Good government doesn't have to be a contradictions in terms. But incoming mayor Sharon Pratt Dixon faces a huge challenge as well as a golden opportunity. So we asked David Osborne, a student of successful innovation in government, to highlight some ways she can make the city work again.

To: Sharon Pratt Dixon

From: David Osborne

Re: How to really make the city work

Congratulations. You promised to clean out the D.C. bureaucracy, fire 2,000 managers and hold the line on taxes. The good news is that you have a mandate. The bad news is that people expect you to deliver.

You're not the first candidate to make such promises, of course. Many a politician has swept into office on the same basic platform, then departed 4 or 8 or 12 years later with little changed. The problem is hardly limited to mayors. Ronald Reagan ran a campaign much like yours in 1980. Michael Dukakis was elected governor of Massachusetts in 1974 on a pledge to clean up state government and hold the line on taxes. Sixteen years later, as the battered Dukakis made his exit, William Weld was elected on the same pledge.

Politicians come and go, in other words, but the bureaucracy remains.

We all know that the D.C. government is bloated. We know that it could shed 2,000 managers without skipping a beat. But firing 2,000 middle managers is not going to solve the District's problems. In fact, it could make them worse. Under the city's archaic civil service system, those managers can just bump lower-level employees who have less seniority; those employees will bump still others; and you could wind up eliminating the fresh faces but keeping the deadwood -- in jobs for which they have little training and less desire.

So as you sweep clean the bureaucracy, you need to keep one thing uppermost in your mind: The people are not the problem. The system is the problem.

The D.C. government, like most governments, operates with a set of incentives that reward sloth, waste and inefficiency. You have a budget system that encourages managers to waste money. You have a personnel system that encourages managers to live with mediocrity. You have a salary system that encourages managers to build empires. You have an accounting system that encourages people to scrutinize how much is spent but ignores the results of that spending. These systems have created a culture that rewards the very behavior you want to eliminate.

You've probably read Peter Drucker, the management sage. He puts it well: "Critics of bureaucracy blame the resistance of public-service institutions to entrepreneurship and innovation on 'timid bureaucrats,' on time-servers who 'have never met a payroll,' or on 'power-hungry politicians.' It is a very old litany -- in fact, it was already hoary when Machiavelli chanted it almost five hundred years ago. . . . Alas, things are not that simple, and 'better people' -- that perennial panacea of reformists -- are a mirage. The most entrepreneurial, innovative people behave like the worst time-serving bureaucrat or power-hungry politician six months after they have taken over the management of a public-service institution, particularly if it is a government institution."

Your basic challenge is this: You have inherited a government designed for the 1940s, and you need a government designed for the 1990s.

The kind of governments that developed during the 1930s, 1940s and 1950s, with their sluggish, centralized bureaucracies, their preoccupation with rules and regulations, and their hierarchical chains of command, no longer work very well. They accomplished great things in their time. They professionalized government, built safety net programs like Social Security and Medicaid, and won a world

war. But somewhere along the line they got away from us. They became bloated, wasteful, ineffective. More important, the environment changed, and they failed to change with it. Hierarchical, centralized bureaucracies simply do not function well in the rapidly changing, information-rich, knowledge-intensive society and economy of the 1990s. They are like luxury ocean liners in an age of supersonic jets: big, cumbersome, expensive and extremely difficult to turn around.

Not surprisingly, new kinds of public sector organizations are beginning to take their place. They are flexible, adaptable, quick to adjust when conditions change. They are lean, decentralized and innovative. They use competition, customer choice and a variety of other non-bureaucratic mechanisms to get things done as flexibly, creatively and effectively as possible. They contract with private firms and nonprofit organizations to provide services. And they change the basic incentive systems that drive public employees' behavior: budget systems, personnel systems, pay systems.

These are entrepreneurial governments, not bureaucratic governments. Your experience in the private sector -- and your campaign rhetoric -- suggests that you have an intuitive understanding of the difference. But to turn around a luxury ocean liner with 48,000 employees, you will need more than intuition. You will need a clear understanding of what makes entrepreneurial governments tick.

I've spent five years studying such governments, trying to understand what drives their employees to behave so differently. I have boiled the answer down into 10 principles -- 10 sometimes-overlapping pieces of what is really a coherent whole, a new way to operate public institutions. Think of them, if you will, as 10 lessons drawn from the experience of others who have gone before you in the task of reinventing government. 1. Use government more to steer than to row.

Entrepreneurial leaders understand that communities are healthy when their families, neighborhoods, schools, voluntary organizations and businesses are healthy -- and that government's most profound role is to steer these institutions to health. Hence entrepreneurial gov- ernments act more as catalysts, brokers and facilitators than traditional governments do.

Nobody embodied this philosophy more than George Latimer, mayor of St. Paul, Minn., from 1976 through 1989. Latimer cut the city's staff by 12 percent, kept budget and property tax growth below the rate of inflation and reduced the city's debt -- but he dramatically increased the activism of St. Paul government. He did it by catalyzing private sector activity: He used nonprofit, voluntary organizations to manage parks and perform energy audits and operate recycling programs; he set up nonprofit corporations to redevelop downtown, invest in low-income housing and organize the nation's first downtown-wide hot water heating system; and he created more partnerships with foundations than any mayor in American history.

E.S. Savas, a veteran of the Lindsay administration in New York City and the Reagan administration in Washington, introduced the metaphor of steering versus rowing. The word "govern," he pointed out, comes from a Greek word that means "to steer." Most of us understand that government's primary role is to steer. What we fail to understand is that when government also rows, it can lose the flexibility it needs to steer effectively, particularly in times of rapid change.

To vary the metaphor: If policy makers can "buy" services only from their own bureaucracy, they become captives of sole-source, monopoly suppliers. This becomes a problem as soon as they decide to change their strategies. When they want to move their welfare departments into the business of training, educating and placing people in jobs, for instance, they are stuck with caseworkers who have the wrong skills. The solution is to keep policy decisions within public hands but contract with or empower separate organizations (public, private, nonprofit or voluntary) to deliver services, depending on who can best do the job. 2. Whenever possible, inject competition into public service.

During your campaign you suggested that you wanted to see more competitive bidding on existing city contracts. Think bigger: Think about injecting competition into every city service. Entrepreneurial governments have discovered that when organizations must compete for funding, they keep their costs down, respond quickly to changing demands and strive mightily to satisfy their customers. This is particularly important in cities where contracting has been a seedbed for corruption. True competition makes it very hard to award contracts to political cronies.

Phoenix, a city of almost a million people, is a compelling model. The city government there uses competitive bidding in garbage collection, street repair, landfill operation, custodial services, security in city buildings and a variety of other areas.

Phoenix first decided to contract out garbage collection in 1978, during a fiscal crisis. It divided the city into five districts and bid out one district at a time, on a long-term contract. Surprisingly, the Public Works Department -- the agency then collecting the garbage -- decided to compete. Three times it submitted bids, and three times it lost a district. But the losses forced its managers and employees to get serious about improving their operations. Management let the drivers redesign their own routes and work schedules. Together they refined their trucks and became a technology leader in the industry. Together they created labor-management committees to work out better ways to do things. And gradually they got their costs down.

In 1984, Public Works finally won a contract. By 1988, it had won back all the districts. "Over a 10-year period you see the costs for all the other city programs going up," says Ron Jensen, who runs the department. "Solid waste costs have gone down by 4.5 percent a year, in real, inflation-adjusted dollars."

Charles Fanniel, president of the union local, would prefer the comfort of guaranteed jobs for his employees. But he acknowledges that working conditions, pay and morale are all better than they were before 1978. "What happens with this bidding system," he says, "is you cut out all the fat."

To make sure the bidders are all competing on a level playing field, the city auditor's office examines each bid, public or private. (The city auditor is the watchdog, a vital role in any contracting process -- especially when corruption is a problem.) Phoenix City Auditor Jim Flanagan says he has discovered there is no truth in the old saw that business is always more efficient than government. The important distinction is not public vs. private, it is monopoly vs. competition: "Where there's competition, you get better results, more cost-consciousness and superior service delivery."

This is simple common sense. We all know that monopolies protect inefficiency and resist change. And yet, to this day we deride competition within government as "waste and duplication." We still assume that each neighborhood should have one school, each city should have one organization picking up its garbage or managing its public housing, each public service should have one provider. It is one of the enduring paradoxes of American politics that we attack monopoly so fervently when it appears in the private sector but embrace it so warmly in government. 3.Tie spending to results.

Unlike businesses, government agencies normally face no market test. The results of their work are secondary; their ultimate test is political -- who gets elected. For an appointed manager, that means the bottom line is pleasing the politicians, not achieving any particular outcomes. And elected politicians normally care about voters' and interest groups' perceptions more than about the performance of public agencies -- many of which they barely understand. So public managers quickly learn that their value and clout flow not from how well their agencies perform, but from how the politicians view them. Because politicians often judge their worth by looking at the size of their budgets and staffs, managers assiduously build their empires. Because politicians make obeisance to organized constituencies, managers become extremely sensitive to interest groups. Because politicians need friends and campaign contributors, managers are sometimes pressured to show favoritism when contracts are let. The way out of this trap is to tie funding and managers' pay to the results of their work. Bureaucratic governments fund according to inputs: how many full-time positions are allotted, how many students enroll, how many people are poor enough to qualify. Entrepreneurial governments try to fund according to outcomes: how much children learn, how many people find jobs and get off welfare, how clean the streets are. They know that when institutions are funded according to inputs, they have little reason to strive for better performance. But when they are funded according to outcomes, they become obsessive about performance. Again, this is particularly important in dealing with corruption: When contractors are not paid until they produce documented, measurable results, corruption becomes far more difficult.

New York City is often written off as beyond hope. But these principles work just as well in a city of 7 million as in a town of 5,000, as the following story shows.

During New York's fiscal crisis in the 1970s, an independent foundation developed a method called "Scorecard" to measure the cleanliness of streets. It then sent out volunteers every month to rate each of 6,000 streets.

The Sanitation Department had always focused on inputs: How many trucks were assigned to each district? How many men were needed on each truck? Now it began to look at the Scorecard information, which rated outcomes: How clean was each street? Using this information, it reassigned its street cleaners and began to reward crews that made the greatest improvements. By 1986, the percentage of streets rated "filthy" had declined from 43 to less than 4 percent. Nearly 75 percent were rated "acceptably clean."

The foundation, called the Fund for the City of New York, went on to develop outcome measures for parks maintenance, job training and placement, foster care, home care services, school maintenance and other programs. For example, the Department of Housing Preservation and Development contracts with community organizations to manage, renovate and sell so-called "In Rem" apartment buildings -- those the city has seized from tax-delinquent owners. The Fund for the City of New York developed standards for rent collection, vacancy rates and expense levels, which the department built into the performance contracts it signs with these community organizations. The In Rem program is one of the city's most successful. 4. Let people choose among many service providers.

The best way to tie funding to performance is to think of citizens as customers of government. Most businesses get their revenues directly from their customers: If they please the customers, sales increase; if someone else pleases the customers more, sales decline. So businesses in competitive environments learn very quickly to pay enormous attention to their customers. But city agencies get most of their funding from taxpayers, via city councils. And most of their "customers" are captive: Short of moving, they have few alternatives to city services. So managers in the public sector learn very quickly to ignore them. Instead, the managers respond to the mayor and city council -- because that's where they get their funding.

The most entrepreneurial governments understand that the best way to force a bureaucracy to get close to its customers is to make it dependent on them for its very livelihood. This takes competition a step further: Rather than government managers choosing service providers in a competitive bidding process, it lets each citizen choose his or her service provider.

How does it work? Consider Community School District 4, in East Harlem. New York City has 32 school districts. Twenty years ago District 4 was at the bottom of the barrel: 32nd out of 32 in test scores.

"It was totally out of control," says Michael Friedman, a teacher who now directs a junior high. "The schools were chaotic, they were overcrowded. There was a lot of violence, gangs roaming the streets. It was sink or swim as a classroom teacher. They closed the door: 'That's your class, do what you can.' And it seemed like nobody could or would try to do anything about it. You had a lot of people who had been there a long time, and they were just punching that clock."

In 1974, out of sheer desperation, district leaders decided they had to get the worst kids out of the schools so other kids could learn. So they created an alternative junior high for troubled students. The task was so daunting, they told the teachers to do whatever it took to get results. Not surprisingly, the result was a very nontraditional school. It worked, so they created another one. Soon teachers began proposing alternative schools for other children: a performing arts school, a traditional school in which children wore uniforms, a marine sciences school, a school that specialized in creative expression, a career preparation school. Each provided a basic core of instruction in math, English and social sciences, but each also had a distinct focus.

Today District 4 boasts 23 junior high schools, all of them schools of choice. (Six grade schools are also schools of choice.) Rather than being assigned to a school and offered the same kind of education as every other child, students are allowed to choose the style of education they prefer. (In other words, they are treated as customers.) They can choose traditional schools or open classrooms; reading institutes for those behind grade level or advanced schools for gifted students; a computer-oriented school; or a school run in conjunction with the Big Apple Circus. Schools are small -- from 50 to 300 students.

"Kids need to be dealt with personally, not imper- sonally," says John Falco, who administers the choice program. "They need to be recognized for what they are, what they could do and what they could accomplish -- and then you need to figure out how to help them accomplish that."

Obviously, Falco is not describing the normal attitude of an inner city teacher or administrator. Why do people in East Harlem think that way? Because they have to. "If you're not operating a program that kids want to come to, you're out of business," says Falco. "That's happened in a couple of cases. You just can't rest on your laurels. You have to continually strive for ways to meet the needs of these kids."

The results speak for themselves. In 1973, 15 percent of District 4 junior high students read at grade level; by 1989, 64 percent did. In the mid-'70s, only a handful of the district's graduates were admitted annually to New York's elite public high schools, like Bronx Science and Brooklyn Technical. In 1987, 139 were -- 10 percent of district graduates, almost double the rate for the rest of New York City. Thirty-six other graduates went to selective private schools, including Andover and the Hill School. 5. Don't just spend money; invest it and measure your return.

When George Latimer became mayor, in 1976, downtown St. Paul was hemorrhaging. It had lost 41 percent of its retail volume over the previous 18 years. The worst section was known as Lowertown, a depressed, 25-square-block warehouse district that made up the eastern third of downtown. Latimer and his deputy mayor, Richard Broeker, dreamed up the idea of a private development bank, capitalized with foundation money, to catalyze investment in the area. In 1978 they asked the McKnight Foundation for $10 million to back the idea, and got it.

The Lowertown Development Corp. brought in developers, offered loans or loan guarantees and put together package deals with banks, insurance companies and anyone else who would listen. Over the previous decade, investors had put only $22 million into Lowertown. In the development corporation's first decade, it triggered $350 million in new investments -- leveraging its own money 30 or 40 to 1. Thirty-nine buildings were ren- ovated or constructed; they contained offices, cinemas, retail shops, restaurants and 1,500 units of new housing. By 1988 Lowertown generated nearly six times the property taxes it had 10 years before. And the development corporation was turning a profit!

Latimer was a master at using the profit motive. He turned many agencies into "revenue centers," fully or partially dependent on user fees for their income, and let them keep a third of all revenues they generated beyond their budgeted targets, as an incentive to make money. (By 1988, his Department of Planning and Economic Development got 83 percent of its funds from fees and other income.) He constantly set up independent corporations like the Lowertown Development Corp. Dick Broeker created a three-day summer festival called "Taste of Minnesota" that turned a profit in its first year, and the Parks and Recreation Department picked up on the idea and sponsored a 12-day "Riverfront Days" festival that became a big moneymaker.

Traditional governments focus entirely on spending decisions. City councils detail how much is to be spent for each line item, managers document how closely they have followed these instructions, and accountants check them for accuracy and honesty. Not one person in this process examines anything related to the organization's return on investment. Nor is anyone re- sponsible for developing new (non-tax) revenues -- otherwise known as profits.

Entrepreneurial managers understand that guaranteed incomes create few incentives to innovate, so they force some agencies to earn much or all of their revenue. When profit is out of the question, as it often is with services like welfare, public education or police protection, entrepreneurial governments still demand a return -- whether measured in job placements, improved student performance, higher tax revenues or greater community health (lower crime rates, lower welfare dependency, lower drug use). Because they define the returns they want and then measure to see if they are achieving them, public entrepreneurs know which initiatives are worth funding and which are not. 6. Use goals -- not rules and budgets -- to drive your organization.

"Never tell people how to do things," Gen. George S. Patton once said. "Tell them what you want them to achieve and they will surprise you with their ingenuity." Most governments do just the opposite. They don't spell out the goals they want programs to achieve. They spell out the exact process by which programs are to be administered and the exact amount managers can spend on every sub-category of every element of every program.

Rule-driven and budget-driven organizations churn out reams of reports on process and spending -- but they rarely deliver quality performance. If you want performance, define an organization's mission, spell out the results you want, and then throw out the rule books and line items.

Start with the set of rules known as "civil service." Civil service systems may have made sense 80 years ago, but they make no sense today. They make it vir- tually impossible to reward people for quality performance, to move people as needs change or to fire incompetents. This is how you get organizations riddled with employees "who are not capable of doing their jobs," as a former director of your Department of Public and Assisted Housing once described his agency.

If you try to "reform" civil service, the bureaucracy will outlast you. Better to throw it out and start over. You will be accused of grasping for power so you can hire every patronage hack in Washington. (You can point out that your predecessor already did that, despite civil service.)

I realize that abolishing civil service may be politically impossible. But your campaign has given you a precious open- ing: You promised to get rid of 2,000 bureaucrats; the people supported you; and the civil service system stands in your way.

Strike while the iron is hot. Don't settle for a one-time waiver; if civil service rules don't make sense this year, they're not going to make any more sense next year. You can build in adequate protections against patronage without the absurd rigidities of civil service.

As for the budget system, my advice is similar. Line-item budgets encourage managers to waste money. If a manager does not spend his entire budget by the end of the fiscal year, three things happen: He loses the money he has saved; he gets less next year; and the budget director yells at him because he asked for too much last year. Who in their right mind would save any money, under those circumstances? This explains the normal end-of-the-year rush to spend money. Most public managers know where they could save 10 to 15 percent, but they have no incentive to do so. Why go through the pain of transferring or laying people off, if you can't keep the money and use it for something more important?

Under the duress of Proposition 13, Fairfield, Calif., invented a solution. A handful of other cities have already copied it, and several nations have moved in the same direction. You might call it a "mission-driven budget."

The mission-driven budget makes two simple changes. First, it does away with line items, leaving one basic budget for each program or agency. This gets legislators out of the business of dictating inputs and frees managers to shift resources to where they can be most productive. (If the city council is smart, it shifts into the business of measuring and funding outcomes -- which gives it far more genuine control.) Second, it lets managers keep part or all of any money they can save, to use on new priorities. This gets them acting like they're spending their own money, rather than someone else's.

Say the police department is spending $500,000 a year on new squad cars but really needs a new computer system. Normally the chief would not risk asking the city council to shift the money, because he might lose the squad car line item but never get the computer funds. Under a mission-driven budget, he could save $200,000 a year on cars for three years running, then use the $600,000 to buy the computer system.

Fairfield's police department dramatically illustrates the contrast between the two budget systems. Its city budget is set up the new way, but it still hustles a fair number of federal grants, which are set up the old way. "It's amazing," says Chuck Huchel, chief of public safety. "The same people behave differently with the two streams of money. With the federal grants, we prepare a budget in advance, and we put on all the bells and whistles, all the frills -- we try to anticipate everything we might need. When we get an authorization, we spend everything that's on the list, whether we need to or not. People don't say, 'Oh, I can save some money here, or I can use it another way now.' Because it's in the plan. You don't have incentives to make the cost savings, because if you don't spend it, you give it back.

"With the city money, they know that any savings they make can be applied to other programs or other equipment. So you say, 'Hey, I don't actually need this to make the program work, so I'm not going to spend it.' Plus they get creative about saving money. We needed a weather covering over a gas pump, to protect people from the rain when they were gassing up their vehicles. The architectural design to make it like a gas station came to around $30,000. We thought that was outra- geous. So somebody said, 'What about these bus stop covers -- the glass-enclosed ones?' We checked, and they cost $2,500. We put one of those up, and it works fine."

The District's budget deficit gives a new mayor the perfect opening to change the budget system. Normally the D.C. Council would hesitate to let go of its power to control line items. But if you take responsibility for cutting spending (and thereby inflicting pain) off its hands, it might go along. Bruce Babbitt cut a deal like this when he governed Arizona, and his managers loved it. Once they're rid of their line-item straitjackets, most managers will have no trouble finding ways to cut their spending by 5 or 10 percent. 7. Decentralize authority.

You campaigned for "decentralized school-based management initiatives," and you were right. Now apply that thinking to the rest of government. Once you have tied spending to out- comes, you can give your mid-level employees a lot more authority to innovate.

The Ford Foundation and the Kennedy School of Government, at Harvard, co-sponsor a series of annual "Innovation Awards" for state and local governments. The biggest surprise, to participants at the Kennedy school, has been the discovery that innovation rarely happens because someone at the top has a good blueprint. It happens because good ideas bubble up from employees who actually do the work and deal with the customers.

One of the first Innovation Award winners was a Minnesota program that finds innovators deep within the bureaucracy and gives them the authority they need. It's called STEP ("Strive Toward Excellence in Performance"). Every year employees who have innovative ideas compete to have them designated as STEP projects. The STEP board is co-chaired by the governor and a leading business executive. Its seal of approval does three things: It gives people permission to innovate; it offers them technical assistance; and it forces their bosses to sit up and listen.

One STEP team persuaded the Department of Natural Resources to change its entire attitude toward its customers. During the mid-1980s, use of the state's 64 parks was declining and budget problems were nibbling away at the parks. A group of people within the department decided that they needed a marketing program. They applied for STEP status and won. First they asked park managers to brainstorm about what they could do to make their customers happier; soon the managers were putting in children's play equipment in parks and electric hookups at campsites. Then they created a "Passport Club" -- a kind of frequent flier program for park users, to lure them to outlying parks that were not heavily used. Next they began accepting credit cards, running advertising and promoting park permits as Christmas gifts. Sales jumped 300 percent. Then they brought in a private company to improve their gift shops, and gift sales increased by 50 percent. Finally they conducted a customer survey of 1,300 park users.

During the first year after the marketing strategy took effect, the number of park visitors jumped by 10 percent. Numbers like these got the department managers' attention; in 1987 they created a "marketing coordinator" position and hired the STEP team leader to fill it.

Another STEP project, in the depart- ment that issues driver's licenses, cut waiting time for the public in half. (Similar innovation, of course, would be a godsend in D.C.) Yet another halved the paperwork involved in paying those who provide health care for handicapped children. STEP has since been copied not only by other states, but within the federal government.

The principle is very basic. In today's environment, things work better if those managing public organizations -- schools, public housing developments, parks, training programs -- have the authority to make many of their own decisions. Centralized, top-down institutions are simply not flexible enough to respond quickly to changing circumstances and customers' needs. 8. Push control of services out of the bureaucracy and into the community.

When professionals and bureaucrats have all the control, the people they are serving -- their "clients" -- have none. They are dependent. This is the situation with most public institutions: our schools, our police departments, our welfare agencies, our public housing authorities. It should come as no surprise when many clients learn dependent behavior -- so many that welfare dependency, for example, is among our most intractable problems. Treat people as dependents, and they will eventually become dependent.

The solution is to give communities control over services. Schools controlled at least in part by parents perform far better than schools controlled by "professionals," according to education research. Head Start centers run by parents make the greatest long-term impact on children's lives. The only gleam of hope in an otherwise bleak public housing landscape comes when residents manage or buy their own properties. Why? Because people act differently when they have some control over their own environment.

You have an inspiring example of this lesson right here in Washington. The Kenilworth-Parkside Resident Management Corp., chaired by Kimi Gray, has transformed one of the city's worst public housing developments. The dealers who once used Kenilworth-Parkside as an open-air drug market are gone. Teenage pregnancies have plummeted. More than 600 residents have gone to college. In five years, the crime rate fell from 12 to 15 crimes a month -- one of the highest levels in the city -- down to 2.

In 1986, the accounting firm Coopers & Lybrand released an audit of Kenilworth-Parkside. During the first four years of tenant management, it reported, rent collections increased 77 percent -- seven times the increase at public housing citywide. Vacancy rates fell from 18 percent -- then the citywide average -- down to 5.4 percent. The Resident Management Corp. helped at least 132 residents get off welfare. If these trends continued, Coopers & Lybrand concluded, the first 10 years of resident management would save the city $4.5 million.

You are undoubtedly familiar with Kenilworth-Parkside's success; in fact, you campaigned in support of "tenant ownership programs." It's time to replicate that success in the city's 59 other public housing developments -- and in dozens of other city services. 9. Choose prevention over cure.

Bureaucratic governments spend little time or money on prevention. Because they are programmed to think of government as service delivery, they typically wait until a problem becomes a crisis, then offer new services to those affected -- the homeless on the street, communities victimized by violence, school dropouts, drug users. Hence we spend enormous amounts treating symptoms -- with more police, more courts, more jails, more housing programs, more welfare payments and higher Medicaid outlays -- while prevention strategies go begging.

Entrepreneurial governments ap- proach problems very differently. Like Scottsdale, Ariz., they require sprinkler systems in all new buildings rather than paying for ever-larger fire departments. Like New Jersey, they help people before they lose their homes rather than building more shelters. Like Suffolk County, N.Y., they ban non-recyclable plastic packaging rather than creating new landfills.

Entrepreneurial governments also do everything they can to anticipate the future -- to give themselves radar. They practice strategic planning. They use "life-cycle costing," which details not just the initial costs of programs or purchases, but their maintenance and other long-term costs. They develop budget and accounting systems that force politicians to look at the long-term implications of all spending decisions.

Sunnyvale, Calif., exemplifies anticipatory government. Sunnyvale's budget projects the consequences of every decision out 10 years. If the City Council is deciding whether to repair a highway, the budget makes it clear that the cost will soon quadruple if nothing is done. If the council is deciding whether to buy land for a park, the budget shows what it will cost to staff and maintain the park for 10 years.

This process makes the long-term costs of decisions painfully clear to the press and the public. It has changed the behavior of the council. "In the right environment, all of the myths about how elected officials behave have come falling down for me," says City Manager Tom Lewcock. "They don't have to be short-range thinkers, they can be long-range thinkers. They don't have to be people who say, 'I know this is more important than that, but we're going to spend our time on that, because I've got constituents on my back.' They don't have to be any of those things." 10. Whenever possible, use market mechanisms rather than administrative mechanisms.

By changing private investment patterns, government can have a hundred times the impact it has when it uses normal administrative programs. Markets are not perfect, and they cannot solve every problem. But they are extremely powerful -- and when governments reshape them to serve a public purpose, they can have a tremendous impact.

In 1985, the year before Sandy Freeman became mayor of Tampa, almost a third of the city's housing units needed some degree of rehabilitation. That year the city's Community Redevelopment Agency, a traditional administrative bureaucracy with 41 employees, rehabilitated 37 units.

Freeman and her redevelopment director, Fernando Noriega, decided to use the marketplace to attack the problem. By guaranteeing loans for the first five years, they persuaded several of the city's banks to lend at low interest rates to owners who wanted to renovate their buildings. Last year the program rehabilitated 778 units.

Meanwhile, the Community Redevelopment Agency is down to 22 employees -- but according to Noriega, they are "happier, more productive and better paid employees." Obviously, they have a far greater impact than they did five years ago.

What Tampa did was simple: It used a market mechanism to achieve a public purpose. "We put the market together," Noriega says, "and presented it to the private lenders." You have talked about doing something quite similar with the $100 million "Capital Growth Fund" you want to create. Stick to your guns. If your opponents question your approach, tell them the story of the 30-year mortgage. In 1930, Americans who wanted to buy homes had to make a 50 percent down payment and repay the mortgage in five years. That's how banks did business. Using the new Federal Housing Administration, the Roosevelt administration pioneered mortgages requiring only 20 percent down and repayment over 30 years. Other government corporations created a secondary market, so banks could resell these loans. And the banking industry converted. Today we take our 30-year, 20 percent down payment loans for granted, because the federal government changed the marketplace. Ask your opponents: Would we be better off if FDR had created half a dozen low- and moderate-income housing programs?

A FINAL POINT: MANY OF THESE LES- sons could be summed up as a shift from administrative mechanisms to market mechanisms. Competition, funding tied to results, customer choice, investment -- all are characteristics of functioning markets. In a complex and rapidly changing world, in which our ability to access information is almost unlimited, markets are often more efficient and effective than administrative processes.

But market mechanisms are only half the equation. Markets are impersonal. Markets are unforgiving. Even the most carefully structured markets tend to create inequitable outcomes. That is why we need the other half of the equation: the empowerment of communities. To complement the efficiency and effectiveness of market mechanisms, we need the warmth and caring of families and neighborhoods and other social groupings. As we move away from administrative bureaucracies, we should embrace both markets and community.

Washington is an intensely political town, and in Washington people would call this moving right and left at the same time. They would label you a conservative for embracing markets and a liberal for empowering communities. But these ideas are not liberal or conservative. They can be used to implement any agenda. They can help you expand social programs if you choose, or shrink social spending and lower taxes if you choose. They dictate how government should work, not what government should do.

And regardless of what you want it to do, wouldn't it be nice if government worked again?

David Osborne is the author of a forthcoming book tentatively titled Reinventing Government: In Search of Excellence in the Public Sector.