The Agony of American Education

How per-student funding can revolutionize public schools.


Imagine a city with authentic public school choice—a place where the location of your home doesn't determine your child's school. The first place that comes to mind probably is not San Francisco. But that city boasts one of the most robust school choice systems in the nation.

Caroline Grannan, a public school advocate and super-involved parent, lobbied hard to wear down the San Francisco school district back in 1996 and get her son William, then an incoming kindergartner, out of his assigned neighborhood school, Miraloma Elementary, and into a "more desirable" alternative school called Lakeshore. In 1996 Miraloma had low test scores and a low-income student body bused in from other neighborhoods; its middle-class neighbors shunned it. Lakeshore had a better reputation and higher student performance.

Once, Grannan remembers, it was conventional wisdom in San Francisco that there were only five decent public schools in the city; if you couldn't get your child into one of them, it was time to move to the suburbs or to find a private academy. But a lot has changed since 1996. Today Grannan could send her child to any school within the city. What's more, she would happily send her kids to Miraloma, one of many elementary schools in San Francisco that now attract eager middle-class clients. Miraloma has a new principal with a parent-friendly attitude, has begun to raise its test scores, and is more diversified. Families now feel secure taking advantage of Miraloma's longstanding positive attributes, including its small size and its sheltered and attractive setting.

Grannan's more recent experience with her children's middle school also reflects how San Francisco schools have changed. Her son William just graduated from Aptos Middle School, and her daughter Anna started sixth grade there this year. This school is now in high demand, but in 1996 parents considered it dirty, dangerous, and academically weak. Today it offers enriched language, arts, and music programs, and its test scores continue to improve.

Grannan is more than just a concerned parent. She is a founding member of the San Francisco chapter of Parents for Public Schools, a PTA board member, and a prolific writer whose articles about local schools appear in the San Francisco Examiner and other publications. She has argued passionately against both vouchers and charter schools, and would wince to be portrayed as a partisan of school choice. Yet she has become an avid supporter of the San Francisco system and the benefits it brings to San Francisco families.

San Francisco is one of a handful of public school districts across the nation that mimic an education market. In these districts, the money follows the children, parents have the right to choose their children's public schools and leave underperforming schools, and school principals and communities have the right to spend their school budgets in ways that make their schools more desirable to parents. As a result, the number of schools parents view as "acceptable" has increased greatly in the last several years. In Grannan's words, "Parents who are willing to go beyond the highest-status schools can now easily find many more acceptable options, and can avoid the fight for a few coveted seats in the most prestigious schools."

Decentralization Rules

Give credit to Arlene Ackerman, San Francisco's superintendent of schools since 2000. Ackerman introduced the weighted student formula, pioneered in Edmonton, Alberta, in 1976, which allows money to follow students to the schools they choose while guaranteeing that schools with harder-to-educate kids (low-income students, language learners, low achievers) get more funds. Ackerman also introduced site-based budgeting, so that school communities, not the central office, determine how to spend their money. Finally, she worked to create a true open-enrollment student assignment system that gives parents the right to choose their children's schools.

In San Francisco the weighted student formula gives each school a foundation allocation that covers the cost of a principal's salary and a clerk's salary. The rest of each school's budget is allocated on a per student basis. There is a base amount for the "average student," with additional money assigned based on individual student characteristics: grade level, English language skills, socioeconomic status, and special education needs. These weights are assigned as a percentage of the base funding. For example, a kindergartner would receive funding 1.33 times the base allocation, while a low-income kindergartner would receive an additional 0.09 percent of the base allocation. In 2005–06 San Francisco's base allocation was $2,561. Therefore, the kindergartner would be worth $3,406, and the low-income kindergartner would generate an additional $230 for his school.

The more students a school attracts, the bigger the school's budget. So public schools in San Francisco now have an incentive to differentiate themselves from one another. Every parent can look through an online catalog of niche schools that include Chinese, Spanish, and Tagalog language immersion schools, college preparatory schools, performing arts schools that collaborate with an urban ballet and symphony, schools specializing in math and technology, traditional neighborhood schools, and a year-round school based on multiple-intelligence theory. Each San Francisco public school is unique. The number of students, the school hours, the teaching style, and the program choices vary from site to site.

The pressure to attract children has produced not just a greater variety of unique schools but new school capacity based on the specific demands of parents. For example, as demand has exceeded the number of available seats the district has added more Chinese and Spanish dual-language immersion programs. The weighted formula ensures that schools have an incentive to recruit and serve students with learning disabilities, limited English proficiency, and other difficulties.

All this diversity is useless if parents don't know about it, so schools have an incentive to market their programs as well. Much of the marketing is done through a local chapter of Parents for Public Schools. The district and the chapters host school enrollment fairs, and the schools offer parent tours throughout the school year. Parents can select up to seven schools on their enrollment application. In the 2005–06 school year 84 percent of parents received one of the schools they listed, with 63 percent receiving their first-choice school. More than 40 percent of the city's children now attend schools outside their neighborhoods.

Decentralized school management is a growing trend in the United States. To date the weighted student formula has been implemented in Cincinnati, Houston, St. Paul, San Francisco, Seattle, and Oakland. This year a weaker version that does not include school choice is being implemented statewide in Hawaii, and pilot programs are underway in Boston, Chicago, and New York City.

By contrast, most districts in the United States use a staffing ratio model, in which the central office directs school sites to spend their resources in a particular way, through allocations of staff and a small supplies budget. For example, a school might be sent one teacher for every 28 students. This system gives individual institutions little control over their financial resources and personnel choices. Under the weighted student formula, each school site receives a budget denominated in dollars instead of positions and decides what staff and nonstaff items to purchase with that money.Oakland, which completed its first year of the weighted student formula in 2004–05, is taking the decentralized concept further than any district in the United States. Edmonton, San Francisco, and the others all charge each school not for the actual salary of each teacher but for "average teacher salaries" in the district. This means that, for the sake of school budgets, differences in teacher salaries are ignored; on paper, a first-year teacher costs the same as a 30-year veteran. This practice hides funding inequities within districts where more desirable schools are stacked with senior teachers and other institutions are staffed with less experienced instructors. In practice, schools with lower-paid teachers end up subsidizing schools with higher-paid teachers. In Oakland, by contrast, schools are charged the actual cost of their employees, so a school with more novice educators has more money left over to pay for training or supplies or even to hire another teacher and reduce class size—all of which could make a school more attractive to potential students.

Another way some districts go further than San Francisco is in the extent to which parents are allowed to choose their children's schools. Edmonton's system is particularly robust, allowing students to apply directly to any school in the system. Similarly, Cincinnati's high school open enrollment system allows students to apply directly to 26 different high school programs on a first come, first served basis. Such systems stand in stark contrast to the form of choice embedded in the federal No Child Left Behind Act. Under federal law students in failing schools are guaranteed the right to transfer to a school that isn't failing. But districts have not made a good-faith effort to implement public school choice. In New York City this year, for example, 11,000 kids applied to leave failing city schools, but only 2,250 city kids received one of their choices. Since the No Child Left Behind Act was passed, fewer than 2 percent of parents nationwide have used the law's provisions to transfer their children to other public schools.

School closure is another prominent feature of the weighted student formula model. In Edmonton, if a school declines to the point that it can't cover its expenses with the per student money, the principal is removed and the remaining teachers and facilities are assigned to a strong principal—or the school is closed altogether, and the staff is moved to other, more successful schools. The San Francisco school district closed five schools in 2005 because of underenrollment and is considering closing or consolidating 19 other schools.

Lifting All Boats

San Francisco's system produced significant academic success for the children in the district. Miraloma Elementary, the school Caroline Grannan would not consider for her children in 1996, has seen test scores for second-graders in English language improve from 10 percent proficient in 2003 to 47 percent proficient in 2005. "Now's the time to get in on the ground floor of one of the most up-and-coming schools in San Francisco," one Miraloma parent recently wrote in an anonymous review for greatschools.net. "Student achievement is rising, parent involvement is soaring and the entire community is working very well together to improve the quality of every aspect of the school.…Parents are moving their kids from private schools to Miraloma because they like what they see. Yes, there is still work to be done but I am very confident that Miraloma will be the next Rooftop or Alvarado." (Rooftop and Alvarado are two previously average schools that are now considered top-notch by parents due to high student achievement.) Greatschools.net had 19 similarly positive reviews for Miraloma.

Similarly, at Aptos Middle School, where Grannan's daughter started this year, the share of students scoring proficient in English language increased from 29 percent in 2002 to nearly 50 percent in 2004–05. Aptos is also the most ethnically diverse school in the district: Its demographic composition in 2004–05 was 26 percent Hispanic, 32 percent Asian, 19 percent black, 13 percent white, 6 percent Filipino, 3 percent multiracial, and 1 percent Native American. Close to 50 percent of the students participate in the federal free lunch program, which is the standard proxy for poverty in public schools—schools with large free lunch populations generally have a more difficult time with academic achievement. California's academic performance index (API) ranks a student body's performance on several standardized tests. Aptos' score has just risen from 6 out of 10 to 7 out of 10 (10 is best); it ranks 8 out of 10 when compared to schools with similar demographics.

Such gains have been made throughout the school district. Every grade level in San Francisco has seen increases in student achievement in math and language arts, and the district is scoring above state averages. (Fifty percent of San Francisco seventh-graders were proficient in language arts in 2005, compared to 37 percent proficiency statewide.) Even high schools, the most intractable of all schools, appear to be improving. Mission made Newsweek's 2005 list of the nation's top 1,000 high schools. Galileo has shown a big jump in test scores—its statewide API ranking jumped from a 3 to a 6 in just one year, while its ranking compared to similar schools climbed from a 2 to an 8. Balboa is on the radar for families who never would have considered it a few years ago.

These gains have been made even as students who used to be excluded from standardized tests are increasingly being tested. In the last year of Superintendent Bill Rojas' administration, 1998–99, only 77 percent of the district's students in the tested grades were included, with kids who were deemed likely to bring scores down left out whenever possible. In 2003–04, 98 percent of students in the tested grades were included.

San Francisco is not alone. William Ouchi of UCLA's Anderson School of Management has done extensive research on the effects of school district decentralization throughout the United States. Ouchi and his team of 12 researchers studied three very centralized public school districts: New York City, Los Angeles, and Chicago; three very decentralized public school districts that used the weighted student formula: Seattle, Houston, and Edmonton; and three very decentralized Catholic school systems: Chicago, New York City, and Los Angeles. In his 2003 book Making Schools Work, Ouchi found that the decentralized public school districts and private Catholic schools had significantly less fraud, less centralized bureaucracy and staff, more money at the classroom level, and higher student achievement.

He also found that most districts merely give lip service to local control. According to Ouchi, the money must follow the child. The only true local control occurs when the principal controls the school budget.

At John Hay Elementary School in Seattle, which Ouchi profiled, the principal controlled about $25,000 a year before decentralization and now controls about $2 million. The principal used her new freedom to hire 12 part-time reading and math coaches and set up a tutoring station outside every classroom, plus another station in a wide hallway, for "turbo-tutoring" the gifted children. Now the school teaches reading in groups of five to seven students while other classes are in larger sections, and every student who is behind grade level receives one-on-one tutoring.During a four-year period following the change, the school's standardized math scores rose from the 36th percentile to the 62nd, and reading scores rose from the 72nd percentile to the 76th. In third grade, black and white students now have identical reading scores, and all of them are at or above grade level.

Such gains also occur in other districts that have implemented public school choice and the weighted student formula. After Oakland's first year of student-based budgeting, a majority of the city's African-American students met basic reading standards at their grade levels in 2005—probably a first in recent district history. In addition, every grade level in Oakland saw increases in the number of students who were proficient in reading and math. Similarly, in 2005 Cincinnati public schools, where 70 percent of students are African-American, improved their state rating from "Academic Watch" to "Continuous Improvement," and test scores were up for most students in most grade levels. Seattle also continues to see increases in student achievement and in 2005 reduced the number of schools rated "failing" under the No Child Left Behind Act from 20 to 18, even as the state raised the bar for proficiency.

As a result of these changes, parents are returning to public schools. In Seattle, the public school district has won back 8 percent of all students from the private schools since implementing the new system. In Edmonton, where it all began, the public schools are so popular that there are no private schools left. Three of the largest private schools voluntarily became public schools and joined the Edmonton district. (This has not held true in San Francisco, where families continue to leave the city, largely because of high housing costs. San Francisco's private schools have lost enrollment as well, as the city's child population reaches an all-time low of 11 percent.)The

Constraints of Public School Choice

Public school choice is not a panacea. In many districts there have been tensions between parents who want more choices and parents who want their children to have a guaranteed spot in a neighborhood school. In Seattle, the district recently considered abolishing the school choice system in favor of the traditional system based on a child's address. The district's reasoning is that busing students all over Seattle is complicated and expensive. So far, a parental outcry has staved off the plans to return to residence-based schools. Parents have suggested charging for transportation or leaving it up to families rather than killing off school choice.

In addition, unlike an actual market system in education, public schools are still strapped with myriad local, state, and federal regulations. No matter how decentralized San Francisco schools become, they still must comply with the No Child Left Behind Act and abide by silly state laws, such as the California statute that forbids parents from bringing home-baked cupcakes to school to celebrate their children's birthdays with classmates.

Public school choice is at best a weak substitute for true school choice, where parents are not bound by excessive government regulations. In support of this point, Ouchi's research found that the three Catholic school systems he examined—Chicago, New York City, and Los Angeles—were the most decentralized. They have very small central staffs, spend the least money per pupil, and have the highest student achievement. (While demographics do not affect the per-pupil spending or smaller centralized staff in Catholic schools, they probably contribute to higher test scores. For example, the New York City Catholic schools in Ouchi's study have only 32 percent low-income children, compared to 74 in the city's public schools.)

Ouchi's findings reinforce the main criticism of decentralized public schools: Is it really necessary to stay within the bounds of the existing public school system and complete the difficult task of changing the system from within? A better alternative would be to move to a direct financing mechanism through vouchers, tax credits, or charter schools—an arrangement under which per-pupil funding immediately empowers parents and leads to the most decentralized schools of all, with 100 percent local budget control.

Yet the better alternative is not always the politically feasible alternative. School decentralization offers a compelling model for restructuring school financing, giving principals and parents true control over their schools, and offering real school choice to all students within the constraints of a public school system. It also gets parents used to the idea that schools need not be linked to real estate. And it demonstrates that even within a limited pseudo-market, when families become consumers of education services with the right of exit, schools quickly improve to attract them.

The San Francisco parents I spoke with probably would be alarmed by the market metaphor. In general, these parents do not support education tax credits or school vouchers. They are for public education. Yet San Francisco has adopted a school district financing system that mimics a school market and has led to a revitalization of the city's public schools. And these parents have taken full advantage.

Caroline Grannan admits she probably could have worked the old residential assignment system to get her kids into good schools. But times have changed in the City by the Bay. When Grannan's son William was applying for high schools, she was one of many middle-class parents now willing to send her child to Balboa High School, which not long ago was viewed as a low-performing, dangerous "ghetto school." William ended up going to SOTA, the School of the Arts, to which students are admitted by audition. But as Grannan says, "Knowing that we were fine with Balboa if he hadn't gotten into SOTA made the entire process much lower-stress." The difference, she says, is "the comfort in knowing that parents have more than one option."