Shaking Things Up
|A broad new study of six cities that have shifted power away from the central office to educators in the schools says that none of them has gone far enough.|
In school districts from Los Angeles to Chicago, principals, teachers, and parents are drawing up budgets, hiring and firing employees, and shaping instruction. Their new authority is part of a push known as decentralization.
School boards in city after city have embraced the management philosophy that educators at the front lines are best qualified to decide how to teach their students, and that allowing them to make those decisions will enhance their involvement and commitment. By shifting money and power away from the central office, districts are gambling that they can win back public confidence and boost test scores.
Like many reform ideas that have come and gone in the past, however, much of this activity has taken place in the absence of proof that it works. And there has been little consensus on exactly how "decentralization" should take place.
So perhaps it shouldn't come as a surprise that an unpublished study by researchers at four universities--the most comprehensive analysis to date--suggests that, so far, decentralization has fallen far short of its goals.
None of the six cities they studied had dramatically raised student achievement. Most failed to envision all the changes needed for schools to gain true freedom and accountability. And all of them floundered when it came to implementation.
"Almost every superintendent and board member and union chief in the country is in favor of decentralization," says Paul T. Hill, a University of Washington researcher who is one of two principal investigators for the study. "But decentralization, as it's been tried so far, has been a half-measure because nobody wanted to make fundamental enough changes in the lives and incentives of adults."
Yet the researchers say their findings do not mean the idea should be abandoned. It just needs to be done right, and they believe they have identified the crucial elements that will help districts go about it.
"We remain convinced that decentralization is both inevitable and essential," they conclude. "If we have a general criticism to offer, it is simply that none of the sites pursued decentralization as aggressively and as fully as was needed."
The Annie E. Casey Foundation provided funds for the study, which brought together researchers from the University of Washington, the University of Chicago, the University of Colorado at Denver, and Pomona College in Claremont, Calif. The project tracked decentralization efforts in Charlotte-Mecklenburg, N.C., Chicago, Cincinnati, Denver, Los Angeles, and Seattle.
Different Starting Points
Each of these districts, the researchers found, began chipping away at their central bureaucracies from far different starting points.
A 1988 law gave Chicago's parents the upper hand in efforts to remake the schools. The law created parent-dominated, elected school councils at each of the city's 557 schools. The councils hire and fire principals, approve annual school improvement plans, and decide how to spend state anti-poverty money.
In contrast, Cincinnati began streamlining its central office in 1991 based on the recommendations of a business-led task force. And in Charlotte-Mecklenburg, a new superintendent hired in 1991 negotiated performance goals with each school and then held principals accountable for meeting them.
The only common feature of decentralization across the districts, the researchers found, was the ability of schools to request waivers from regulations they considered onerous.
The reforms in each city became more detailed and complex over time. But whatever the starting point, the researchers say, each of the plans suffered from three common weaknesses:
The authority and resources granted to schools were never clear.
"Almost every superintendent and board member and union chief in
the country is in favor of decentralization."
Paul T. Hill,
School-level educators weren't sure what decisions they could make without permission from the school board, the central office, or the teachers' union.
Budgets were never truly "site based" in any system. Most money still went to salaries and benefits for teachers and other fixed costs. Schools generally had little, if any, control over who worked in their buildings and only a small chunk of money to spend on school improvement.
Most plans started with only a vague idea of how schools would get help.
Most principals are not accountants or trained personnel managers. If they are to perform those functions, they need help. For decentralization to work, the researchers conclude, the central office cannot just go away, but must change its role to one of support.
"Most of the districts started with a chop-the-top philosophy," says Anthony Bryk, the other principal investigator for the study and a professor of education and sociology at the University of Chicago. In other words, they reduced the number of administrators who could have provided guidance and support to their counterparts in the schools.
Nor did the remaining central-office administrators relish a supporting role. "You still had central offices that thought of themselves as controlling what happens in schools," Mr. Bryk says.
The result: Many central offices continued to perform the same old regulatory functions, only with fewer people.
"Some schools moved forward because they were able to figure out what to do on their own," Mr. Bryk observes. "But probably many more could have moved forward under a more robust system."
Many schools struggled to make sense of a disjointed array of improvement efforts. And funding for retraining and staff development across the six districts remained modest.
"You're talking about changing the basic governance structure of
education. And that's not something a superintendent and an
elected school board will want to do."
Accountability remained the "black box" in most decentralization efforts.
None of the six cities fully changed the system of rewards and sanctions for educators to increase the focus on learning. Most sites are still weighing how to measure performance, what consequences will follow poor results, and who will make those decisions.
As early as 1993, "we had identified a significant subset of schools in Chicago that looked to us to be dead in the water," Mr. Bryk says. "But they basically could hide under decentralization because there was no viable mechanism to identify those schools and to intervene in them."
It wasn't until Illinois lawmakers passed another round of legislation in 1995 that a corporate-style district management team gained the ability to overhaul, or reconstitute, failing schools.
In each of the cities, the driving force for change most often came from the outside, the researchers found. Educators in each district often participated "hesitantly and sporadically, if at all," in the early stages.
In four of the six cities--Charlotte, Chicago, Cincinnati, and Los Angeles--the business community wielded the most resources and influence over decentralization. Private foundations also helped bankroll the early stages in Chicago, Cincinnati, Denver, and Los Angeles.
"Left on their own," the researchers conclude, "school system insiders are unlikely to think boldly enough to take enough risks to make a difference." Even now, the researchers note, most local decentralization efforts continue to rely on the energy, initiative, and money of outsiders.
In Chicago, for example, both the teachers' union and the administrators' union initially opposed efforts to restructure the bureaucracy.
"You're talking about changing the basic governance structure of education," explains Dorothy Shipps, the managing director of the Consortium on Chicago School Research, who was the project director for the six-city study. "And that's not something that a superintendent and an elected school board will want to do."
In looking at decentralization in the six cities, the researchers identified several trends:
- Gains in student achievement remained relatively modest. Although elementary schools in cities such as Chicago and Charlotte showed improvement, high school performance barely budged.
- In several cities, notably Chicago and Denver, the effort has encouraged teachers and principals to be more responsive to their communities. And it has helped some elementary schools create a professional culture that is more focused on learning.
- The roles of principals changed the most. In Chicago, principals lost tenure and had to work with the local school councils to keep their jobs. In Cincinnati, their pay was linked to performance.
In both Charlotte-Mecklenburg and Chicago, more than two-thirds of the principals retired, changed schools, or were fired in the years immediately after decentralization. In schools that made dramatic changes, strong building leaders typically became entrepreneurs, seeking out new ideas, money, and assistance.
- But the work lives of many teachers remained untouched, even though they were positive about the opportunities afforded them. Although some teachers gained more say over issues such as curriculum and staff development, only a minority of teachers assumed dramatically new roles in schools.
- The support of ethnic and racial groups for decentralization has been mixed. African-Americans in all six cities were conspicuously divided about whether decentralization was a good idea.
Hispanics appeared to benefit from the transfer of power in some cities. In Chicago, the number of Hispanic principals more than doubled in the year after decentralization, from 17 to 43. In Los Angeles, the number of Hispanic teachers jumped from 3,991 in 1991 to 5,361 four years later. Nonetheless, Asian-Americans and Hispanics typically were not leading actors in decentralization in most cities.
A Fragmented Landscape
"Decentralization requires work, threatens established interests, and creates conflict," the researchers found. Not all cities were equally well-positioned to sustain their efforts.
"Some schools moved forward because they were able to figure out
what to do on their own. But probably many more could have moved
foward under a more robust system."
At best, the researchers say, decentralization creates the conditions that allow schools to improve one at a time. But it doesn't guarantee it. Districts with a cohesive political structure and a vital civic sector appeared more likely to stick with their efforts to decentralize.
The clearest contrast was between Chicago and Los Angeles. The Windy City has a history of centralized, machine-dominated politics, and it also has a solid base of civic and business organizations that helped craft the reform plan.
"The Chicago reform, as it turns out, had a much better chance of actually coming to fruition," Ms. Shipps says.
In contrast, politics and government in Los Angeles are much more fragmented. The 625,000-student Los Angeles Unified School District encompasses portions of 29 municipalities outside the city. Competing power centers--and competing priorities--have led to spotty implementation of its decentralization plan, known as LEARN. ("Second Thoughts About LEARN Surface in L.A.," May 28, 1997.)
The role of the teachers' union also varied. In Los Angeles, Cincinnati, and Seattle, the unions supported decentralization. But the price has been a slower, more phased-in approach. Although Chicago was able to move faster, Ms. Shipps says, it must now reach out more aggressively to teachers and principals.
If communities want decentralization to work, the researchers say, they must pay attention to autonomy, assistance, and accountability:
- School-level educators must control the checkbook. They need control over hiring, evaluating, and firing staff members. And they must be free to pursue different instructional strategies. Parents should be able to choose among schools.
- Schools should be free to select help from a range of public and private sources. States and districts should not attempt to deliver "one-size-fits-all training and assistance."
- Districts and states should nurture a "rich system of school-specific accountability." That means new forms of testing, performance agreements between schools and local school boards, and real consequences for schools that fail to educate children.
None of the six cities they studied, the researchers conclude, has done all of these things adequately.
"We're hoping that we're going to push the conversation forward about what it means to decentralize," Mr. Bryk says. But, he acknowledges, "this is a long-term process.
Vol. 17, Issue 02, Page 29-31Published in Print: September 10, 1997, as Shaking Things Up