Feature

Declining Recources, Targeted strategies

Making the best of a shrinking budget by connecting spending to a district’s vision and what matters most by NATHAN LEVENSON

As resources shrink, the need to do more with less becomes critical. As a business CEO turned superintendent, I have seen firsthand that many options exist, but none are fun, easy or politically rewarding.

In selecting me, the board of education in Arlington, Mass., hired a nontraditional superintendent. A public announcement said I was the man to navigate a tough financial situation and bring new ideas to the school district. Having not come up through the ranks made it easier to see ways to improve learning while lowering expenses. It also meant I couldn’t see many of the political land mines.

Nathan LevensonAs superintendent in Arlington, Mass., Nathan Levenson identified student literacy as a target for beefed-up funding. Photo by Richard Viard



With contract signatures still damp in 2005, I launched a 50-person listening tour with school committee members, central-office staff, curriculum leaders and principals. The conversations revealed three truths:

•  The school board lacked a plan, vision or strategy, but it controlled the budget;

•  The administrators knew what was needed to raise student achievement, but no one asked them; and

•  No one, including the superintendent, chief financial officer and school board, really knew where we spent our money.

The district spent about $50 million a year, roughly $11,000 per pupil, which was reasonable, but the trends were problematic. Enrollment was growing, wages and benefits were growing even faster, and tax revenue was capped. This combination meant we faced a structural deficit each year. Budget cuts were a given.

Optimum Targets
With this backdrop, I set out to do more with less. The effort had phases: (1) clarify a strategy; (2) allocate resources to the strategy; (3) foster teamwork; (4) think creatively about funding sources; and (5) take a fresh look at the district’s approach to special education. It turned out we had enough money; we just spent it on the wrong things.

An old Russian proverb teaches, “If you don’t know where you are going, any path will take you there.” The Arlington Public Schools, with its 4,500 students, needed a strategy for raising student achievement. We had a collection of programs, purchased curricula and pet projects supported by influential staff, but no coherent strategy.

After much internal and community debate, we reached broad agreement that standards-based education, a common curriculum, districtwide formative assessments, an unrelenting focus on reading, lots of accountability, and sustained, targeted professional development would help the students of Arlington. This was a comprehensive, interconnected strategy, not a bunch of unrelated programs.

While the final result was a clean, thoughtful plan, the debate was anything but. The public discussion quickly turned nasty and laid bare deep divisions within the community. An influx of better-educated residents wanted gifted programs, engineering courses and college for all. Many long-term, more blue-collar residents fought to to preserve the jobs of veteran employees in home economics (who had tiny teaching loads), crossing guards (some of whom stood at corners with no children passing by) and long-term administrators (even though all their staff had been let go).

In the past, sufficient budgets meant difficult choices about whether to cut underenrolled courses and underutilized staff were avoided because there were enough new funds to support them. Going forward, our analysis showed no new money was coming. In fact, inflation-adjusted per-pupil spending was projected to decline over the next few years. Only one way existed to implement our new plan: We had to shift spending from existing efforts. The good news was we had enough money if we spent it on what mattered most to children.

To prioritize our limited resources, we needed to understand, really understand, our current spending. This meant recasting the budget down into a meaningful level of detail. Some line items contained only $200, one had $1 million marked “other,” and a few combined a half dozen unrelated programs. Because the business office staff seldom spoke to frontline staff or building principals, they had to guess where to charge many expenses. It took five months and 10 revisions to create a clear, straightforward listing of all prior expenses by activity and program, encompassing grants and revolving funds, as well.

Yet with the new strategy in hand, we now knew what was needed: more reading teachers, paid mentors, elementary math coaches, a better student information system, a data analyst and stipends for teachers to develop the common curriculum and districtwide formative assessments.

Winners and Losers
Having documented the school district’s spending, it was time to support our priorities and trim everything else. I first tackled the administrative staff. We had far too many noncore-subject department heads. Past efforts to be respectful meant every department, even those with few or just one staff member, had an administrative leader.

The next target was federal grants. While Arlington was meticulous in following the letter of the law, nearly all these funds were frittered away on ineffective programs, secretaries for the grant owners and nice-to-have programs, completely unconnected to the core standards-based curriculum.

Between restructuring the administrators and shifting grants to fund the strategy, we had enough for everything we wanted, including expanded mentoring, new reading coaches, many more reading teachers, staff-developed standards, teacher-created common assessments, and greatly expanded professional development.

After about a year, a subtle shift took place in the district. As curricula were standardized, reading became centrally managed and professional development focused on supporting the strategy, many decisions affecting the entire district came before the leadership team. The problem, however, was that the leadership team wasn’t really a team. It was a group of smart, caring people who viewed their jobs as protecting their turf, rather than thinking about the good of the whole district.

The vehicle for creating teamwork was the budget. In the past, the budget had been done to the leadership team. I needed the team to do the budget — to take the lead, make prudent, cost-effective decisions, find savings and sell the plan to others.

Turning budget development into a team exercise took more than 50 hours of meetings, an outside facilitator, training in budgeting and a constant reminder of the school district’s vision. These could have been 50 useless hours, but, because we shared detailed student achievement and staffing data, we were able to avoid turf battles and use the conversations to shift resources to the most effective and cost-effective programs.

Perhaps the most powerful outcome of collective budgeting was reversing the incentive system. My 16-member cabinet started to think about the districtwide master plan, not their parochial interests. Because the administrators now had meaningful say in how money would be spent, they wanted to find ways of freeing up funds. This dramatically improved the effectiveness of our budgeting and allowed less money to achieve more.

During my second year, I realized there was a limit to how much we could shift and cut. The district had enough money to meet the academic needs of our students, but I knew we were underserving their social and emotional needs.

I created a task force to research the local services available to students and families seeking counseling. Focus groups with local mental health providers and clinics revealed many counselors who serve children don’t have much business during the day, and older children don’t want to go after school. Within two years, we had partnered with five different organizations, providing over $1 million of services a year to our students at almost no cost to the district.

Buoyed by the success of our mental health partnerships, the district enlarged the concept to expand programs and raise revenue. Collaborating with a local interfaith group, we solicited bids for someone to expand a fee-based after-school program (with scholarships for children who couldn’t pay). Through third parties, we added summer programs, day care for our staff, enhanced preschool programs and a foreign-student-exchange initiative.

These alliances provided services we could never have afforded on our own and we charged market-based rent for most of them to be housed in our schools. This raised substantial revenue to fund the strategy.

Special Education
Any plan to do more with less needed to address special education. From the day I arrived in Arlington, I knew our students with special needs achieved at unacceptably low levels, despite ever-increasing resources.

By replacing common practices with best practices, we achieved impressive gains by rethinking special education.

•  The number of special education students scoring proficient or better on the state accountability test increased by 26 percent in English and 22 percent in math over three years in the schools participating in the reform program. Achievement declined in the control group.

•  The special education achievement gap at the high school declined by 65 percent. 

•  The number of formal parent complaints to the state Department of Education dropped from 25 to zero, and parent satisfaction, as measured by an independent survey company, increased substantially.

All while per pupil special education costs grew by less than 2 percent per year, well below the rate of salary increases. In the past, costs had been growing by 10 percent a year.

Revamping Reading
We found that increasing student achievement would actually lower costs. This is true for two reasons. First, nothing is more expensive than a child not learning. If you help students reach grade level quickly, you save years of future remediation costs. Second, we were spending substantial sums of money on ineffective programs. We didn’t need more money; we needed programs that worked better.

In Arlington, like most districts, the No. 1 reason students enter special education is difficulty with reading. Nationwide, up to 40 percent of all students on individual education plans have reading problems, and 80 percent of students with the most common disability diagnosis, specific learning disability, simply haven’t learned to read, according to the Rennie Center for Education Research & Policy in Cambridge, Mass.

At the start of my second year, we ended five ineffective reading programs and placed all reading staff under one leader. All teachers — classroom, remedial and special education — used the same program and pacing. Principals were evaluated based on reading results in their building, and we mandated the use of the National Reading Panel recommendations, including 30 minutes a day of supplemental small-group instruction for all struggling readers. Teachers had a great deal of input into designing and adjusting the districtwide plan, but they couldn’t individually change it.

This worked better than I could have hoped, with these outcomes:

•  The number of students K-5 not reading at grade level dropped by 65 percent;

•  Two-thirds of K-2 students who started the year behind in reading made more than a year’s progress; and

•  Referral rates to special education dropped. Within three years, reading ceased to be a special education service for nearly all children.

Despite the remarkable success, many veteran teachers felt the new efforts were an implicit criticism of their past work.

Cost Neutrality
The success in elementary reading encouraged us to look at math and English at the secondary level. One day in a special education resource room, I watched a bright, caring, passionate veteran teacher stand at the board and try to explain math to one student, English to another, biology to a third and U.S. history to a fourth.

Several thoughts hit me at once: (1) We would never allow this teacher to teach any of these subjects to general education students, for she wasn’t certified in any of these fields; (2) All the students sitting in front of her had already been taught that day by a certified teacher and still they struggled; and (3) Not a single general education math, science or English teacher would agree to teach outside his or her field, yet we expected special educators to teach all subjects. None of this made sense to me, yet I knew this was the norm in most districts across the country.

We implemented a common-sense plan. Math and English remediation became open to all students (general education and special education alike), taught by general education teachers with expertise in the subject matter. The classes were 50 to 100 percent longer, and the teachers were handpicked, similar to how we selected skilled teachers for Advanced Placement classes.

These programs were cost-neutral in the short run because they required only shifting resources from special education to general education. In the long run, they will decrease costs by reducing the need for future services. High school test scores for students with special needs rose so dramatically that Arlington was profiled by the Rennie Center for Education Research & Policy as one of the fastest-improving districts in the state.

This strategy created some pushback. Special education teachers felt devalued when their numbers decreased as we shifted to content experts. Hurt feelings did take a toll on morale. Yet despite reduced special education staffing, the special education achievement gap between 2004 and 2008 at the high school declined by 65 percent.

Para Staffing
Paraprofessionals were another big expense. Conventional wisdom was “the more, the better,” and the budget reflected this thinking.

But overuse of paraprofessionals can be detrimental to many students. They hover beside their assigned child, creating a social barrier and stifling peer interaction, which defeats one of the primary benefits of inclusion. What’s more, having a paraprofessional often decreases the instruction a student receives from the classroom teacher, who feels a student with a teaching assistant already has 100 percent of another adult’s time.

To determine whether a paraprofessional was really necessary for a given student, we followed a straightforward process. First, we made sure aides weren’t being assigned for the wrong reason, such as parental pressure.

Second, for those students who did require paraprofessionals, we were specific about the needs. If a student struggled in reading, we provided reading help, not an aide. If a student had behavior problems, we hired a behaviorist to teach the student coping skills, and if a student had trouble transitioning to school, the student received paraprofessional support in the morning only.

Finally, we scheduled the aides on a buildingwide basis, rather than student by student, to more effectively share aides who now can change classrooms throughout the day.

Unlike for classroom teachers, we had few rules for workloads for related-service therapists.

Speech and language, occupational therapy and counseling are valuable services, but they need staffing guidelines. We used benchmarking, the process of comparing how others do the same things we do. It turned out that despite desperate pleas for more therapists, our staffing levels exceeded like communities by 25 percent.

What was wrong? We weren’t good at scheduling. One therapist scheduled 25 percent of her time to escort students to and from class, while others had the students come to them. We also realized that staff were assigned to buildings, not by student caseload. If school A really needed 1.25 therapists, and school B needed 0.75 therapists, each got 1.0. We moved all scheduling to one administrator. Because she was a peer of the principals, she could say, “We need Mary and John at the same time. Let’s find a way to work this out.” The net effect was that we increased effective capacity by 25 percent. As a result, we added new in-house programs without hiring more staff.

We also learned we could run our own top-quality programs for students with significant special needs for 40 percent less per student than the out-of-district programs we used. The savings came from less transportation (much shorter rides), lower facility costs (we owned the buildings) and no marketing costs. Many parents preferred their children attending school in the community, so these new programs were popular. Over the next few years, Arlington opened eight programs and saved more than $5 million cumulatively.

Unavoidable Backlash
By focusing like a laser on our strategy and not funding programs and staff who were no longer part of the plan, a typical middle-class school district had enough money to serve all students well.

Looking back, I’m proud of our results, but battered and bruised from the process. A relatively small, but mean-spirited, group of protesters regularly packed school board meetings, sent anti-Semitic notes, e-mails and voice mails, made threats against my family, and whispered ugly, fabricated innuendo to discredit me, staff and community members who supported the reforms.

Why the backlash? There is no nice way to cut administrators, many of whom lived in town. No kind way to end ineffective programs. No painless way to cut staff in one area to fund new efforts, and no one enjoys losing authority over spending grant money. A Boston Magazine article published more than a year after I had left the school district in 2008 concluded: “In the end, it turned out that Levenson had made the mistake of thinking that when Arlington residents said they were ready for change, they actually meant it.”

I’m often asked, “Would you do it again?” or “What would you do differently next time?” The more important question, however, is, “What would make it less difficult to make the same decisions next time?”

The point of creating the strategy was to prioritize our time, energy and resources. Our limited funds should be focused on the strategy, and administrators should be accountable for implementing the plan. While many lauded this, most failed to understand that given limited resources we could not continue to spend money on what wasn’t critical to the strategy. The reality is that you can’t do more with less unless you make hard, thoughtful budget decisions, rather than across-the-board cuts.

The challenge of managing declining resources isn’t the choice between doing less for children or discovering new revenue. It is finding the political skill, cover and buy-in to do what matters most for students, even at the expense of the adults. It is not easy, but it was the job I signed up for.

Nathan Levenson, former superintendent in Arlington, Mass., is managing director of the District Management Council in Boston. He was a contributor to Stretching the School Dollar (Harvard Education Press). E-mail: nlevenson@dmcouncil.org