvement in working conditions is a test of how serious management is about improving the organization:

The quicker that top management produces the new copier, the better its credibility will be.

In fact, before asking frontline workers what should be done to improve the organization, its leaders ought to know the answer they will hear. Before top management meets with the workers, leaders ought to find out what kind of improvements the workers will request. Before the meeting, they ought to check out exactly what they will have to do to produce the improvement and how long it will take. Then, when confronted with the request, they can commit to making the improvement and also state clearly whether the improvement will be completed in a day, a week, a month, or a year. To promise a new photocopier in a week and then produce it in a month is certainly better than never producing it at all; but the delay does bring into question how seriously management takes its own deadlines.

To identify the needs of frontline workers, the agency's leadership ought to ask the union. In fact, in a unionized agency, if the organization's leaders go straight to their frontline workers, the union will view this as a direct threat, an effort to undermine its role. At the Bureau of Motor Equipment, the agency's management asked the leaders of BME's different trade unions to help identify members to represent frontline workers on a labor-management committee. Hint 1 1/2 might be: Don't ignore (or try to go around) the union.

Hint 2: Support mistakes (or sit next to the first honest innovator called before a legislative committee).

Innovative organizations make mistakes, lots of mistakes (Behn 1991a). And how the organization treats these mistakes and those who make them sends important signals throughout the organization. If the mistaken innovators are punished in any way, even if they are just perceived to be punished, frontline workers will relearn a basic lesson of bureaucratic life: It does not pay to experiment with new ideas.

Unfortunately, a lot of people make their living catching mistaken innovations. These mistake catchers (journalists, legislators, and now, inspectors general) get their jollies and their professional recognition from uncovering and exposing mistakes. The moral fervor with which they take on this assignment combined with the well-known and easily implemented strategy for publicizing any mistake creates the Ten Commandments of Government: "Thou shalt not make a mistake. Thou shalt not make a mistake. Thou shalt not make a mistake." After all, the mistake catchers do not want to catch the mistake: They really want to catch the mistake-maker.

If frontline workers learn that no mistake, even an honest mistake, goes unpunished, they will certainly be reluctant to be innovative. Consequently, leaders who wish to create an innovative organization have to figure out ways to prevent those who make mistakes from being punished.

Private sector organizations obviously have an advantage here. They can make most of their mistakes without being publicly exposed. U.S. attorneys and journalists will be concerned about illegalities, and financial analysts (and journalists) will be concerned about very expensive mistakes. But in business, small expenditures that merely prove unproductive or inefficient are not cause for a moral crusade. Mistake catchers have a hard time making a living off of the private sector.

In contrast, the smallest public sector mistake can easily become front-page news, with investigative reporters, legislators, and inspectors general all competing to get the credit for exposing this latest waste of the taxpayers' dollar. Can you imagine a public agency copying the approach to failure taken by the Ore-Ida subsidiary of H. J. Heinz? Every time it identifies a "perfect failure," it shoots off a cannon in celebration. Peters and Waterman observe:

Shooting off the cannon serves another purpose: By formally calling an end to a mistake in a positive way, it ensures that people do not continue to pour more resources into a mistaken idea in a futile attempt to prove that they were really right all along.

What is the public sector equivalent of Ore-Ida's cannon? How can the leaders of a public agency convince their frontline workers that mistakes are an acceptable and even a necessary part of improving agency performance? In government, if a cannon is shot off to celebrate a failure, most frontline workers will think that the cannon is aimed at them.

Unfortunately, the agency's leaders are not the only ones with cannons. Journalists, legislators, and inspectors general all have cannons too--and these do not fire mere ceremonial blanks. Their cannons can easily take out a public employee--be that a frontline worker or an agency head.

Thus in the long run, it would be desirable to convince not just the frontline workers but also the general public that mistakes are acceptable--indeed, a necessary part of improving agency performance. This is not an easy sell. Indeed, it may never be possible. But the manager can at least publicly stand up for any frontline worker accused of making an honest mistake.

What happens when the first frontline worker who has made a creative but ultimately unsuccessful effort to improve performance is called to testify before a legislative committee? This is, of course, the nightmare of every government employee: to be publicly accused of incompetence, stupidity, or theft in a forum designed to ensure both that you are unable to offer a coherent defense and that all your friends will learn of your presumed failures and vices. The usual management strategy is to control the problem, and to limit the damage, by making sure that the fewest people and the smallest part of the organization are affected and that everyone and everything else is isolated from the accused. For the good of the entire organization, according to this strategy, one individual or one unit should accept all the blame.

In the short run, this strategy may indeed protect the rest of the agency both from guilt by association and from any adverse legislative action such as a budget cut. But over the long run, it only reinforces the well-known message: Mistakes are not tolerated, and those who make them do so at their personal peril.

Suppose however, that the agency director shows up at the legislative hearing and sits right next to the accused frontline worker so that both are automatically included in any journalist's picture. The frontline worker is no longer the story; the agency head is. Legislators can punish not merely a lowly worker or small unit; they can punish the agency head and the entire organization. Although the leader will offer a better-reasoned and more articulate defense, it will still appear to be purely defensive. Even the well-established business sector argument--that mistakes are required to make progress--may still produce the headline: "Agency head defends worker's mistake. "

Yet, for internal purposes, that is precisely the headline the agency head should be seeking. The objective is not to convince frontline workers that honest mistakes are essential to improvement. Intellectually, everyone understands that, and distributing a copy of the director's testimony can reinforce this basic truth. The objective is to convince people throughout the organization that people who make honest mistakes will not be merely tolerated but will be vigorously defended and that those in the organization who are willing to experiment with innovative ways to improve performance will be protected, even at personal cost to the agency director.

Clearly, the internal signals have to be consistent with this external message. To help frontline workers understand that leadership is on their side, it is important that people suffer no internal penalty for trying. The agency's leaders have to value initiative. This means they should not penalize people for doing something. They should only penalize people for doing nothing.

This also means that the agency's leaders may have to accept what middle managers or frontline workers do even if it does not meet the leadership's definition of perfection. When the agency's frontline workers take initiative and work hard, the leadership needs to recognize their successes.

All this helps establish the trust that is essential for innovation. To become innovative, frontline workers must trust their leadership. They have to believe that they will not be punished for the inevitable mistakes that flow from any serious effort to develop innovative ways to achieve the organization's purposes.

Hint 3: Create an explicit mission and related performance measures (or give people a real reason to be innovative).

We want people throughout the organization to be innovative, but toward what end? An innovative organization needs a clear mission and a set of performance goals. Otherwise, people within the organization can simply pursue their own ends and rationalize their actions by claiming they were only trying to follow the instructions to be innovative. To engage everyone throughout the organization in the task of creating and implementing new ways to achieve the organization's purposes, we need an explicit statement of these goals. Innovative organizations need never use the word innovative; but they do need explicit purposes.

Often, such purposes are made explicit through an inspiring mission and operational goals. The mission provides the general statement of what the organization is trying to do: it suggests an exciting vision of the future, but this vision is necessarily vague. In contrast, the operational goals are mundane: they state the explicit performance targets to be achieved in the next year, quarter, or month. They provide a measure, though not a comprehensive measure, of how well the organization is doing in realizing its mission.

Mission and goals provide the necessary rationale for innovation. The organization may be an air force base attempting to train pilots by flying 17,000 sorties a year (Behn 1992). It may be a state welfare department attempting to move welfare recipients from dependence to independence by placing 50,000 of them in jobs over five years (Behn 1991b). Or it may be a city's vehicle maintenance shop attempting to put a fleet of sanitation trucks on the road daily while keeping costs below 95 percent of the private sector's costs (Behn 1996). Regardless, the explicit goal provides a basis for measuring performance, and the overarching mission provides a means of checking to be sure that the goal has not been obtained at the sacrifice of the organization's true purpose.

At Homestead Air Force Base, the goal of flying 17,000 sorties gave the maintenance and supply crews a clear rationale for being innovative (Behn 1992). Frontline mechanics understood that, by developing ways to repair airplanes more quickly or by developing ways to maintain the planes so they needed fewer repairs, they were helping to achieve the base's mission. They were not asked to be innovative. Rather, they were asked to make sure that planes were available to achieve the base's overall sortie goal (as well as the sortie subgoals for subunits). They were educated to understand how their work influenced the base's ability to achieve those goals. Innovation was not the purpose. Training pilots and flying sorties were the goals. But as people began to understand how their individual tasks influenced the organization's ability to achieve its goals, they began to develop new ways to accomplish those tasks. That is, they began to become innovative.

Similarly, at the Massachusetts Department of Public Welfare, the goal of placing 50,000 welfare recipients in jobs over five years gave the agency's Employment and Training (ET) workers a clear rationale for being innovative (Behn 1991b). To find jobs for welfare recipients, they had to be inventive. Moreover, when a case-management system was implemented, the department made clear to the local units that they--not headquarters--would be figuring out how to make the new system work (Behn 1991b, 119, 112).

The mission of the Bureau of Motor Equipment (BME) within the Department of Sanitation in New York City is clear: to ensure that the city's sanitation crews have enough trucks each day to collect the city's refuse (Behn 1996). Yet on any day in 1978, BME was providing only three-quarters of the trucks needed. Thus, the bureau's leadership created the obvious goal: to provide the number of trucks needed every day. Then, it created a second goal: to keep maintenance costs below those the private sector would charge. The mission and goals provided the rationale not only for redesigning the maintenance process but also for redesigning the vehicles so that they would need less maintenance.

By itself, innovation has no merit. The innovation takes on value only to the extent that it helps to achieve important public purposes. Similarly, innovative organizations are not inherently superior. They, too, become valuable only to the extent that they focus their innovations on achieving their purposes. Creating an innovative organization requires a clear understanding of mission and goals so that individual innovations can be examined to see whether and how much they actually contribute to achieving the organization's purposes. Innovative organizations are not trying to be innovative. Rather, they are trying to achieve purposes.⁷

Hint 4: Broaden job categories (or don't let each individual do only one narrow task).

The traditional, hierarchical organization is designed to minimize the number of different functions that a frontline worker must perform. Moreover, the assumption is that, if these functions are properly defined, a frontline worker need know nothing more than precisely how to perform his or her narrow function. As Frederick Winslow Taylor wrote, "one type of man is needed to plan ahead and an entirely different type to execute the work" (1967, 38). In such an organization, management has the sole responsibility for thinking:

In an innovative organization, however, the responsibility for thinking about how best to accomplish the organization's mission is spread throughout the organization. It is difficult, however, to think innovatively if you see only a small part of the picture. If you are merely told to turn the bolt to the left three times or to fill out the green form and the blue card completely but are not told why these actions must be performed, you will hardly figure out that a plastic fastener could replace the bolt or that a red form could replace both the green form and the blue card.

If frontline workers are going to understand the big picture, they need bigger jobs. Narrow jobs inhibit innovative thinking. Taylor and other advocates of scientific management designed narrow jobs in part to prevent frontline workers from doing any thinking. Broadening the operational responsibilities of frontline workers is necessary if they are to achieve their innovative responsibilities. From her work with private sector organizations, Kanter concluded:

Hint 5: Move people around (or don't let workers think they need learn only one job for life).

Moving frontline workers to different jobs is another way of helping them understand the big picture. Even a worker with a broadly defined job sees only a small part of the organization--and thus understands only a small part of what the organization is trying to accomplish. Thus, moving people into jobs with different responsibilities helps to broaden their understanding of the overall purposes and functioning of the organization.

By training, some people have well-defined (if not narrow) jobs. It is quite possible to move someone from the position of transmission mechanic to that of a sanitation truck driver. That, however, might waste the talents of the transmission mechanic. Conversely, it would be difficult to move the truck driver into the position of transmission mechanic without years of training and apprenticeship. Indeed, even moving the mechanic into the cab of the truck would require some investment in training. Frontline workers with specialized skills should not be moved all around the organization. This does not mean, however, that the ability of transmission mechanics to be innovative would not be helped by occasional rides in the cab of the truck.

There is obviously a trade-off between the benefits gained by moving people into new jobs and the costs of training people for these jobs. People who are highly trained to do specialized jobs cannot be moved without great cost, but moving people into new positions not only revitalizes them by giving them something new to do. It also helps give workers the big picture that they need to contribute innovative ideas to the organization. In the private sector, individuals moving up the corporate hierarchy are given a wide variety of jobs in different parts of the firm; this ensures that, if they are promoted to a position with overall responsibilities, they will understand the big picture. Would not frontline workers charged with being innovative be more effective if they too understood that big picture?

Hint 6: Reward teams, not individuals (or find ways to beat the formal performance-appraisal and promotion systems).

Successful innovations are rarely the work of a solitary individual. To convert an innovative idea into a functioning innovation requires the work of many people contributing to its implementation and adapting the initial idea to fit the operational realities and organizational environment. Certainly this is how many public sector innovations actually come into being (Golden 1990). Moreover, as Katzenbach and Smith wrote, "The team is a basic unit of performance for most organizations" (1993, 27). Teams of mechanics, not individual mechanics, repair and maintain airplanes and sanitation trucks. Teams of social workers, not individual social workers, find jobs for welfare recipients. Teams of people, not individual employees, actually produce the organization's results.

Consequently, it makes little sense to create a system of rewards that focuses entirely on individuals when teams, committees, or groups actually do the thinking and the work. Innovative organizations are not a collection of innovative individuals but of innovative teams.

Unfortunately, our public sector system of rewards has been designed not to enhance performance but to prevent corruption. Therefore, it focuses strictly on the individual. Public sector personnel systems ignore the team. Fortunately, these systems also employ the kinds of rewards that often matter least to people. Indeed, public sector personnel systems usually offer only two kinds of rewards: pay increases and formal promotions. For those people who have worked their way up Maslow's hierarchy of needs (1943) beyond food, safety, and love, to esteem, the formal rewards offered in the public sector are less important than the informal recognition of peers.

If the only forms of rewards are pay and promotion, however, these also become by default the only basis of esteem. This need not be the case.. Effective leaders create a wide variety of rewards, visible recognitions for accomplishment that contribute to the individual's esteem. Such forms of recognition are not artificially constrained by the personnel system. They come in the form of plaques, parties, and praise. Effective leaders use these forms of recognition to provide people with the public credit that both reinforces self-esteem and inspires esteem from others.

Moreover, these informal and more significant forms of motivation are not limited to individuals. Leaders can use plaques, parties, and praise to recognize teams, too. Indeed, leaders can reinforce an individual's ties to an innovative team simply by recognizing the performance of that team, by tying individual self-esteem and the esteem of others directly to the individual's membership on a high-performing team. Plaques, parties, and praise that help celebrate the team's innovative accomplishments reinforce the essential message that it is the work of the team, not that of the individual, that really counts.⁸

At the Bureau of Motor Equipment, the union contracts contained a variety of individual work standards, such as the amount of time it should take a mechanic to fix a headlight. As the labor-management committee moved beyond working conditions to issues of productivity, the unions brought up the detested work standards. After much discussion, labor and management agreed to ignore the work standards and simply not collect the data.⁹ Instead, BME decided to focus on the total productivity of each shop, comparing, for example, the average cost for the radiator shop to repair a radiator with the equivalent private sector cost. No longer was the focus on individual performance; now it was the performance of an entire shop, a team of mechanics, that mattered.

Innovative organizations need more than innovative individuals. They need innovative teams. Unfortunately, much about modern public sector organizations undermines an individual's willingness to be part of a real team and to consider whatever responsibilities they have for innovation to be part of a group effort. Consequently, as agency leaders seek to provide rewards for successful innovations, they should focus the recognition on teams.

Hint 7: Make the hierarchy as unimportant as possible (or at least walk around without an entourage).

Effective teams have to be quite flat. Otherwise they are not real teams; they are simply small units carrying out the orders from a different boss. And to create an innovative organization (whether a large agency or a small team), everyone in the organization needs to feel responsible for helping to produce the organization's real results, not for carrying out orders.

The formal hierarchy of most public sector organizations undercuts any feeling of individual responsibility for the performance of the whole. ("That's the boss's job. That's why he's the boss. I just follow the procedures manual.") Moreover, the formal hierarchy is intimidating. ("If I suggest this idea to the boss, I'll be lucky if she laughs at me. She's more likely not to even recognize that my idea exists--or that I do.")

At Homestead Air Force Base, the base commander, Col. William A. Gorton, made it a point to mix with his frontline workers: "I'd sit around and drink beer with the enlisted men and get to know them more. After a couple of beers, you get to hear a lot of things." Indeed, once Gorton established this rapport, "I had all kinds of people coming out of the woodwork...everybody started confiding in me" (Behn 1992).

Innovative organizations depend, by definition, upon the ideas of everyone from chief executive to frontline worker. Yet if the frontline workers believe that the differences in hierarchical status reflect not only differences in responsibilities but also differences in how their ideas are judged, they will keep these ideas to themselves. No one wants to be told that an idea is silly or to have an idea ignored. So rather than risk embarrassment, frontline workers will simply keep their mouths shut. If the leaders of an organization silence their frontline workers' mouths, they also turn off these workers' minds.

Organizations have formal and informal hierarchies. The operational issue is how much these hierarchies affect the behavior of the individuals, particularly those on the lower rungs in the organization. Does the hierarchy intimidate people from offering suggestions? Does it prevent people from recommending solutions? If a team is to work together to solve a problem, everyone must feel free to contribute; every member of the team must feel that his or her contribution will be valued.

The members of the team also need a "shared sense of accountability" (Katzenbach and Smith 1993, 32). They will never feel that they are sharing accountability if they perceive major differences in status. To be innovative is to take responsibility for improving performance. Hierarchical organizations create not only differentials in status but also differentials in responsibility. To create an innovative organization requires making these hierarchical differences as unimportant as possible (Lawler 1988).

Hint 8: Break down functional units (or don't let the procurement guys tell everyone "no").

Throughout government and often in business, the job of the oversight or overhead units is to say "no." At least, that is how the budget office, the procurement office, the finance office, and the personnel office often define their roles: "People are always trying to pull a fast one, and our job is to protect the integrity of the agency." Unfortunately, in protecting the agency from charges of fraud, waste, or abuse, they are often preventing that same agency from improving performance. Yet the behavior of such overhead units is not irrational. They have never been charged with improving performance; they have simply been charged with administering a set of formal rules. Someone violating the rules can get in trouble--often big trouble. If the agency's performance improves or declines, it makes little difference to those in personnel or procurement.

Creating an explicit mission and related performance measures (Hint 3) is an important first step, but it is not enough. The people in the overhead units will certainly salute the agency's overall mission and goals, but they will only be truly loyal to their own rules and regulations. After all, these are the concerns they must deal with every day. Moreover, the people with whom they also deal every day are similarly loyal to the same rules and regulations. The organization's overall mission and goals may be posted on the wall or repeated in the monthly newsletter, but they are not relevant to the overhead unit's real work.

Consequently, the people who work in budget, procurement, finance, and personnel must be made an integral part of the teams that are charged with producing the organization's results. They may still be responsible, for example, for procurement. Indeed, they may still be responsible for ensuring that the team follows all the procurement rules. Now, however, they have a dual responsibility, for they are also responsible for using the procurement rules in ways that will help the team improve performance. If they are truly members of the team, they will know that they cannot get away with a simple "no." If they cannot answer with an unequivocal "yes," they will feel compelled to respond either "yes, if..." or at least "no, but..."

Homestead Air Force Base was divided into four functional units (Behn 1992). The air squadrons actually flew the planes that trained the pilots and achieved the base's sortie goals. The three other functional units--maintenance, supply, and support--neither flew planes nor trained pilots though their work was obviously critical to achieving the base's mission and goals. Those in maintenance were a little removed from the daily concerns of flying sorties, those in supply were more removed, and those in support still further removed.

The leadership at Homestead broke down the barriers between these four units by visibly identifying each person in every maintenance unit, supply unit, and support unit with a specific squadron. Because workers' rewards were tied to the success of their team, they understood that their most explicit responsibility was to help their squadron achieve its sortie goal.

Functional units do not carry out the mission of the organization; they carry out functions. It takes the work of many different functional specialists to achieve the organization's purpose. If left in their own functional unit, they will never see the big picture; they will only see their own narrow specialty. They will certainly never be in a position to work with other functional specialists to create innovative ways to help the organization achieve its mission.

Dedicated innovators will get around the boundaries between functional units. By creating a skunk works or practicing Jesuit management (It is better to ask forgiveness than permission), they will figure out how to beat the system.

But the truly innovative organization is not the product of a lone skunk works or a few cross-functional mavericks. The innovative organization engages everyone, regardless of their primary functional responsibility, in thinking about the work of the entire organization. To get everyone thinking and behaving innovatively requires that they see their job from the perspective of the entire organization. This is why the leaders of innovative organizations have to break up the functional units into ones that focus on the real product of the organization, whether that be sorties or jobs for welfare recipients.

Hint 9: Give everyone all the information needed to do the job (or don't let the overhead units hoard the critical data).

Information is power. Indeed, one of the best ways that the overhead and oversight units of an organization obtain, keep, and use power is to control the organization's information. These units collect, organize, analyze, and dispense information whenever they find it convenient or useful.

But innovative organizations require information. People need information to understand the organization's performance and to judge how innovative changes will affect that performance. Such information may be easily available because it is well publicized, common knowledge, or attainable from multiple sources. People also need the information necessary to manage the implementation of their innovations, to figure out how best to arrange the technical aspects to meet the formal requirements of some overhead system. This information may not be as available, because it is the type of information that overhead units hoard for the power it gives them in allowing them to justify a "no."

Innovative organizations are designing not only the conceptual framework for doing something differently. They are also designing, implementing, and adjusting the details. These processes of design, implementation, and readjustment require access to detailed and immediate information.

Hint 10: Tell everyone what innovations are working (or have frontline workers report their successes to their colleagues).

How will frontline workers discover that innovation is going on? How will they learn about the innovations that might help them do their jobs better? How will they know that innovation is possible? How will they come to understand that innovation is truly expected of them? One solution is to have their peers, the real innovators, tell them.

In attempting to encourage teams in local welfare offices to experiment with new ways to find jobs for welfare recipients, the Massachusetts Department of Welfare followed precisely this strategy. The agency held all sorts of meetings--from monthly meetings of the directors of the 50 local welfare offices, to large annual conferences attended by all their frontline workers. A standard on the agenda of these meetings was the case-management panel--a team presentation from a local welfare office with each member explaining different tasks that had to be accomplished to get a specific welfare recipient a specific job. The former welfare recipient was also there to describe the process from her perspective (Behn 1991b, 106-7).

Such presentations serve several purposes. Obviously, they can provide for technology transfer, giving other middle managers and frontline workers new information about how to do their jobs better. Such panels can do even more; they can help create an innovative organization. When frontline workers explain how they took an innovative approach to accomplishing the agency's purpose, they dramatize better than any memo or speech from the agency's director that innovation really is possible. Moreover, as different panels of frontline workers make their presentations at different meetings, their colleagues also begin to sense that innovation is more than merely possible. They begin to comprehend that it ought to be the norm.

Frontline workers may have been told numerous times that they were supposed to be innovative, but who demonstrated that the organization had produced innovations that worked? Who demonstrated that frontline workers were actually the ones producing these innovations? Having the inventive and resourceful frontline workers explain their innovations is the most effective way to deliver the message that the organization can produce innovations that work.

Necessary and Sufficient Hints
Are any of these 10 hints (listed in Table 1) sufficient to create an innovative organization? Probably not. No one action--not even a few select actions--will get people throughout the organization experimenting with new ways to achieve its purposes. No one action will convince frontline workers that the agency's leadership is on their side. No one action will give them the big picture. Indeed, even this list of hints may not be sufficient.

Which of these hints are essential? With the exception of Hint 3 about mission and goals, I suspect that none are. It may well be possible to create an innovative organization, for example, without being responsive to requests for improved working conditions and without supporting mistakes. I suspect that scholars and practitioners may offer such examples. I doubt, however, that it is possible to create an innovative organization without providing the mission and performance goals that make it clear what the innovations should accomplish.

Table 1: Ten Hints for Involving Frontline Workers in Creating Innovative Organizations

Hint 1:	Be immediately responsive to requests for improved working conditions (or when they ask for a new photocopier, produce it).
Hint 2:	Support mistakes (or sit next to the first honest innovator who is called before a legislative committee).

Hint 3:	Create an explicit mission and related performance measures (or give people a real reason to be innovative).
Hint 4:	Broaden job categories (or don't let each individual do only one narrow task).
Hint 5:	Move people around (or don't let workers think they need learn only one job for life).
Hint 6:	Reward teams, not individuals (or find ways to beat the formal performance-appraisal and promotion systems).
Hint 7:	Make the hierarchy as unimportant as possible (or at least walk around without an entourage).
Hint 8:	Break down functional units (or don't let the procurement guys tell everyone "no").
Hint 9:	Give everyone all the information they need to do the job (or don't let the overhead units hoard the critical data).
Hint 10:	Tell everyone what innovations are working (or have frontline workers report their successes to their colleagues).

Creating an innovative public agency is, itself, a task of innovation. Each innovative organization will be different. It will be pursuing different purposes. Or it will be pursuing them in a different organizational context, within a different political environment, or within different legal constraints. There is no recipe for replicating an innovation. Similarly, there is no recipe for replicating the innovative organizations mentioned here.

Moreover, there may be many different ways to convert a moribund organization into an innovative one. There may well be another set of hints (that includes the hint about creating mission and goals) that may, in some contexts, prove equally effective. Leadership is not like physics. In physics, the acceleration of an object is always equal to the force on it divided by its mass. You cannot get different answers in physics; you always get precisely the same one.

In contrast, there are many possible answers in biology. For example, there is not just one kind of bird, but there are birds of many different sizes and colors with quite different styles of flying. Yet they are all quite successful; each has found an ecological niche in which to thrive. In fact, to be a successful bird, you do not even have to fly.

Leadership is more like biology than physics. Different leaders can accomplish similar purposes with different strategies and styles. These hints reflect one set of styles implemented by different leaders in different settings. They are not necessarily the only hints that will create an innovative public agency, but they certainly have done so in the past.

Notes
1. Galbraith argues, "organizations that want to innovate or revitalize themselves need two organizations, an operating organization and an innovating organization" (1982, 6). I, however, argue the opposite: in a truly innovative organization, the operating divisions feel a real responsibility to be innovative. The operating divisions do not merely produce the firm's products or the agency's services; they share the responsibility for designing new and better products and services and for developing new and better ways to produce those products and services. At the same time, not everyone in the organization is engaged daily in completely redesigning his or her job. Most of the people spend most of their time carrying out their job as defined by the last innovation. My point is that the task of creating better ways to achieve the organization's purposes is not that of only a few people; this responsibility is in the job description of everyone.

2. Certainly the chief executive could have been repeatedly promoted merely by avoiding any action that was incompetent or evil.

3. This assumes that innovative organizations can actually be created and are not solely the result of some combination of external forces and pure luck. The Western mind, however, rejects the suggestion that external forces and/or pure luck are the sole factors that create innovative organizations. Our Western assumption is that some person or group within the organization took some actions that, perhaps combined with outside forces and a bit of luck, helped to foster the innovative organization. Indeed, we would hope that the actions of these leaders to create an innovative organization were designed to exploit the external forces and the luck.

4. An innovation that does not help the organization achieve its mission is no innovation at all. We do not want the frontline workers at the offices of the state division of motor vehicles to be innovative about the fees they charge for drivers' licenses. We do not want the frontline workers in the city's welfare offices to be innovative about who should receive assistance checks. Innovation only makes sense within the overall framework of the agency's mission and goals, though frontline workers might well challenge that mission, and the agency's leadership team certainly ought to give them a reasonable hearing.

5. Several of my colleagues have argued that, in creating an innovative organization, the biggest challenge is not the frontline workers but the middle managers who may feel threatened.

6. At BME, mechanics are responsible for providing their own tools. Without cages in which to lock their tools, however, the mechanics were continually having to buy new ones.

7. A focus on achieving purposes also serves to help protect people from attacks on their mistakes. This will never provide complete protection because some people will disagree with the mission and performance measures. A focus on purposes does, however, provide a clear defense: We were trying to achieve this goal, and I thought this might be an effective way to do that. Obviously I was wrong, but I was trying to accomplish my organization's mission.

8. The individual rewards of pay and promotion will still exist and can undercut the emphasis on teams. In the public sector, however, salary increases and bonuses often are extremely small. In an era of cutbacks and flattened hierarchies, the opportunities for formal promotion to a higher job classification (as opposed to an informal promotion to larger responsibilities) are also limited. Consequently, the leadership of a public agency can create an informal system of reward and recognition. This system is completely separate from the formal processes imposed by the personnel office and will override for most people in the agency any motivational disincentives created by the formal structure.

9. The work standards could not, however, be deleted from the union contracts. The citywide overhead agency that negotiated with all the city's unions insisted on it.

References
Behn, Robert D.1991a. Innovation and public values: Mistakes, flexibility, purpose, equity, cost control, and trust. Presented at the conference on The Fundamental Questions of Innovation, Duke University, Durham, N.C. (May 3-4).
_____1991b. Leadership counts: Lessons for public managers. Cambridge, Mass.: Harvard University Press.
_____1992. Homestead Air Force Base (a public-management teaching case with sequel and teaching note). Durham, N.C.: The Governors Center at Duke University, Terry Sanford Institute of Public Policy.
_____1993. The myth of managerial luck: Success from an unlikely place. Governing (November): 68.
_____1996. The Bureau of Motor Equipment (a public-management teaching case with sequel and teaching note). Durham, N.C.: The Governors Center at Duke University, Terry Sanford Institute of Public Policy, forthcoming.

Galbraith, Jay R. 1982. Designing the innovating organization. Organizational Dynamics 11, no. 1 (Winter): 5-25.

Golden, Olivia. 1990. Innovation in public sector human services programs: The implications of innovation by "groping along." Journal of Policy Analysis and Management 9, no. 2 (Spring): 219-48.

Kanter, Rosabeth Moss. 1988. When a thousand flowers bloom: Structural, collective, and social conditions for innovation in organizations. Organizational Behavior 10:169-211.

Katzenbach, Jon R., and Douglas K. Smith. 1993. The wisdom of teams: Creating the high-performance organization. Boston: Harvard Business School Press.

Lawler, Edward E., III. 1988. Substitutes for hierarchy. Organizational Dynamics 71: 5-15.

Light, Paul C. 1994. Creating government that encourages innovation. In New paradigms for government: Issues for the changing public service, 63-89. Patricia W. Ingraham and Barbara S. Romzek & Associates, eds. San Francisco: Jossey-Bass.

Maslow, Abraham H. 1943. A theory of human motivation. Psychological Review 50 (July): 370-96.

Osborne, David, and Ted Gaebler. 1992. Reinventing government. Reading, Mass.: Addison-Wesley.

Peters, Thomas J., and Robert H. Waterman, Jr. 1982. In search of excellence: Lessons from America's best-run companies. New York: Harper & Row.

Taylor, Frederick Winslow. 1967. The principles of scientific management. 1911. Reprint, New York: W. W. Norton.