Boston Public Schools routinely overestimates future enrollment. Here’s why it matters.

Boston Public Schools consistently overestimates its future enrollment.Megan Lam/Megan Lam/Globe Staff; Adobe

Boston Public Schools officials predicted this month that enrollment would surpass 50,000 students in the next academic year.

But they’re way off. And they know it.

For proof, look no further than the annual census that the district — like all others in the state — took in October, which counted fewer than 49,000 students. And all signs indicate those numbers will probably decline further next year, as they have every year for most of a decade. No one — not even the district — is expecting a rebound.

So why the unrealistically rosy prediction? Because Boston, unlike most other districts, constructs its budget around what peak enrollment could be for all of its schools and classes, rather than that October student count or the average daily attendance. While other districts set aside money for the possibility of more kids than expected enrolling in various classes or schools, Boston bakes the high estimates into its budget and, until last week, did not go back to reconcile those estimates with actual student numbers.

The practice has been going on for years, and each year causes a wide gap between what the district plans and budgets for, and how many students actually show up. It translates this year into over $26 million allocated in places where it may not be needed. It adds to the millions the district spends propping up the budgets of dozens of schools that would otherwise face cuts to staff and programs due to falling enrollment. The inaccurate count also confuses the School Committee and the public.

In a period when enrollment has declined by almost 8,000 students, BPS has added more than 1,400 employees, justified in part by these generous estimates. More than 200 more positions will be added next year. Backed up by an even-worse long-term projection that wrongly predicted enrollment would rebound, the district put off hard decisions to close schools, even though it knew it already had excess space.

The district’s unusual methods for predicting enrollment — and with them, how it allocates funding — now face an unusually high-stakes test: a $2 billion overhaul of the district’s deteriorating facilities, including the construction or major renovation of at least 14 schools, the merging of others, and the challenge of making every single building meet the needs of students with disabilities and climate-friendly standards set by Mayor Michelle Wu.

Ultimately, the district’s enrollment projections and the corresponding school budgets are the responsibility of the School Committee and superintendent. The current superintendent, Mary Skipper, took the helm just months ago following a turnstile of leaders who oversaw the district when the calculations were made. She and the School Committee now face difficult decisions squaring declining enrollment with excess classroom space, rising needs, and the fast-approaching shutoff of a spigot of federal money.

In a statement, a spokeswoman for the district noted that enrollment is constantly changing over the course of the year, especially among populations such as immigrants and students from low-income families.

”If Boston just took a snapshot of our student population in October, it would not account for the students who arrive later in the year,” the spokeswoman, Gabrielle Farrell, said. “Our goal is to be a district that is ready to serve the highest-need students whenever they arrive. Any other enrollment projection would further create inequities and widen the opportunity and achievement gaps.”

Facing those stakes and following months of inquiries from the Globe, BPS on Thursday, during a budget hearing, for the first time published estimates of enrollment that look back on how their preferred way of counting has panned out over the last five school years, using peak enrollment by program type and grade range. The numbers were closer to district forecasts than the official October counts, but still significantly lower in four out of five years.

A new comprehensive plan for the system’s buildings is due in December, and yet BPS continues to use a custom student forecasting measure that analysts in other districts around the state and country say they have never heard of.

A reliable long-term forecast matters, if the last building plan is any indication.

Right now, the furthest-out the district has an estimate for is the school year 2025-2026, when it is predicting 56,000 students will be enrolled, not including pre-kindergarten. That number comes from a prior 10-year forecast, which was created in 2016 as part of the $1 billion BuildBPS construction plan and made no forecast for the district’s thousands of pre-K students.

But that forecast was too optimistic from day one. Consultants hired by the district predicted that despite years of stagnant enrollment, the schools would see gradual growth. BPS’s own analysts disagreed with the optimistic outlook contained in the 2016 plan, a district official said, but lost that battle.

That forecast is now off by over 10,000 kids: the district now has just 45,258 K-12 students, versus 55,591 in the forecast.

And it’s reverberated through the system. At the time, the district already had excess classroom space, so the prospect of more students meant the district wouldn’t have to close schools, a political minefield. Instead, the district could just shrink classroom capacity in existing schools, trading it for gyms, cafeterias, and other facilities.

A handful of schools have been closed since. But with enrollment continuing to decline, not increase as initially forecast, the amount of excess space is now vast. One recent analysis found the district has at least 16 schools too many.

Accurate projections are important, said Marguerite Roza of Georgetown University’s Edunomics Lab, because they give school communities more warning of impending school consolidations and closures and potentially more say in how they happen.

“Most districts delay way too long before doing anything about it, which makes it worse,” Roza said. “All this information feels like it comes out of nowhere, and people feel blindsided.”

Case in point: the district’s efforts to merge four schools into two have met fierce resistance from parents who question why their schools are being singled out. Recently, those mergers were officially delayed to 2024 and 2025 at the earliest. Hard decisions were pushed off in part due to a lack of clarity.

Skipper has said she is committed to making those sorts of hard decisions, but with no long-term forecast, annual budgets are the only window into central office thinking. That’s where Boston’s unusual methodology comes in.

The annual forecast is based on current enrollment. So, for example, the number of second-graders the district should expect next year begins with the count of the current first grade, layered with an assessment of how many usually return for second grade.

That part is normal; that’s what many districts use, based either on student censuses as of October or average daily attendance.

What’s unusual is what, exactly, Boston officials say they’re trying to estimate.

BPS projects a so-called high water mark, the peak possible attendance in every grade, program, and school, chief financial officer Nathan Kuder has said.

To simplify massively: Imagine the district had just two classes, and in each the population changed through the year, say between 10 and 20. Rather than calculating an average or tallying the students across the two classes on a particular date, Boston would simply take the peak number from each class and predict there will be 40 students the following year. That’s true even if some of the changes were caused by children switching from one class to the other — those students would be double-counted.

Kuder said Boston’s generous projections are needed to fulfill the teachers union contract, which has hard limits on class size. If the limit for the two classes is 15 students per teacher, then the district would budget for at least three teachers to accommodate the maximum possibility of 40 students.

In reality, BPS averages fewer than 11 students per teacher, and many classes are significantly underenrolled. In the 2021-2022 school year, around 700 students joined the district over the course of the academic cycle. That was an unusually unstable year — but the district’s forecast would still have been more than 2,000 students too high even if those 700 were added to the official count. (After projecting 51,642 students, the district ended up having just 48,654 that October).

In the retrospective estimates released last week, the district said enrollment by grade range and program type peaked at 50,486 last year, over 1,000 students below the forecast. In total, 52,040 students were enrolled at any point from Oct. 1 on. The disparities have not prompted BPS to overhaul the way it estimates enrollment numbers.

District officials say these intentional overestimates are necessary because Boston’s enrollment is particularly fluid and the system has many small programs. But other districts have different methods to manage those types of shifts, such as setting aside money to hire more teachers. In fact, Boston also has mechanisms in place to add or redeploy staff if needed due to enrollment fluctuations, officials said.

Boston’s method confuses even School Committee members. Member Brandon Cardet-Hernandez has repeatedly questioned why the projected number for this year was higher than last year’s official enrollment, when district officials were raising the prospect of enrollment declining.

“I‘ve never seen it done that way before,” Cardet-Hernandez said in a meeting last year. “It feels like the goal is to have an enrollment projection that, as a budget exercise, would be as close to your target as possible.”

The district already pads the budget at underenrolled schools to protect them from severe staffing and program cuts — $49 million this year in so-called soft-landing funds. Overly optimistic enrollment numbers can result in even larger sums redirected to low population schools.

With the new long-term plan due by December and the district rethinking how it allocates school funding, BPS has an opportunity to do its student count differently. But officials are still assessing how to produce a new multiyear projection. That decision is crucial as Boston prepares to spend billions on construction, experts said: if the district is to spend billions of dollars on refurbishing, replacing, and merging schools, it’s important it knows how many students those schools will need to serve.

“It is important that we have accurate projections as we go into master-facilities planning and that it is built into the process, separate from annual budget projections,” said Kerry Donahue of the Boston Schools Fund, an organization that funds public, private, and charter schools in Boston.

At a recent School Committee meeting, Skipper said there would have to be “right-sizing” of the district, driven by underenrollment and the fast-approaching end of federal relief funds. The new long-term plan gives district leaders an opportunity to finally make clear to families the student enrollment they actually expect over the next decade, as well as what they intend to do about it.

The Great Divide team explores educational inequality in Boston and statewide. Sign up to receive our newsletter, and send ideas and tips to [e-mail].