Problem: Current methods of managing and evaluating management often lead to inefficient outcomes. Proposed solution and evaluation: A shift to outcome-based management might eliminate some problems. Outcome-based management is new in the public sector. Some U.S. cities have developed it over the past two decades; some states are beginning to; and foreign countries such as Great Britain, Australia, and New Zealand are on their way. Sunnyvale, California, a city of 120,000 in the heart of the Silicon Valley, began the experiment 20 years ago. In each policy area, the city defines sets of "goals," "community condition indicators," "objectives," and "performance indicators." "In a normal political process, most decisionmakers never spend much time talking about the results they want from the money they spend," says City Manager Tom Lewcock. "With this system, for the first time they understand what the money is actually buying, and they can say yes or no." Sunnyvale measures performance to reward successful managers. If a program exceeds its objectives for quality and productivity, its manager can receive a bonus of up to 10 percent. This generates pressure for ever-higher productivity. The result: average annual productivity increases of four percent. From 1985 to 1990, the city's average cost of service dropped 20 percent, in inflation- adjusted dollars. According to a 1990 comparison, Sunnyvale used 35 to 45 percent fewer people to deliver more services than other cities of similar size and type. At least a half-dozen states hope to follow in Sunnyvale's footsteps. Oregon has gone farthest. In the late 1980s, Governor Neil Goldschmidt developed long term goals, with significant citizen input. He set up the Oregon Progress Board, comprising public and private leaders, to manage the process. The board developed goals and benchmarks through 12 statewide meetings and written materials from over 200 groups and organizations. "Oregon," the board stated, "will have the best chance of achieving an attractive future if Oregonians agree clearly on where we want to go and then join together to accomplish those goals." The legislature approved the board's recommended 160 benchmarks, measuring how Oregon is faring on three general goals: exceptional individuals; outstanding quality of life; and a diverse, robust economy. Seventeen measures are deemed short-term "lead" benchmarks, related to urgent problems on which the board seeks progress within 5 years. They include reducing the teen pregnancy rates, enrolling people in vocational programs, expanding access to basic health care, and cutting worker compensation costs. Another 13 benchmarks are listed as "key"--fundamental, enduring measures of Oregon's vitality and health. These include improving basic student skills, reducing the crime rate, and raising Oregon's per capita income as a percentage of the U.S. average. Barbara Roberts, today's governor, has translated the broad goals and benchmarks into specific objectives for each agency. This year, for the first time, objectives were integrated into the budget--giving Oregon the first performance- based budget among the states. Great Britain has instituted performance measurement throughout its national government. In addition, the government has begun writing 3-year performance contracts, called "Framework Agreements," with about half its agencies. These agencies are run by chief executive officers, many from the private sector, who are hired in competitive searches and then negotiate agreements specifying objectives and performance measures. If they don't reach their objectives, the CEOs are told, their agencies' services may be competitively bid after the 3 years. Citation: 1993 National Performance Review, Chapter 3. Keywords: management efficiency